Evaluation of SVM and RF Machine Learning Algorithms in Land Use/Land Cover Change Assessment: Tessa Watershed Case Study (Northwest of Tunisia)

Evaluation of SVM and RF Machine Learning Algorithms in Land Use/Land Cover Change Assessment: Tessa Watershed Case Study (Northwest of Tunisia) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Evaluation of SVM and RF Machine Learning Algorithms in Land Use/Land Cover Change Assessment: Tessa Watershed Case Study (Northwest of Tunisia) Noamen BACCARI, Mohamed Hafedh HAMZA, Tarek SLAMA, Abdelaziz SEBEI, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4359112/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Jan, 2025 Read the published version in Earth Systems and Environment → Version 1 posted You are reading this latest preprint version Abstract This study aims to highlight the changes in LULC (land use and Land cover) in the Tessa watershed (Northwest of Tunisia) between 1993 and 2023. Remote sensing coupled with geographic information systems (GIS) and statistical analysis, are used. Accuracy metrics make it possible to evaluate the performance of the image classification method, using the calculation of the producer’s accuracy, the user’s accuracy, overall accuracy, and the Kappa coefficient. Two Machine Learning (ML) algorithms related to the supervised classification are used for two Landsat images related to 1993 and 2023: the Support Vector Machine (SVM) and the Random Forest (RF). These algorithms are integrated into the SCP plugin of the QGIS software used in this study. The overall accuracy achieved by applying the SVM algorithm to the Landsat 5 TM image from 1993 is 88.24% with a Kappa value of 0.8, whereas the overall accuracy obtained for the Landsat 8 OLI image from 2023 is 99.4% with a Kappa value of 0.99. By applying the RF algorithm, the overall accuracy obtained for the 1993 Landsat 5 TM image is 86% with a Kappa value of 0.8, while for the 2023 Landsat 8 OLI image, the overall accuracy obtained is 81% with a Kappa value of 0.77. Using the transition matrix, it was possible to detect LULC changes spatiotemporally. A comparison of the classification results obtained from SVM and RF algorithms with ground truth showed that the SVM classifier was more accurate in the study area. LULC changes. SVM and RF algorithms. Accuracy metrics. Transition matrix Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 1. Introduction Analyses of land use and cover, LULC, are crucial in environmental research and regional planning. Many researchers have endorsed the importance of LULC types on the hydrological response of watersheds (Hoang 2007 ; Singh et al. 2014 ; Zhang et al. 2019 ; Seyam et al. 2023 ). This influence is particularly doubled in the case of small watershed areas. LULC affects soil erosion, flooding, sustainable development, and biodiversity (Gajbhiye et al. 2015 ). In northern Tunisia, the hydrological behavior of watersheds is undergoing major changes affecting climate, vegetation cover, surface conditions, and LULC (Chebbi et al. 2019 ; Souissi et al. 2022 ). For systematic mapping and monitoring of spatial and temporal LULC dynamics, remote sensing is the only practical way to obtain complete, accurate, quantitative, and cost-effective time-series data (Fallati et al. 2017). The use of GIS simplifies data analysis since it provides high capabilities for handling large spatial data, processing, examining spatial distributions and temporal variations, and for thematic mapping (Baccari et al. 2008 ; Slama and Sebei 2020 ; Hamza and Chmit 2022 ). The use of remote sensing combined with GIS techniques to map LULC changes has been tested in several research case studies around the world (Coppin et al. 2004 ; Bazgeera et al. 2008 ; Wu et al. 2018 ; Wiatkowska et al. 2021 ; Abebe et al. 2022). The accuracy of the satellite image classification is a determining factor in the quality of the LULC maps (Hütt et al. 2016 ). In the present study, the Semi-Automatic Classification (SCP) model integrated as a plugin into the QGIS software (Congedo 2016 ; Congedo 2021 ), was used. The plugin in question is open-source and enables semi-automatic classification of remote sensing images. It provides tools for pre-and post-processing, raster calculation, as well as free image downloads. This plugin can perform complementary processing such as the resampling of new training sites to improve the supervised classification of satellite images. The SCP plugin allows the evaluation of the accuracy of the classification which is a fundamental step in the process because it quantifies the quality of the map obtained and allows the possibility of carrying out additional processing such as merging classes of strong confusion. The generated confusion matrix shows how the LULC data and the ground truth are in agreement (Lillesand et al. 2015 ; Chuvieco 2010). Results are evaluated through the measurement of numerous classification accuracies (Fisher et al. 2018 ; Congalton and Green 2019 ; Borra et al. 2019 ). The most valued are the user’s accuracy, the producer’s accuracy, and the overall accuracy (Yuan et al. 2005 ; Jia et al. 2014 ; Flood et al. 2019 ; Tian et al. 2020 ). The Kappa coefficient is also often used to characterize and test classification accuracy or reliability (Smits et al. 1999 ; Kraemer 2014 ; Hamza and Saegh 2023 ). Different methods of supervised classification are integrated into the SCP plugin including Maximum Likelihood, Minimum Distance, Multi-Layer Perceptron, Random Forest, Spectral Angle Mapping, and Support Vector Machine. Compared to parametric approaches such as maximum probability, Machine Learning methods such as Random Forest and Support Vector Machine appear to be able to generate a more accurate classification for remote sensing satellite data (Maxwell et al. 2018 ; Singh and Pandey 2021 ). In this study, carried out on the watershed of Tessa located in the northwest of Tunisia, we focused on the comparison of the accuracy of two Machine learning (ML) algorithms: the Support Vector Machine (SVM) and the Random Forest (RF) that are widely used today for classifying satellite images for mapping the Earth's surface and consequently LULC. Machine learning-based algorithms, which are nonparametric approaches, have gained popularity in remote sensing applications over the last decade. Between 2007 and 2017, Thanh Noi and Kappas ( 2018 ) examined several articles in the ISI Web of Knowledge (indexed by SCI-EXPANDED and SSCI) that utilized machine learning methods in LULC mapping. During the research period, the utilization of SVM and RF classification algorithms increased significantly. SVM is known as a flexible statistical learning methods. Regarding the form or underlying distribution of the function f, no assumptions are made (James et al. 2021). Cortes and Vapnik proposed the SVM software package in 1993, and it was published in 1995 (Cortes and Vanpik, 1995). A hyperplane in SVM is a decision frontier that separates two classes. (Fig. 1). SVM is frequently employed in studies on remote sensing classification (Mountrakis and Ogole 2011) and it has been observed that SVM is capable of handling complex Land Use and Land Cover (LULC) classification with effectiveness (Jozdani et al. 2019 ). One of the most effective, practical, and widely used kernels in the SVM classifier family is the Radial Basis Function (RBF) kernel (Ding et al. 2021 ). In our study, the RBF kernel was selected for the SVM classifier. Breiman introduced the RF classifier in 2001 (Breiman 2001 ). The Random Forest (RF) classifier, is a machine learning system that combines numerous tree classifiers (Fig. 2). RF has shown its outstanding performance in a variety of regions and application fields (Zoungrana et al., 2024 ). To classify an input vector, each tree classifier provides one unit vote for the class that is most common in the tree (Adam et al., 2014 ). In image classification studies using a variety of data formats, RF is one of the most popular machine learning algorithms. RF enhances classification accuracy by producing numerous decision trees. A review published by Belgiu and Drăgut (2016), demonstrated that RF classifiers are fast, non-overfitting, and can handle high dimensionality and multicollinearity. In this study, the SVM and RF Machine learning algorithms have been applied to two satellite images covering the study area: a 1993 Landsat 5 TM image and a 2023 Landsat 8 OLI image. The objective is to identify the LULC in 1993 and 2023 based on satellite image classification, to analyse LULC change between these two dates, and to evaluate the accuracy of SVM and RF algorithms using a spatiotemporal and statistical analysis to determine the most effective technique for change detection in the study area. 2. Study area The watershed of Tessa lies between the governorates of El-Kef and Siliana in the northwest of Tunisia. It is located between latitudes 35° 48′ and 36° 22′ North and longitudes 8° 45′ and 9° 15′ East (Fig. 3 ). With a surface area of 2036 km², Wadi Tessa is an ephemeral river that falls within the semi-arid Mediterranean bioclimatic class. It is one of the major tributaries of the Medjerda River on its right bank (Hamed and Dhahri 2013 ). It belongs to the semi-arid, Mediterranean bioclimatic level. Rainfall averages are 480 mm per year (1985–2020). According to topographic data, the watershed's highest point is 1240 m southwest of it, while its lowest point is 280 m at its outlet to the northeast (Abidi et al. 2017 ). The selection of the study region was based on the watershed's changing hydrological regime, which has seen periods of flooding and drought. 3. Data used and methodological approach 3.1. Data used In this study, Landsat 5TM image acquired on 13-3-1993 and Landsat 8 OLI image acquired on 15-3-2023, images belonging to the joint NASA/USGS program, are used. Table 1 shows the feature characteristics of these two types of images. These images were extracted from the USGS data portal ( https://earthexplorer.usgs.gov ) to assess LULC change over the three decades between 1993 and 2023. Table 1 Feature characteristics of Landsat 5 TM and Landsat 8 OLI images Bands Wavelength (µm) Resolution (m) Landsat 5TM Band 1: Blue 0.45–0.52 30 Band 2: Green 0.52–0.60 30 Band 3: Red 0.63–0.69 30 Band 4: Near Infrared (NIR) 0.76–0.90 30 Band 5: Short-wave Infrared (SWIR) 1 1.55–1.75 30 Band 6: Thermal 10.40–12.50 120 Band 7: Short-wave Infrared (SWIR) 2 2.08–2.35 30 Landsat 8OLI Band 1: Ultra Blue (coastal/aerosol) 0.435–0.451 30 Band 2: Blue 0.452–0.512 30 Band 3: Green 0.533–0.590 30 Band 4: Red 0.636–0.673 30 Band 5: Near Infrared (NIR) 0.851–0.879 30 Band 6: Shortwave Infrared (SWIR) 1 1.566–1.651 30 Band 7: Shortwave Infrared (SWIR) 2 2.107–2.294 30 Band 8: Panchromatic 0.503–0.676 15 Band 9: Cirrus 1.363–1.384 30 3.2 Methodological approach The methodological approach used for the detection of changes is essentially based on the pre-classification, classification, and post-classification steps (Peiman 2011 ). This study focuses on the use of the semi-automatic classification SCP plugin which is a Python utility for downloading and manipulating remote sensing images. This utility was built in the QGIS environment. Many research studies have applied the SCP to land cover classification for an array of purposes (Arekhi et al. 2019 ; Furukawa et al. 2020 ). Operating semi-automatically, the SCP allows the user to download satellite images, preprocess them, and then carry out unsupervised or supervised classification automatically by selecting specified parameters in the user interface (Tempa et al. 2022). An overview of the SCP interface's functions is provided following: (i) Platform for uploading remote sensing images (ii) Pre-processing including atmospheric correction, conversion to reflectance and clipping (iii) Processing: classification and analysis; (iv) Post-processing: refinement of the classification and data interpretation; and (v) Creating training sites inputs (ROI), calculation spectral signatures and accuracy assessment. An overview of the different steps taken in the classification of images and the detection of changes is illustrated in Fig. 4 . 3.2.1 Pre-processing Pre-processing techniques are applied aiming to improve classification accuracy (Sree Sharmila et al. 2013; Okolie and Smit 2022 ). This consists of a series of operations necessary to organize the data before any classification. The preprocessing techniques help to improve the image's contrast and the edges across each zone of the image (Rajendran et al. 2020 ). These operations allow in particular the standardization of the data when the data are from different acquisition dates, but also the possibility of generating additional information useful for the classification. These operations include the conversion of images into surface reflectance and then the derivation of different spectral indices, making it possible to associate the information to reveal the discriminating properties of the surfaces (Kamusoko 2019 ). Landsat imagery is known to be distorted. Therefore, preprocessing techniques such as radiometric, atmospheric, and geometric corrections are made to provide a more direct link between data and biophysical phenomena. Radiometric, atmospheric correction, and geometric corrections of the two images were also performed using QGIS software. All image data were geometrically corrected to the WGS 84 / UTM zone 32N coordinate reference system. In addition, image enhancement, and reduction processing were performed. We had also performed preliminary image interpretation using Red, Green, and Blue (RGB) pseudocolor composites. 3.2.2 Classification The supervised classification is related to the extrapolation of the training sites or Region Of Interest (ROI) previously identified, and for which a thematic class of LULC can be granted via a classification algorithm (model). All of the image's pixels are classified by the algorithm based on a comparison of its spectral properties with reference items in the training database. Conventional methods for automatically classifying land use using data from remote sensing rely on pattern recognition algorithms (Richards, 1992). Machine learning algorithms are used in remote sensing to generate and explore Land Use classification from multi-source remote sensing data (Zhang et al. 2019 ). Based on moderate-resolution data, such as those from Landsat, several efficient techniques and classifiers have been developed to enhance the classification of land use and land cover. Classical approaches for categorizing remote sensing data include logistic regression, maximum likelihood, distance measure, and clustering. There were later more sophisticated techniques for LULC mapping such as neural networks, decision trees, Support Vector Machine, Random Forest and k-nearest neighbours (Rana et al. 2020). Random Forest (RF) and Support Vector Machine (SVM), the two machine learning (ML) methods that are applied in this work, are algorithms widely used today for classifying satellite images (Akar and Oguz 2012 ; Dabija et al. 2021 ; John and Varghese 2022 ; Avci et al. 2023 ). Before the selection of ROI training sites, Google Earth images and topographic maps related to the Tessa watershed were carefully analyzed. To improve discrimination when selecting training sites from image classes, the results of an unsupervised algorithm were used to define the training sites. In addition, RGB synthesis and visual interpretation of the reference map were considered to better select the training sites. The minimum number of training sites was 20 across all classes. 1993 Landsat 5 TM and 2020 Landsat 8 OLI images classified using SCP-QGIS supervised semi-automatic classification, made it possible to map the LULC of the Tessa watershed. Six classes have been identified and classified according to the classification adopted by the Food and Agriculture Organization of the United Nations, the FAO (Latham et al. 2014 ): Forest, Scrubland, Rangeland, Water bodies, Agricultural areas, and Bare ground (Table 2 ). Table 2 Study area’s LULC classes description (Latham et al. 2014 ) Land Use/Land Cover Class Description Forest At least 10% overlap of wooded area, a minimum crown width of 15 m, and a minimum surface area of 4 ha or more than 250 young trees per hectare Scrubland Low, loose woody formations, often found on limestone soils in semi-arid and arid regions of Tunisia (such as Romari Scrubland) Rangeland A natural herbaceous landscape (steppes, meadows) with a coverage of 10% or more. Composed of native vegetation, as opposed to species introduced by humans, sets them apart from pasture areas. Water bodies Natural or artificial surface water. Agricultural areas Crop farming (land dedicated to growing crops, cereals), Herbaceous and/or woody crops (olive). Bare ground The area is not covered with vegetation on at least 90% of its surface or it is covered with lichens and mosses. 3.2.3 Post-classification Post-classification is a very important step to ensure the accuracy of the classification of each image (Congalton and Green 2019 ). It is a method in classification-based change detection (Wu et al. 2017 ). A thorough tabulation of the changes between two classified images is created using this statistical technique. By recognizing the classes in which pixels have changed in the final state, it is possible to identify the pixels that have changed. By using this technique, we can understand the transformation of pixels and the change of class in depth, which gives us a better understanding of how LULC changes over time. This study used the SCP-QGIS plugin to run change detection statistics. QGIS visualizes LULC statistics in a set of confusion and transition matrices. The confusion matrix reflects the agreement between the current LULC raster and the ground truth using a set of accuracy measures (Olofsson et al. 2014 ; Seyam et al. 2023 ). Among these accuracy metrics, we have the Producer’s Accuracy (PA), the User’s Accuracy (UA), the Overall Accuracy (OA), the Kappa Coefficient (Ka), and the Percent Correct (PC) skill measure (Esch et al. 2020 ; Yang et al. 2021). The term "Producer's Accuracy" describes a map's accuracy from the producer's perception. This accuracy type equals 100% minus Omission Error (Franquesa et al. 2022 ). The Overall Accuracy indicates how many cases are correctly classified (Türk 1979 ). The User's Accuracy can be evaluated by comparing the number of correctly classified cases of a class to the number of cases assigned to that class in the classification (Stehman and Foody 2009 ). Stated differently, this process involves dividing the number of classified points corresponding to the reference data by the total number of points classified in this class (Zhang et al. 2016 ). To measure the degree of agreement among classes, the Kappa Coefficient is commonly used (Warrens 2010 ; Vanbelle et al. 2012 ; Yang and Zhou 2015; Ramadhani et al. 2020 ). Moreover, the Percent Correct is one of the most popular skill metrics (Wang et al. 2019 ). In the present study, User, Producer, and Overall accuracies, Kappa Coefficient, and Percent Correct skill measures were mathematically calculated as follows: \(User{\prime }s Accuracy= \frac{{n}_{ii}}{{n}_{i+}}\) (Eq. 1) \(Producer{\prime }s Accuray= \frac{{n}_{ii}}{{n}_{+i}}\) (Eq. 2) \(Overall Accuracy = \frac{1}{N}\sum _{i=1}^{r}{n}_{i}\) (Eq. 3) \(Kappa Cofficient= \frac{N\sum _{i=1}^{r}{n}_{ii}-\sum _{i=1}^{r}{n}_{i+ *}{n}_{+i}}{{N}^{2}-\sum _{i=1}^{r}{n}_{i+}* {n}_{+i}}\) (Eq. 4) \(Percent Correct= \frac{\sum _{i=1}^{r}{n}_{ii}*100}{n}\) (Eq. 5) where N: stands for the pixels in total, r for the class number, and n ii for the sum of the pixels in rows "i" and columns "i," respectively. In the error matrix, the total samples in column "i" are represented by subscription n + i , while the total samples in row "i" are represented by n i+ Evaluating the accuracy of the results is an important step in the analysis. The LULC evolution is analyzed by comparing the states of the landscape in 1993 and 2023 and then by evaluating the changes that have occurred. Cross-referencing the 1993 and 2023 LULC maps made it possible to extract the evolution statistics and the matrices reflecting the conversions of the different landscape units. 0.5 m orthophotos and high-resolution satellite images were superimposed on the dataset to analyze it visually at local scales. The map shows a comparison between the pixels of different LULC classes and the actual LULC pixels observed in the field (ground truth data). 4. Results and discussion 4.1 Application of the SVM classification 4.1.1 SVM 1993 and 2023’s LULC maps The map of LULC extracted using the SVM classification for 1993 (Fig. 5 a) shows 6 LULC classes: Agricultural areas, Rangeland, Scrubland, Bare ground, Forest, and Water bodies. It is distinguished that the Agricultural area (625 km²) and Rangeland (1159 km²) have the highest superficies (Table 3 ). Agricultural areas are mainly scattered along the hydrographic network and the watershed plains, while Rangelands are primarily located on the watershed's boundary. The smallest superficies are occupied by Bare ground (102 km²) and Scrubland (112 km²) which are mainly concentrated in the south and northwest of the study area. It should be noted that during 1993, the rains were very weak and consequently there were almost no Water bodies. The second map related to 2023 (Fig. 5 b) shows that Forests occupy 101 km², Scrubland 323 km², Bare ground 25 km², Rangelands 853 km², Agricultural areas 732 km², and Water bodies only 0.5 km ² . 4.1.2 Accuracy assessment and LULC change statistics 4.1.2.1 SVM LULC classification in Tessa watershed (1993 and 2023) Table 3 represents the area in km² of each LULC class of Tessa watershed in 1993 and 2023, as well as their percentages. For the year 1993, Rangelands and Agricultural areas respectively cover the highest areas, 57% and 31% of the total area of the watershed. The rest of the basin is occupied by Forest (2%), Scrubland (6%), and Bare ground (5%). For the year 2023, there is an increase in Agricultural areas, Scrubland, and Forest reaching 36%, 16%, and 5% respectively. On the other hand, there is a regression for Rangeland (42%) and Bare ground (1%). In addition, it should be noted that the years 1989 and 1993 were years of drought, which explains the absence of Water bodies on the LULC 1993 map (Fig. 6 ). Table 3 Areas of LULC classes (SVM classification) observed in 1993 and 2023 Classes Area in1993 (%) Area in 1993 (Km²) Area in 2023 (%) Area in 2023 (Km²) Regression trangression (Km²) Forest 2 38 5 101 63 Scrubland 6 112 16 323 211 Agricultural area 31 625 36 732 107 Rangeland 57 1159 42 853 -306 Bare ground 5 102 1 25 -77 Water bodies 0 0 0 0.5 0.5 Total 100 2036 100 2036 0 Performance metrics for classification models are often determined by confusion matrices (Luque et al. 2019 ). A confusion matrix is created to test the reliability of the classified images. The column component of the confusion matrix represents the classified data, whereas the row section contains ground truth or reference data collected during field trips. Comparing these data points with the study area's current LULC facilitates the validation of the classified results (Aryal et al. 2023 ). The confusion matrix is critical in analyzing key metrics based on reference or ground truth data, such as overall accuracy, producer accuracy, user accuracy, and the Kappa Coefficient of the classified maps. Using the SCP-QGIS plugin, both the error matrix and Kappa coefficient were generated for the classified images of 1993 and 2023. The error matrix illustrates the accuracy of categorization, with columns representing classes described as "user value" derived from the reference value, and rows corresponding to classes denoted as "producer value" from the classified image. The matrix's sidebars display the total number of correctly detected points for each class in both the classified and reference data. 4.1.2.2 SVM confusion matrix and accuracy assessment for 1993 In Table 3 , the confusion matrix (in pixels) is shown for the 1993 Landsat 5 TM image. Rangelands and Scrubland are the most confusing classes, while Bare ground is the least confusing. The confusion matrix's diagonal line displays the degree of agreement between ground truth points and LULC. The overall accuracy of the classified image was calculated to be 88.24%, which is considered satisfactory for confirming the 1993 image classification. The Kappa coefficient of 80.08% indicates that the classes are in high agreement. The producer accuracy (PA) was calculated by dividing the values along the major diagonal (chord) by the total number of sample points inside the specified class on the map. This calculation was conducted according to the ground truth data compared to the result obtained by SVM image processing. The process for calculating the producer accuracy of each class is illustrated in Table 4 . The computation of producer accuracy (PA) for various classes revealed that Bare ground had the highest producer accuracy among all classified classes (100%). This was followed by Rangeland (99.77%), Scrubland (98.98%), and Forest (78.82%). The lowest value of producer accuracy was observed for Agricultural areas (73.38%). Additionally, user accuracy (UA) was calculated for all generated classes of the classified image (Table 5 ). The highest user accuracy was recorded for Bare ground (100%), followed by Agricultural areas (99.93%), Forest (98.2%), and Rangeland (81.66%). The lowest user accuracy value was associated with Scrubland (75.43%). Table 4 SVM LULC confusion matrix related to the 1993 Landsat TM image Class Forest Scrubland Agricultural areas Water bodies Rangeland Bare ground Forest 873 16 0 0 0 0 Scrubland 27 215 37 0 6 0 Agricultural areas 0 0 9382 0 6 0 Water bodies 0 0 0 0 0 0 Rangeland 1 0 131 0 588 0 Bare ground 0 0 0 0 0 62 Table 5 Accuracy assessment related to the 1993 Landsat 5 TM image Class Accuracy assessment Forest Scrubland Agricultural areas Water bodies Rangeland Bare ground PA [%] 78.82 98.98 73.38 0 99.77 100 UA [%] 98.2 75.43 99.93 0 81.66 100 Kappa coefficient 0.98 0.74 0.99 0 0.65 1 OA [%] 88.24 Kappa 0.80 4.1.2.3 SVM confusion matrix and accuracy assessment for 2023 In Table 6 , the confusion matrix (in pixels) is presented for the 2023 Landsat 8 OLI image. The Scrublands have the most confusion compared with the other classes. There is an overall accuracy of 99.4% for this classification. The Kappa coefficient of 0.99 indicates substantial concordance between the classes. Among all the classified classes, Bare ground producer accuracy (PA) was the highest (100%). It was followed by Agricultural areas (99.97%), Rangeland (99.92%), Scrubland (99.82%), Water bodies (25.61%) and Forest (95.16%). User accuracy (UA) was also calculated for all classes of the classified image (Table 7 ). It was high enough for all LULC classes. We obtained an overall accuracy (OA) of 88.24% for the 1993 image and 99.4% for the 2023 image. In terms of spatial occupation of landscape elements, Agricultural areas, and Rangeland dominate, whereas Forest, Scrubland, Bare ground, and Water bodies are the least dominant. Table 6 Statistical metrics used to analyze the SVM classification accuracy for the 2023 Landsat 8 OLI image Classes Forest Scrubland Rangeland Water bodies Agricultural areas Bare ground Forest 425 2 2 0 0 0 Scrubland 1 64 0 1 0 0 Rangeland 1 0 718 0 0 0 Water bodies 0 0 0 559 0 0 Agricultural areas 0 0 0 0 310 0 Bare ground 0 0 0 0 2 272 Table 7 Accuracy assessment related to the 2023 Landsat 8 OLI image Accuracy assessment Forest Scrubland Agricultural areas Water bodies Rangeland Bare ground PA [%] 95.16 99.82 99.92 25.61 99.97 100 UA [%] 99.06 96.96 99.86 100.00 100.00 99.27 Overall accuracy (%) 99.4 Kappa 0.99 The changes detected by using the SVM classification were determined by highlighting the areas of the different LULC units between 1993 and 2023. The overall classification accuracies obtained were 88.24% and 99.01% and the Kappa coefficients were 0.8 and 0.99, respectively for the 1993’s and the 2023’s images. According to Lea and Curtis ( 2010 ), the ratio accuracy assessment requires an overall classification accuracy above 80% and a Kappa coefficient above 0.7 (De Souza et al. 2015 ), something that was successfully achieved in the present study. 4.2 Application of the RF Classification 4.2.1 RF 1993’s and 2023’s LULC maps The map of LULC extracted using the RF classification for the year 1993 shows the same 6 LULC classes (Fig. 7 a). The study area's smallest superficies are composed of Bare ground (59 km2) (Table 8 ). The Agricultural areas and Rangeland have the highest superficies (932 and 933 km² respectively). Forest and Scrubland occupy respectively 37 and 72 km². There were almost no Water bodies in 1993. The second map related to 2023 (Fig. 7 b) shows that Forests occupy 118 km², Scrubland 324 km², Bare ground 25 km², Rangelands 716 km², Agricultural areas 851 km², and Water bodies only 0.5 km² (Fig. 8 ). Determining the classification accuracy is essential after image classification. Using the stratified sampling technique, the QGIS SCP plugin accuracy assessment tool randomly generated 500 reference points for each of the 1993 and 2023 classified images (Fig. 8 ). A color and a pixel value are allocated to each point. The user will then have to manually select the right class when the generated points have been detected. Random positions are selected for field verification. It should be noted that field trips, as well as the use of high-resolution Google Earth Pro images, are essential for regional mapping of land use and cover. To identify the changes detected using RF classification, the areas of the various LULC units between 1993 and 2023 were highlighted. The Kappa coefficients for the 1993 and 2023 images were 0.82 and 0.77, respectively, while the overall classification accuracy was 86% and 81%. To evaluate the reliability of the classified images, a confusion matrix and an accuracy evaluation table were generated for each classified image (1993 and 2023) (Tables 9 and 10 ). Table 8 Areas of LULC classes (RF classification) observed in 1993 and 2023 Classes Area in1993 (%) Area in 1993 (Km²) Area in 2023 (%) Area in 2023 (Km²) Regression/ Progression (Km²) Forest 2 37 6 118 81 Scrubland 4 72 16 324 252 Agricultural area 46 932 42 851 -81 Rangeland 46 933 35 716 -217 Bare ground 3 59 1 25 -34 Water bodies 0 0 0 0.5 0.5 Total 100 2036 100 2036 0 4.2.2 RF LULC confusion matrix and accuracy assessment in 1993 Tables 9 and 10 present the confusion matrix (in pixels) for the Landsat 5 TM image from 1993. The overall accuracy (OA) of the classified image is 86%, which is deemed satisfactory for validating the classification. The Kappa coefficient is 0.82, indicating a high level of agreement among the classes. The computation of the producer accuracy (PA) and the user accuracy (UA) were performed. Producer accuracy (PA) computation for the different RF classes indicated that Bare ground exhibited the highest PA among all the classes (93%), followed by Forest and Scrubland (86% for each of them), and Agricultural areas (85%). The lowest PA was observed for Water bodies (80%). Furthermore, UA was calculated for all generated classes of the classified image. The highest UA was recorded for Scrubland (90%), followed by Agricultural areas (88%), Bare ground (87.5%), Forest (85.71%), Rangeland (80%), and Water bodies (80%). Table 9 RF LULC confusion matrix for the Landsat 5 TM 1993 image Forest Scrubland Rangeland Water bodies Agricultural areas Bare ground Total Forest 60 5 0 0 5 0 70 Scrubland 10 90 0 0 0 0 100 Rangeland 0 5 80 0 10 5 100 Water bodies 0 0 0 20 5 0 25 Agricultural areas 0 0 10 5 110 0 125 Bare ground 0 5 5 0 0 70 80 Total 70 105 95 25 130 75 500 Table 10 Accuracy assessment for the Landsat TM 1993 image Forest Scrubland Rangeland Water bodies Agricultural areas Bare ground PA 86 86 84 80 85 93 UA 85.71 90.00 80.00 80.00 88.00 87.50 OA (%) 86 Kappa 0.82 4.2.3 RF LULC confusion matrix and accuracy assessment in 2023 For the Landsat 8 OLI image from 2023, a confusion matrix (in pixels) is provided in Table 11 . Compared to the other classes, the Scrubland class is the most accurate. In this classification, the OA is 0.81 and the Kappa coefficient is 0.77, indicating substantial agreement between classes (Table 12 ). Among all the extracted classes, Bare ground has the highest PA at 100%, followed by Agricultural areas (91%), Forest (83%), Rangeland (74%), and Scrublands (73%). Water bodies recorded the lowest PA (56%). UA was also calculated for all classes of the RF image (Table 12 ), showing high values for all LULC classes. OA is 81%. Agricultural area and Rangeland classes are predominant, while Forest, Scrubland, Bare ground, and Water bodies are less prevalent. The high value of OA (0.81) of our mapping indicates its reliability for LULC analyses. Regarding the spatial arrangement of landscape elements, our mapping reveals that Agricultural areas and Rangelands dominate, while Forest, Scrubland, Bare ground, and Water bodies constitute a smaller proportion. Table 11 RF LULC confusion matrix for the Landsat 8 OLI 2023 image Forest Scrubland Rangeland Water bodies Agricultural areas Bare ground Total Forest 50 10 0 5 5 0 70 Scrubland 10 80 0 5 0 0 95 Rangeland 0 5 70 5 0 0 80 Water bodies 0 0 5 25 5 0 35 Agricultural areas 0 10 15 0 100 0 125 Bare ground 0 5 5 5 0 80 95 Total 60 110 95 45 110 80 500 Table 12 RF accuracy assessment for the Landsat 8 OLI 2023 Forest Scrubland Rangeland Water bodies Agricultural areas Bare ground UA 71.43 84.21 87.50 71.43 80.00 84.21 PA 0.83 0.73 0.74 0.56 0.91 1.00 OA (%) 0.81 Kappa 0.77 4.3 Transition matrix and change detection of LULC from 1993 and 2023 One of the different ways for detecting land cover changes is the transition matrix, which contributes to revealing the spatiotemporal land cover transformation in table form (Bagwan and Sopan Gavali 2021 ). The transition matrix's primary purpose is to show the historical and current state of the various classes. These statistics can offer real assistance to decision-makers in tracking LULC changes and formulating suitable strategies for national development and natural resource preservation. Many Remote Sensing and GIS studies utilize the LULC transition matrix to assess LULC change patterns quantitatively (Takada et al. 2010 ). The area that has undergone a transition from one LULC class to another between two times, t 0 and t 1 , is visible thanks to the change matrix comparison technique (Daba and You 2022 ). For each LULC class, the transition matrix shows the areas of changes in LULC, and the diagonal values indicate the areas of LULC that persist between the initial and final times, with rows indicating the transitions from the initial time and columns indicating the transitions from the final time (Viana et al. 2020). The developed transition matrix related to the SVM classification depicts the various transition routes for each LULC class between 1993 and 2023 (Table 13 ). Regarding Forest class (FR), approximately 6 km² have been converted into Scrubland (SL), 1 km² into Rangeland (RL), and 1 km² into Agricultural areas (AA). 30km² of Forest areas (FR) remained unchanged. 40 km² of Scrubland areas (SL) were transformed into Forests (FR). Overall, the Forest class (FR) has gained 63km² of surface area between 1993 and 2023. For the Scrubland class 51 km² are unchanged, around 31 km² have been converted to Forest (FR), 4 km² to Rangeland (RL), and 40 km² to Forest (FR), while 134 km² of Agricultural areas (AA) have gone to Scrubland areas (SL). The Scrubland areas (SL) increased from 112 km² in 1993 to 323 km² in 2023 with a gain of 211 km². For Agricultural areas (AA), 326 km² remains unchanged, about 134 km² has been converted to Scrubland class (SL), 133 km² to Rangeland (RL), and 1 km² to Water bodies (WB). For the Rangeland class (RL), 646 km² remain unchanged, about 7 km² have been converted into Bare ground (BG) and 1 km² into Water bodies (WB). At the same time, 374 km² of Rangeland class (RL) went to Agricultural areas (AA) and 131 to Scrubland class (SL). For Bare ground (BG) 18 km² remain unchanged and around 14 km² have been converted into Agricultural areas (AA), 69 km² into Rangeland (RL), and 1 km² have gone into Scrubland class (SL). This implies a decrease of 77 km² for Bare ground (BG) during the period 1993–2023. Regarding the Water bodies class (WB), it did not exist in 1993 since it was a dry year where the quantities of rain were very low with a rate of 230 mm (DGRE, 1994). It can be concluded that for the period 1993–2023, major changes were observed in the Agricultural areas (AA) and Rangeland (RL) classes. Agricultural areas (AA) increased from 625 km² in 1993 to reach an area of 732 km² in 2023. This transgression was mainly at the expense of the Rangeland class (RL) with an increase of around 107 km². Rangeland (RL) went from 1159 km² in 1993 to 853 km² in 2023, with a regression of 306 km² for 30 years. Table 13 Transition matrix of LULC related to SVM classification, observed between 1993 and 2023 Year 2023 Regression/ transgression (km²) (1993–2023) 1993 LULC Class Forest Scrubland Agriculture areas Rangeland Bare ground Water bodies Total 1993 (km²) Forest 30 6 1 1 0 0 38 + 63 Scrubland 40 51 17 4 0 0 112 + 211 Agricultural areas 31 134 326 133 0 1 625 + 107 Rangeland 0 131 374 646 7 1 1159 − 306 Bare ground 0 1 14 69 18 0 102 − 77 water bodies 0 0 0 0 0 0 0 + 0.05 Total 2023 (km²) 101 323 732 853 25 2 2036 Table 14 related to the RF classifier shows that concerning the Forest class (FR), around 6 km² were converted into Scrubland (SL) and 1 km² into Rangeland (RL), and around 30 km² remained unchanged. For Scrubland class (SL), 18 km² remained unchanged but around 42 km² were converted to Forest (FR), 2 km² to Rangeland (RL), and 10 km² to Agricultural areas (AA). For Agricultural areas (AA), 439 km² remained unchanged but about 165 km² were converted to Scrubland class (SL), 294 km² to Rangeland (RL), 33 km² to Forest (FR), and 1 km² to Bare ground (BG). For Rangeland (RL), 481 km² remained unchanged but about 8 km² were converted to Bare ground (BG), 14 km² to Forest (FR), 134 km² to Scrubland (SL), while 296 km² were converted to Agricultural areas (AA). For Bare ground (BG), 15 km² were unchanged. But around 6 km² have been converted into Agricultural areas (AA), and 38 km² into Rangeland (RL). The Water Bodies (WB) did not exist in 1993 because it was a dry year. For the period 1993–2023, major changes were observed (Fig. 9 ): a decrease in the areas of the three classes, namely RL (− 117 km²), AA (− 81 km²), and BG (− 34 km²). On the other hand, increases in the areas of the three other LULC classes, namely SL (+ 252 km²), FR (+ 81km²), and WB (+ 0.5 km²). We can see in Fig. 9 the RF and SVM changes between the different LULC classes from 1993 and 2023. Table 14 Transition matrix of LULC related to RF classification, observed between 1993 and 2023 Year 2023 Regression/ transgression (km²) (1993–2023 1993 LULC Class Forest Scrubland Agriculture areas Rangeland Bare ground Water bodies Total 1993 (km²) Forest 30 6 0 1 0 0 37 + 81 Scrubland 42 18 10 2 0 0 72 + 252 Agricultural areas 33 165 439 294 1 0 932 − 81 Rangeland 14 134 296 481 8 0 933 − 117 Bare ground 0 0 6 38 15 0 59 − 34 water bodies 0 0 0 0 0 2 3 + 0.5 Total 2023 (km²) 118 324 751 816 25 2 2036 In the field, the total surface of Forest (FR) and Agricultural areas (AA) has significantly increased in the study area according to the registered 1993 data and the predicted data for 2023 by the National Agricultural Observatory (ONAGRI) service of the Ministry of Agriculture, Hydraulic Resources and Fisheries in collaboration with the Food and Agriculture Organization of the United Nations (FAO) (ONAGRI 2022; Di Gregorio 2022). The surface area in km² of each LULC class in the Tessa watershed in 1993 and 2023, along with their corresponding percentages, are shown in Table 15 extracted from the ONAGRI report (ONAGRI 2022). This report explains the food security policy, and strategies for the preservation and development of agricultural resources followed by the Ministry of Agriculture, Hydraulic Resources, and Fisheries. According to this report detailed in Table 15 , Rangeland (RL) and Agricultural areas (AA) together accounted for 31% and 54% of the watershed's total area in 1993, respectively. The remaining area of the watershed is made up of Bare ground (7%), Scrubland (5%) and Forest (2%). It should be noted that for the year 1993, a very dry year, water bodies were absent. In 2023, the predicted percentage of agricultural, scrubland, and forest areas is expected to rise to 37%, 15%, and 5%, respectively. However, a regression for Bare ground (BR) (3%), and Rangelands (RL) (40%) was expected. All of these ground truth data are consistent with the results obtained by the SVM classifier and conflict with the results of the RF classifier. Table 15 LULC classes and change observed for 1993 and predicted for 2023 (ONAGRI 2022) Class Area in1993 (%) Area in 1993 (Km²) Area in 2023 (%) Area in 2023 (Km²) Regression/ transgression (Km²) Forest 2 42 5 98 56 Scrubland 5 99 15 306 207 Agricultural area 31 640 37 761 121 Rangeland 54 1105 40 810 -295 Bare ground 7 150 3 60 -90 Water bodies 0 0 0 1 1 Total 100 2036 100 2036 0 Based on the results of Table 16 the areas relating to the SVM classifier are very close to those of ground truth (GT) data provided by the Ministry of Agriculture, Hydraulic Resources and Fisheries (ONAGRI 2022). On the other hand, the areas relating to the RF classifier, are less similar. It was clear from the SVM results that the Forests FR increased by 56 km², the Scrublands SL by 207 km², and the Agricultural areas AA by 121 km². Furthermore, the Rangeland RL and the Bare ground BG classes decreased by 295 and 90 km² respectively. By subtracting the results of the ground truth from those of the SVM classifier, we found that for the same period, ranging from 1993 to 2023, there is only a very slight increase for classes AA, SL, and WB (water bodies) as well as for classes FR, SL, and BG. It is clearer from Fig. 10 that there is a higher degree of conformity between the ground truth and SVM results than between the ground truth and RF results. In Fig. 11 , we compared the change predictions made by SVM and RF classifiers from 1993 to 2023 with the ground truth GT change data. This comparison shows that the SVM conforms to the GT ground truth more closely than the RF classification. Tableau 16 Analysis of 1993–2023 change results (transgression or regression) related to SVM and RF classifiers compared to ground truth GT change SVM change RF change GT change GT change - SVM change GT change - RF change FR + 63 + 81 + 56 − 7 − 25 SL + 211 + 252 + 207 − 4 − 45 AA + 107 − 81 + 121 14 202 RL − 306 − 217 − 295 11 − 78 BG − 77 − 34 − 90 -13 − 56 WB + 0.5 + 0.5 + 1 0.5 0.5 4.4 SVM and RF Algorithms Accuracy Comparison Particularly in the area of Remote Sensing image analysis, where dimensionality is very high, the SVM is considered as a useful classifier (Chi and Bruzzone 2007 ). Furthermore, only one subset of training data is needed to make decisions. SVM is one of the most efficient techniques in memory since just this fraction of training data needs to be kept there. On the other hand, since Breiman introduced the RF classifier in 2001, it has gained popularity for use in classification, prediction, analysis, selection, and external detection. Due to its good classification results and simple, comprehensible decision-making process (Bassa et al 2016 ), as well as its ease of implementation in a parallel structure to accelerate geo-big data processing (Wright and Ziegler 2015 ), RF has become more and more popular in the land cover classification field. In our study, when the Landsat 5 TM image from 1993 is put through to the SVM method, the overall accuracy obtained is 88.24% with a Kappa value of 0.8, whereas the Landsat 8 OLI image from 2023 yields an overall accuracy of 99.4% with a Kappa value of 0.99. Using the RF algorithm, the 2023 Landsat 8 OLI image yielded an overall accuracy of 81% with a Kappa value of 0.77, while the 1993 Landsat 5 TM image yielded an overall accuracy (OA) of 86% with a Kappa value of 0.8. Change detection analysis between 1993 and 2023 proved that compared to the ground truth data, the SVM results are more accurate. Sheykhmousa et al. ( 2020 ) examined and compared RF and SVM applications in 251 peer-reviewed journal papers and found that the SVM classifier is more effective when fewer classes are used. Statistically speaking, input data with fewer than or equal to six classes is with higher accuracy. In this instance, the mean number of classes is roughly 5.5, while the mean number of classes where RF performs better is 8.4. This is consistent with the current study, in which the number of classes is 6 and the SVM method was found to be the most accurate, according to the results of the different skill measures. 5. Conclusion In this study, the main objective was to test changes in land use/land cover (LULC) in Tessa watershed (Northwestern Tunisia) between 1993 and 2023. Remote sensing, Geographic Information Systems (GIS), and statistical analysis were employed to achieve this goal. Our work focused on evaluating the accuracy of two machine learning (ML) methods: Support Vector Machine (SVM) and Random Forest (RF). These algorithms are widely employed to classify satellite images to map the Earth's surface and, therefore, LULC. Supervised classifications were carried out on Landsat images from 1993 and 2023 using the SVM and RF Machine Learning classifiers. Several accuracy metrics were employed to assess the effectiveness of the image classification technique, including Producer Accuracy, User Accuracy, Overall Accuracy, and Kappa Coefficient. On the other hand, changes in the LULC coverage were identified using transition matrices. Applying the SVM classifier on the Landsat 5 TM image from 1993 yielded an overall accuracy of 88.24% with a Kappa value of 0.8, whereas the Landsat 8 OLI image from 2023 yielded an overall accuracy of 99.4% with a Kappa value of 0.99. Using the RF classifier, the overall accuracy acquired for the Landsat 5 TM image from 1993 is 86%, with a Kappa value of 0.8, and the overall accuracy obtained for the Landsat 8 OLI image from 2023 is 81%, with a Kappa value of 0.77. The results indicated that the Tessa Watershed underwent many changes between 1993 and 2003. By comparing these classification results with field reality data, it was concluded that the SVM algorithm produced the most efficient classification result and was thus more realistic in the study area compared to the RF. Particularly in the Agricultural Areas AA class, the SVM classifier's excellent performance is clear. The SVM method assessed a transgression of 107 km² for the same class between 1993 and 2023; this is strongly closer to the ground truth data, where there was a transgression of 121 km². According to the RF classifier results, there was a regression of 81 km2 for the Agricultural Areas AA class, which is completely contrary to the ground truth. For the other LULC classes, the results of the SVM classifier follow also the ground reality with transgressions in the Forest FR, Scrubland SL, and Water bodies WB classes, and regressions in the Rangeland RL and BG Bare ground classes. SVM is widely used to conduct remote sensing classification studies, and it has been noted that it can handle complex LULC classification well. Meanwhile, the Random Forest (RF) classifier can handle high dimensionality and multicollinearity, and it is fast and without overfitting. Many studies have shown that the effectiveness of the SVM classifier is most visible when there are less than or equal to six classes, while RF performs better on average when there are nine or more classes. This is in accordance with the findings of the present study, which uses six LULC classes and where the SVM classifier results are more valid compared to the ground truth. Further studies may be carried out in other study areas in Tunisia to validate the results obtained. Declarations Funding: There is no Funding for this research. Data availability: The data used in this study is either open or public. The datasets generated or analyzed during the current study are available from the corresponding author upon request. Ethics approval : There is no approval committee. Conflict of interest: On behalf of all authors, the corresponding author declares that there are no conflicts of interest. References Abdu HA (2019) Classification accuracy and trend assessments of land cover-land use changes from principal components of land satellite images. Int J Remote Sens 40(4):1275–1300. https://doi.org/10.1080/01431161.2018.1524587 Abebe G, Getachew D, Ewunetu A (2021) Analysing land use/land cover changes and its dynamics using remote sensing and GIS in Gubalafito district, Northeastern Ethiopia. SN Applied Sciences https://doi.org/10.1007/s42452-021-04915-8 . 4 Abidi S, Hajji O, Habaieb H (2017) Study of Rainfall Variations In Tessa Subwatershed of Medjerda River in Tunisia. Water Resources in Arid Areas: The Way Forward, Springer Water Book. pp. 59–74. https://doi.org/10.1007/978-3-319-51856-5 Adam E, Mut anga O, Odindi J, Abdel-Rahman EM (2014) Land-use/cover classification in a heterogeneous coastal landscape using RapidEye imagery: evaluating the performance of random forest and support vector machines classifiers. Int J Remote Sens 35(10):3440–3458. https://doi.org/10.1080/01431161.2014.903435 Afifi HA (2011) Evaluation of change detection techniques for monitoring land-cover changes: a case study in new Burg El-Arab area. Alexandria Eng J 50(2):187–195. https://doi.org/10.1016/j.aej.2011.06.001 Akar Ö, Oguz G (2012) Classification of multispectral images using Random Forest algorithm. J Geodesy Geoinf 1(2):105–112. https://doi.org/10.9733/jgg.241212.1 Arekhi M, Goksel C, Sanli BF, Senel G (2019) Comparative evaluation of the spectral and spatial consistency of Sentinel-2 and Landsat-8 OLI data for Igneada Longos Forest. Int J Geo-Inf 8:1–13. https://doi.org/10.3390/ijgi8020056 Aryal J, Sitaula C, Frery A (2023) Land use and land cover (LULC) performance modeling using machine learning algorithms: a case study of the city of Melbourne, Australia. Sci Rep 13:13510. https://doi.org/10.1038/s41598-023-40564-0 Avci C, Budak M, Yağmur N, Balçik F (2023) Comparison between random forest and support vector machine algorithms for LULC classification. Int J Eng Geosci 8(1):1–10. https://doi.org/10.26833/ijeg.987605 Baccari N, Boussema MR, Lamachere JM, Nasri S (2008) Efficiency of contour benches, filling-in and silting-up of a hillside reservoir in a semi-arid climate in Tunisia. CR Geosci 340(1):38–48. https://doi.org/10.1016/j.crte.2007.09.020 Bagwan WA, Sopan Gavali R (2021) Dam-triggered Land Use Land Cover change detection and comparison (transition matrix method) of Urmodi River Watershed of Maharashtra, India: a Remote Sensing and GIS approach. Geology, Ecology, and Landscapes, 7, 189–197. https://doi.org/10.1080/24749508.2021.1952762 Bassa Z, Bob U, Szantoi Z, Ismail R (2016) Land cover and land use mapping of the iSimangaliso Wetland Park, South Africa: Comparison of oblique and orthogonal random Forest algorithms. J Appl Remote Sens 10(1):015017–015017. https://doi.org/10.1117/1.JRS.10.015017 Bazgeera S, Sharma PK, Maheya RK, Hundala SS, Sood A (2008) Assessment of land use changes using remote sensing and GIS and their implications on climatic variability for Balachaur watershed in Punjab, India. Desert 12, 139–147. https://doi.org/10.22059/JDESERT.2008.27113 Belgiu M, Drăguţ L (2016) Random Forest in remote sensing: A review of applications and future directions. ISPRS J photogrammetry remote Sens 114:24–31. https://doi.org/10.1016/j.isprsjprs.2016.01.011 Breiman L (2001) ST4 Method Random Forest, Mach. Learn., vol. 45, no. 1, pp. 5–32 Borra S, Thanki R, Dey N (2019) Satellite image analysis: clustering and classification. Springer, Singapore Chebbi HE, Pellissier JP, Khechimi W, Rolland JP (2019) Rapport de synthèse sur l’agriculture en Tunisie. Rapport de recherche CIHEAM-IAMM, pp.99, hal-02137636. https://hal.science/hal-02137636 Chi M, Bruzzone L (2007) Semisupervised classification of hyperspectral images by SVMs optimized in the primal. IEEE Trans Geosci Remote Sens 45(6):1870–1880. https://doi.org/10.1109/TGRS.2007.894550 Chuvieco E (2020) Fundamentals of Satellite Remote Sensing An Environmental Approach. CRC Press Taylor & Francis Group, London Coppin P, Jonckheere I, Nackaerts K, Muys B, Lambin E (2004) Review ArticleDigital change detection methods in ecosystem monitoring: a review. Int J Remote Sens 25(9):1565–1596. https://doi.org/10.1080/0143116031000101675 Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20:273–297. https://doi.org/10.1007/BF00994018 Congalton RG, Green K (2019) Assessing the Accuracy of Remotely Sensed Data: Principles and Practices, Third Edition, 3rd edn. CRC Press Taylor & Francis Group. https://doi.org/10.1201/9780429052729 Congedo L (2016) Semi-automatic classification plugin documentation. Release 4 (0.1), 29, 2016 Congedo L (2021) Semi-Automatic Classification Plugin: A Python tool for the download and processing of remote sensing images in QGIS. Journal of Open Source Software 6, n o 64 3172. https://doi.org/10.21105/joss.03172 Daba MH, You S (2022) Quantitatively Assessing the Future Land-Use/Land-Cover Changes and Their Driving Factors in the Upper Stream of the Awash River Based on the CA–Markov Model and Their Implications for Water Resources Management. Sustainability 14:1538. https://doi.org/10.3390/su14031538 Dabija A, Kluczek M, Zagajewski B, Raczko E, Kycko M, Al-Sulttani AH, Tardà A, Pineda L, Corbera J (2021) Comparison of Support Vector Machines and Random Forest s for Corine Land Cover Mapping. Remote Sens 13:777. https://doi.org/10.3390/rs13040777 De Souza CH, Mercante E, Johann JA, Lamparelli RA, Uribe-Opazo MA (2015) Mapping and discrimination of soya bean and corn crops using spectro-temporal profiles of vegetation indices. Int J Remote Sens 36:1809–1824. https://doi.org/10.1080/01431161.2015.1026956 DGRE Direction Générale des Ressources en Eau (1994) Annuaire pluviométrique de Tunisie Di Gregorio A, Essifi B, Henry M, Selmi K, Belhaj Salah S, Salis A, Aloui K, Cherif I, Saad K, Dannunzio R, Khemakhem F, Karray B (2022) Un système national de référence de l'occupation du sol pour la Tunisie. Rome, FAO et Tunis, DGF. https://doi.org/10.4060/cc0834fr Ding X, Liu J, Yang F, Cao J (2021) Random radial basis function kernel-based support vector machine. J Franklin Inst 358(18):10121–10140. https://doi.org/10.1016/j.jfranklin.2021.10.005 Esch T, Zeidler J, Palacios-Lopez D, Marconcini M, Roth A, Mönks M, Leutner B, Brzoska E, Metz-Marconcini A, Bachofer F et al (2020) Towards a Large-Scale 3D Modeling of the Built Environment-Joint Analysis of TanDEM-X, Sentinel-2 and. 12:2391. Open Street Map Data. Remote Sens. https://doi.org/10.3390/rs12152391 Fisher JR, Acosta EA, Dennedy-Frank P, Kroeger T, Boucher T (2018) Impact of satellite imagery spatial resolution on land use classification accuracy and modeled water quality. Remote Sensing in Ecology and Conservation https://doi.org/10.1002/rse2.61 . 4 Flood N, Watson F, Collett L (2019) Using a U-net convolutional neural network to map woody vegetation extent from high resolution satellite imagery across Queensland, Australia. Int J Appl Earth Obs Geoinf 82. https://doi.org/10.1016/j.jag.2019.101897 Franquesa M, Stehman SV, Chuvieco E (2022) Assessment and characterization of sources of error impacting the accuracy of global burned area products. Remote Sens Environ. https://doi.org/10.1016/j.rse.2022.113214 Furukawa F, Morimoto J, Yoshimura N, Kaneko M (2020) Comparison of conventional change detection methodologies using high-resolution imagery to find forest damage caused by typhoons. Remote Sens 12:1–17. https://doi.org/10.3390/rs12193242 Gajbhiye S, Sharma SK, Tignath S, Mishra SK (2015) Development of geomorphological erosion index for Shakker watershed. J. Geol. Soc. India. 2015 a; 86(3). http://doi.org/10.1007/s12594-015-0323-3 Hamed Y, Dhahri F (2013) Hydro-geochemical and isotopic composition of groundwater, with emphasis on sources of salinity, in the aquifer system in Northwestern Tunisia. J Afr Earth Sc 83:10–24. https://doi.org/10.1016/j.jafrearsci.2013.02.004 Hamza MH, Chmit M (2022) GIS-Based Planning and Web/3D Web GIS Applications for the Analysis and Management of MV/LV Electrical Networks (A Case Study in Tunisia). Appl Sci 12:2554. https://doi.org/10.3390/app12052554 Hamza MH, Saegh AM (2023) Flash Flood Risk Assessment Due to a Possible Dam Break in Urban Arid Environment, the New Um Al-Khair Dam Case Study, Jeddah, Saudi Arabia. Sustainability 15:1074. https://doi.org/10.3390/su15021074 Hoang KH (2007) Les changements de l'occupation du sol et ses impacts sur les eaux de surface du bassin versant. Le cas du bassin versant de la rivière Câu (Viêt-nam). Mémoire de Maître ès Science (M.Sc.) en Sciences de l'eau. Université du Québec INRS. Terre et Environnement 127 pages. Hütt C, Koppe W, Miao Y, Bareth G (2016) Best Accuracy Land Use/Land Cover (LULC) Classification to Derive Crop Types Using Multitemporal, Multisensor, and Multi-Polarization SAR Satellite Images. Remote Sens 8:684. https://doi.org/10.3390/rs8080684 Latham J, Cumani R, Rosati I, Bloise M (2014) FAO global land cover (GLC-SHARE) beta-release 1.0 database. Division LaW James G, Witten D, Hastie T, Tibshirani R (2013) An introduction to statistical learning, vol 112. Springer, New York, p 18 Jia K, Wei XQ, Gu XF, Yao YJ, Xie XH, Li B (2014) Land cover classification using Landsat 8 Operational Land Imager data in Beijing. China Geocarto Int 29(8):941–951. https://doi.org/10.1080/10106049.2014.894586 John S, Varghese AO (2022) Analysis of support vector machine and maximum likelihood classifiers in (LULC) classification using Sentinel-2 images. Proc.Indian Natl. Sci. Acad. 88, 213–227. https://doi.org/10.1007/s43538-022-00078-1 Jozdani SE, Johnson BA, Chen D (2019) Comparing deep neural networks, ensemble classifiers, and support vector machine algorithms for object-based urban land use/land cover classification. Remote Sens 11(14):1713. https://doi.org/10.3390/rs11141713 Kamusoko C (2019) Pre-processing. Remote Sensing Image Classification in R. Springer Geography. Springer, Singapore. https://doi.org/10.1007/978-981-13-8012-9_2 Kraemer HC (2014) Kappa Coefficient. Wiley StatsRef: Statistics Reference Online https://doi.org/10.1002/9781118445112.stat003 Kung SY (2014) Kernel methods and machine learning. Cambridge University Press Lea C, Curtis AC (2010) Thematic accuracy assessment procedures: National Park Service Vegetation Inventory, version 2.0. Natural Resource Report NPS/2010/NRR––2010/204, National Park Service, Fort Collins, Colorado, USA Lillesand TM, Kiefer RW, Chipman JW (2015) Remote sensing and image interpretation. 7th ed, Hoboken, USA: John Wiley&Sons, Inc.; 2015 Lu D, Mausel P, Brondízio E, Moran E (2004) Change detection techniques. Int J Remote Sens 25(12):2365–2401. https://doi.org/10.1080/0143116031000139863 Luque A, Carrasco A, Martín A, Heras AD (2019) The impact of class imbalance in classification performance metrics based on the binary confusion matrix. Pattern Recognit 91:216–231. https://doi.org/10.1016/j.patcog.2019.02.023 Maxwell AE, Warner TA, Fang F (2018) Implementation of machine learning classification in remote sensing: an applied review. Int J Remote Sens 39(9):2784–2817. https://doi.org/10.1080/01431161.2018.1433343 Mountrakis G, Im J, Ogole C (2011) Support vector machines in remote sensing: A review. ISPRS J Photogramm Remote Sens 66(3):247–259. https://doi.org/10.1016/j.isprsjprs.2010.11.001 Okolie CJ, Smit JL (2022) A systematic review and meta-analysis of Digital Elevation Model (DEM) fusion: pre-processing, methods and applications. ISPRS J Photogramm Remote Sens 188:1–29. https://doi.org/10.1016/j.isprsjprs.2022.03.016 Olofsson P, Foody GM, Herold M, Stehman SV, Woodcock CE, Wulder MA (2014) Good ractices for Estimating Area and Assessing Accuracy of Land Change. Remote Sens Environ 148:42–57. https://doi.org/10.1016/j.rse.2014.02.015 ONAGRI Observatoire National de l'Agriculture (2023) Tableau de bord de l’agriculture en tunisie. Ministère de l’Agriculture, des Ressources Hydrauliques et de la Pêche, Tunisie Peiman R (2011) Pre-classification and post-classification change-detection techniques to monitor land-cover and land-use change using multi-temporal Landsat imagery: a case study on Pisa Province in Italy. Int J Remote Sens 32:4365–4381. https://doi.org/10.1080/01431161.2010.486806 QGIS Development Team (2021) QGIS geographic information system. QGIS Association. https://www.qgis.org Rajendran GB, Kumarasamy UM, Zarro C, Divakarachari PB, Ullo SL (2020) Land-Use and Land-Cover Classification Using a Human Group-Based Particle Swarm Optimization Algorithm with an LSTM Classifier on Hybrid Pre-Processing Remote-Sensing Images. Remote Sens 12:4135. https://doi.org/10.3390/rs12244135 Ramadhani F, Pullanagari R, Kereszturi G, Procter J (2020) Mapping of rice growth phases and Bare land using Landsat-8 OLI with machine learning. Int J Remote Sens 41:8428–8452. https://doi.org/10.1080/01431161.2020.1779378 Rana VK, Venkata Suryanarayana TM (2020) Performance evaluation of MLE, RF and SVM classification algorithms for watershed scale land use/land cover mapping using sentinel 2 bands. Remote Sens Applications: Soc Environ 19:100351. https://doi.org/10.1016/j.rsase.2020.100351 Seyam MMH, Haque MR, Rahman MM (2023) Identifying the land use land cover (LULC) changes using remote sensing and GIS approach: A case study at Bhaluka in Mymensingh, Bangladesh. 7:, Article 100293. Case Studies in Chemical and Environmental Engineering https://doi.org/10.1016/j.cscee.2022.100293 Sheykhmousa M, Mahdianpari M, Ghanbari H, Mohammadimanesh F, Ghamisi P, Homayouni S (2020) Support vector machine versus random Forest for remote sensing image classification: A meta-analysis and systematic review. IEEE J Sel Top Appl Earth Observations Remote Sens 13:6308–6325. https://doi.org/10.1109/JSTARS.2020.3026724 Singh P, Gupta A, Singh M (2014) Hydrological inferences from watershed analysis for water resource management using remote sensing and GIS techniques. Egypt J Remote Sens Space Sci 17:111–121. https://doi.org/10.1016/j.ejrs.2014.09.003 Singh G, Pandey A (2021) Evaluation of classification algorithms for land use land cover mapping in the snow-fed Alaknanda River Basin of the Northwest Himalayan Region. Appl Geomat 13:863–875. https://doi.org/10.1007/s12518-021-00401-3 Slama T, Sebei A (2020) Spatial and temporal analysis of shallow groundwater quality using GIS, Grombalia aquifer, Northern Tunisia. J Afr Earth Sc 170:103915. https://doi.org/10.1016/j.jafrearsci.2020.103915 Smits PC, Dellepiane SG, Schowengerdt RA (1999) Quality Assessment of Image Classification Algorithms for Land-Cover Mapping: A Review and a Proposal for a Cost-Based Approach. Int J Remote Sens 20 no 8:1461–1486. https://doi.org/10.1080/014311699212560 Souissi D, Souie A, Sebei A, Mahfoudhi R, Zghibi A, Zouhri L, Amiri W, Ghanmi G (2022) Flood hazard mapping and assessment using fuzzy analytic hierarchy process and GIS techniques in Takelsa, Northeast Tunisia. Arab J Geosci 15:1405. https://doi.org/10.1007/s12517-022-10541-4 Sree Sharmila T, Ramar K, Sree Renga Raja T (2014) Impact of applying pre-processing techniques for improving classification accuracy. SIViP 8:149–157. https://doi.org/10.1007/s11760-013-0505-7 Stehman SV, Foody GM (2009) Accuracy assessment. The SAGE handbook of remote sensing, 297–309. https://doi.org/10.4135/9780857021052 Takada T, Miyamoto A, Hasegawa SF (2010) Derivation of a yearly transition probability matrix for land-use dynamics and its applications. Landsc Ecol 25:561–572. https://doi.org/10.1007/s10980-009-9433-x Tempa K, Aryal KR (2022) Semi-automatic classification for rapid delineation of the geohazard-prone areas using Sentinel-2 satellite imagery. SN Appl Sci 4:141. https://doi.org/10.1007/s42452-022-05028-6 Thanh Noi P, Kappas M (2018) Comparison of Random Forest, k-Nearest Neighbor, and Support Vector Machine Classifiers for Land Cover Classification Using Sentinel-2 Imagery. Sensors 18(1):18. https://doi.org/10.3390/s18010018 Tian H, Pei J, Huang J, Li X, Wang J, Zhou B, Qin Y, Wang L (2020) Garlic and Winter Wheat Identification Based on Active and Passive Satellite Imagery and the Google Earth Engine in Northern China. Remote Sens 12:3539. https://doi.org/10.3390/rs12213539 Türk G (1979) Gt index: A measure of the success of prediction. Remote Sens Environ 8:65–75. https://doi.org/10.1016/0034-4257(79)90024-5 Vanbelle S, Mutsvari T, Declerck D, Lesaffre E (2012) Hierarchical modeling of agreement, statistics in medicine. 31:3667–3680. https://doi.org/10.1002/sim.5424 Viana CM, Rocha J (2020) Evaluating Dominant Land Use/Land Cover Changes and Predicting Future Scenario in a Rural Region Using a Memoryless Stochastic Method. Sustainability 2020, 12, 4332. https://doi.org/10.3390/su12104332 Wang WX, Watkins AB, Jones DA (2019) A user-oriented forecast verification metric: Weighted Percent Correct. https://doi.org/10.1127/metz/2019/0882 . Meteorologische Zeitschrift Warrens MJ (2010) Inequalities between multi-rater kappas. Adv Data Anal Classif 4:271–286. https://doi.org/10.1007/s11634-010-0073-4 Wiatkowska B, Słodczyk J, Stokowska A (2021) Spatial-Temporal Land Use and Land Cover Changes in Urban Areas Using Remote Sensing Images and GIS Analysis: The Case Study of Opole, Poland. Geosciences 11(8):312. https://doi.org/10.3390/geosciences11080312 Wright MN, Ziegler A (2015) ranger: A fast implementation of random Forest s for high dimensional data in C + + and R. J Stat Softw 77(1):1–17. https://doi.org/10.48550/arXiv.1508.04409 Wu C, Du B, Cui X, Zhang L (2017) A post-classification change detection method based on iterative slow feature analysis and Bayesian soft fusion. Remote Sens Environ 199:241–255. https://doi.org/10.1016/j.rse.2017.07.009 Wu T, Luo J, Fang J, Ma J, Song X (2018) Unsupervised object-based change detection via a Weibull mixture model-based binarization for high-resolution remote sensing images. IEEE Geosci Remote Sens Lett 15:63–67. https://doi.org/10.1109/LGRS.2017.2773118 Yang Z, Zhou M (2014) Kappa statistic for clustered matched-pair data. Statistics in Medicine, 33. https://doi.org/10.1016/j.csda.2014.08.004 Yang J, Xu J, Zhai D-L (2023) Integrating Phenological and Geographical Information with Artificial Intelligence Algorithm to Map Rubber Plantations in Xishuangbanna. Remote Sens 13:2793. https://doi.org/10.3390/rs13142793 Yuan F, Sawaya KE, Loeffelholz BC, Bauer ME (2005) LULC classification and change analysis of the Twin Cities (Minnesota) metropolitan area by multitemporal Landsat remote sensing. Remote Sens Environ 98:317–328. http://dx.doi.org/10.1016/j.rse.2005.08.006 Zhang C, Sargent I, Pan X, Li H, Gardiner A, Hare J, Atkinson PM (2019) Joint Deep Learning for (LULC) Classification. Rem Sens Environ 221:173–187. https://doi.org/10.1016/j.rse.2018.11.014 Zhang W, Qi J, Wan P, Wang H, Xie D, Wang X, Yan G (2016) An Easy-to-Use Airborne LiDAR Data Filtering Method Based on Cloth Simulation. Remote Sens 8:501. https://doi.org/10.3390/rs8060501 Zoungrana LE, Barbouchi M, Toukabri W, Ben Khatra N, Annabi M, Bahri H (2024) Sentinel SAR-optical fusion for improving in-season wheat crop mapping at a large scale using machine learning and the Google Earth engine platform. Appl Geomat 16:147–160 https://doi.org/10.1007/s12518-023-00545-4 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 01 Jan, 2025 Read the published version in Earth Systems and Environment → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4359112","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":305890794,"identity":"943bfd20-683c-4c92-b848-09e965c55ad9","order_by":0,"name":"Noamen BACCARI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA90lEQVRIiWNgGAWjYBCDBAhlYMNPspY0yQY2IH2AgUGCSC0MhwlrMWdvfrrh4w6GPP72ww8/VxScl5Cf38Am/TGHoc4chxbLnmNmN2eeYSiWOJNmLHnG4LaEwTEGNomD2xgkLBuwazG4kWB2m7eNIbHhBg+DZIPB7ToDNqgWgwO4tKR/u/0XqGX+DR7mnw0G5yTk2whqyTG7zQjUsuEGDxvQlgMSDMcIaTlzpuxmb5tEseGZNDPLBoNkoF8Smy3ObpOQ3IBLy/H2bTd+ttnkyR0//Phmwx87CfnmwwdvVG6z4cdlCxSgRAJjA7rIKBgFo2AUjAISAQAHLlvIFiImwQAAAABJRU5ErkJggg==","orcid":"","institution":"University of Tunis El Manar","correspondingAuthor":true,"prefix":"","firstName":"Noamen","middleName":"","lastName":"BACCARI","suffix":""},{"id":305890795,"identity":"5a89c5ed-7694-4e86-9132-9bd36f0328e9","order_by":1,"name":"Mohamed Hafedh HAMZA","email":"","orcid":"","institution":"University of Tunis El Manar","correspondingAuthor":false,"prefix":"","firstName":"Mohamed","middleName":"Hafedh","lastName":"HAMZA","suffix":""},{"id":305890796,"identity":"06f7a54f-fad1-49b2-b976-0fec95c3dfa1","order_by":2,"name":"Tarek SLAMA","email":"","orcid":"","institution":"University of Tunis El Manar","correspondingAuthor":false,"prefix":"","firstName":"Tarek","middleName":"","lastName":"SLAMA","suffix":""},{"id":305890797,"identity":"7c971779-31a5-4b69-8cc1-b29d0fc8baed","order_by":3,"name":"Abdelaziz SEBEI","email":"","orcid":"","institution":"University of Tunis El Manar","correspondingAuthor":false,"prefix":"","firstName":"Abdelaziz","middleName":"","lastName":"SEBEI","suffix":""},{"id":305890798,"identity":"4f5d9e48-9439-4c6a-8c3f-6a0589746854","order_by":4,"name":"Noamen REBAI","email":"","orcid":"","institution":"University of Tunis El Manar","correspondingAuthor":false,"prefix":"","firstName":"Noamen","middleName":"","lastName":"REBAI","suffix":""}],"badges":[],"createdAt":"2024-05-02 12:21:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4359112/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4359112/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s41748-024-00562-2","type":"published","date":"2025-01-02T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":57444949,"identity":"12f8ed56-eebd-4352-8fd3-1a5e7b3c4942","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":46680,"visible":true,"origin":"","legend":"\u003cp\u003eSupport Vector Machine (SVM) classifier\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/b5366c5f49b592e1ce5f8417.jpg"},{"id":57444948,"identity":"8b807d40-1b4d-45e9-8b71-f41ccabfcd21","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":43751,"visible":true,"origin":"","legend":"\u003cp\u003eRandom Forest \u0026nbsp;(RF) classifier\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/99dc2133b8420a56f3ab9af3.jpg"},{"id":57444952,"identity":"d4cc6403-d27c-4031-8331-dfbeb5aa1e1d","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":341352,"visible":true,"origin":"","legend":"\u003cp\u003eGeographical location of Tessa watershed\u003c/p\u003e","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/59fa988afa3f0e87e5522a9d.jpg"},{"id":57444954,"identity":"6f39bdd0-ec0b-49a3-a87d-7061c848d930","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":134054,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart summarizing the different steps followed in the classification of images and the detection of changes\u003c/p\u003e","description":"","filename":"Figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/c158b38a6a1505c2f78c39e5.jpg"},{"id":57445940,"identity":"58689cc2-1e0e-4029-a88f-704f3587fbeb","added_by":"auto","created_at":"2024-05-30 19:28:08","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":125922,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea \u003c/strong\u003e\u0026nbsp;SVM Map of LULC in 1993, \u003cstrong\u003eb\u003c/strong\u003e SVM Map of LULC in 2023\u003c/p\u003e","description":"","filename":"Figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/dad859213c9eccc4061686f3.jpg"},{"id":57445484,"identity":"552643ed-b519-4b67-8e6d-1728de94b4bc","added_by":"auto","created_at":"2024-05-30 19:20:08","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":35584,"visible":true,"origin":"","legend":"\u003cp\u003eSVM LULC change histogram in 1993 and 2023\u003c/p\u003e","description":"","filename":"Figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/1b38b2d7ed271933215ee139.jpg"},{"id":57444953,"identity":"c4adc9b6-35b0-43f4-9a53-937a79ef0028","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":122832,"visible":true,"origin":"","legend":"\u003cp\u003eRandom points used for the accuracy assessment \u003cstrong\u003ea)\u003c/strong\u003e superimposed on the RF classification map of LULC in 1993, \u003cstrong\u003eb)\u003c/strong\u003e superimposed on the RF classification map of LULC in 2023\u003c/p\u003e","description":"","filename":"Figure7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/44b593dfcc71a3d66a36fafa.jpg"},{"id":57445486,"identity":"7bb88c9a-e0c0-4a17-be35-d4f2abc0b7b9","added_by":"auto","created_at":"2024-05-30 19:20:08","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":41928,"visible":true,"origin":"","legend":"\u003cp\u003eRF LULC\u003cem\u003e \u003c/em\u003eareas histogram between 1993 and 2023\u003c/p\u003e","description":"","filename":"Figure8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/c00ca0941ecbcefcbdfd7165.jpg"},{"id":57445487,"identity":"effde73e-a7b3-4a09-b0ba-64cc9dcef03f","added_by":"auto","created_at":"2024-05-30 19:20:09","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":225241,"visible":true,"origin":"","legend":"\u003cp\u003eRF and SVM changes from 1993 and 2023 between the different LULC classes\u003c/p\u003e","description":"","filename":"Figure9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/01904fa223a7b846871582dd.jpg"},{"id":57444958,"identity":"72bb92fd-ec95-4d14-a6d3-b88f6fd9d7bc","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":35383,"visible":true,"origin":"","legend":"\u003cp\u003eLULC class changes observed between 1993 and 2023 for ground truth GT, and for SVM and RF classifiers\u003c/p\u003e","description":"","filename":"Figure10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/9da58fc6c0b77eb917242cfb.jpg"},{"id":57444957,"identity":"c89d7fd1-3044-4ef5-b702-bd49851094a5","added_by":"auto","created_at":"2024-05-30 19:12:08","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":28260,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of SVM and RF classifiers' change predictions from 1993 to 2023 with ground truth GT change\u003c/p\u003e","description":"","filename":"Figure11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/281b4ed0bca86d2f2c66be9d.jpg"},{"id":73305919,"identity":"28bd7543-5446-4aac-8c6e-18eeffccd9bf","added_by":"auto","created_at":"2025-01-08 17:01:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2692601,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4359112/v1/cc3591f6-97d2-42e6-8cc8-4baf1ac1b318.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Evaluation of SVM and RF Machine Learning Algorithms in Land Use/Land Cover Change Assessment: Tessa Watershed Case Study (Northwest of Tunisia)","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAnalyses of land use and cover, LULC, are crucial in environmental research and regional planning. Many researchers have endorsed the importance of LULC types on the hydrological response of watersheds (Hoang \u003cspan class=\"CitationRef\"\u003e2007\u003c/span\u003e; Singh et al. \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zhang et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Seyam et al. \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). This influence is particularly doubled in the case of small watershed areas. LULC affects soil erosion, flooding, sustainable development, and biodiversity (Gajbhiye et al. \u003cspan class=\"CitationRef\"\u003e2015\u003c/span\u003e). In northern Tunisia, the hydrological behavior of watersheds is undergoing major changes affecting climate, vegetation cover, surface conditions, and LULC (Chebbi et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Souissi et al. \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eFor systematic mapping and monitoring of spatial and temporal LULC dynamics, remote sensing is the only practical way to obtain complete, accurate, quantitative, and cost-effective time-series data (Fallati et al. 2017). The use of GIS simplifies data analysis since it provides high capabilities for handling large spatial data, processing, examining spatial distributions and temporal variations, and for thematic mapping (Baccari et al. \u003cspan class=\"CitationRef\"\u003e2008\u003c/span\u003e; Slama and Sebei \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hamza and Chmit \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e). The use of remote sensing combined with GIS techniques to map LULC changes has been tested in several research case studies around the world (Coppin et al. \u003cspan class=\"CitationRef\"\u003e2004\u003c/span\u003e; Bazgeera et al. \u003cspan class=\"CitationRef\"\u003e2008\u003c/span\u003e; Wu et al. \u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e; Wiatkowska et al. \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e; Abebe et al. 2022). The accuracy of the satellite image classification is a determining factor in the quality of the LULC maps (H\u0026uuml;tt et al. \u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e). In the present study, the Semi-Automatic Classification (SCP) model integrated as a plugin into the QGIS software (Congedo \u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e; Congedo \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e), was used. The plugin in question is open-source and enables semi-automatic classification of remote sensing images. It provides tools for pre-and post-processing, raster calculation, as well as free image downloads. This plugin can perform complementary processing such as the resampling of new training sites to improve the supervised classification of satellite images. The SCP plugin allows the evaluation of the accuracy of the classification which is a fundamental step in the process because it quantifies the quality of the map obtained and allows the possibility of carrying out additional processing such as merging classes of strong confusion. The generated confusion matrix shows how the LULC data and the ground truth are in agreement (Lillesand et al. \u003cspan class=\"CitationRef\"\u003e2015\u003c/span\u003e; Chuvieco 2010). Results are evaluated through the measurement of numerous classification accuracies (Fisher et al. \u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e; Congalton and Green \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Borra et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e). The most valued are the user\u0026rsquo;s accuracy, the producer\u0026rsquo;s accuracy, and the overall accuracy (Yuan et al. \u003cspan class=\"CitationRef\"\u003e2005\u003c/span\u003e; Jia et al. \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e; Flood et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Tian et al. \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e). The Kappa coefficient is also often used to characterize and test classification accuracy or reliability (Smits et al. \u003cspan class=\"CitationRef\"\u003e1999\u003c/span\u003e; Kraemer \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e; Hamza and Saegh \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eDifferent methods of supervised classification are integrated into the SCP plugin including Maximum Likelihood, Minimum Distance, Multi-Layer Perceptron, Random Forest, Spectral Angle Mapping, and Support Vector Machine. Compared to parametric approaches such as maximum probability, Machine Learning methods such as Random Forest and Support Vector Machine appear to be able to generate a more accurate classification for remote sensing satellite data (Maxwell et al. \u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e; Singh and Pandey \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this study, carried out on the watershed of Tessa located in the northwest of Tunisia, we focused on the comparison of the accuracy of two Machine learning (ML) algorithms: the Support Vector Machine (SVM) and the Random Forest (RF) that are widely used today for classifying satellite images for mapping the Earth\u0026apos;s surface and consequently LULC. Machine learning-based algorithms, which are nonparametric approaches, have gained popularity in remote sensing applications over the last decade. Between 2007 and 2017, Thanh Noi and Kappas (\u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e) examined several articles in the ISI Web of Knowledge (indexed by SCI-EXPANDED and SSCI) that utilized machine learning methods in LULC mapping. During the research period, the utilization of SVM and RF classification algorithms increased significantly. SVM is known as a flexible statistical learning methods. Regarding the form or underlying distribution of the function f, no assumptions are made (James et al. 2021). Cortes and Vapnik proposed the SVM software package in 1993, and it was published in 1995 (Cortes and Vanpik, 1995). A hyperplane in SVM is a decision frontier that separates two classes. (Fig.\u0026nbsp;1). SVM is frequently employed in studies on remote sensing classification (Mountrakis and Ogole 2011) and it has been observed that SVM is capable of handling complex Land Use and Land Cover (LULC) classification with effectiveness (Jozdani et al. \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e). One of the most effective, practical, and widely used kernels in the SVM classifier family is the Radial Basis Function (RBF) kernel (Ding et al. \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e). In our study, the RBF kernel was selected for the SVM classifier. Breiman introduced the RF classifier in 2001 (Breiman \u003cspan class=\"CitationRef\"\u003e2001\u003c/span\u003e). The Random Forest (RF) classifier, is a machine learning system that combines numerous tree classifiers (Fig.\u0026nbsp;2). RF has shown its outstanding performance in a variety of regions and application fields (Zoungrana et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). To classify an input vector, each tree classifier provides one unit vote for the class that is most common in the tree (Adam et al., \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e). In image classification studies using a variety of data formats, RF is one of the most popular machine learning algorithms. RF enhances classification accuracy by producing numerous decision trees. A review published by Belgiu and Drăgut (2016), demonstrated that RF classifiers are fast, non-overfitting, and can handle high dimensionality and multicollinearity.\u003c/p\u003e\n\u003cp\u003eIn this study, the SVM and RF Machine learning algorithms have been applied to two satellite images covering the study area: a 1993 Landsat 5 TM image and a 2023 Landsat 8 OLI image. The objective is to identify the LULC in 1993 and 2023 based on satellite image classification, to analyse LULC change between these two dates, and to evaluate the accuracy of SVM and RF algorithms using a spatiotemporal and statistical analysis to determine the most effective technique for change detection in the study area.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"2. Study area","content":"\u003cp\u003eThe watershed of Tessa lies between the governorates of El-Kef and Siliana in the northwest of Tunisia. It is located between latitudes 35\u0026deg; 48\u0026prime; and 36\u0026deg; 22\u0026prime; North and longitudes 8\u0026deg; 45\u0026prime; and 9\u0026deg; 15\u0026prime; East (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e3\u003c/span\u003e). With a surface area of 2036 km\u0026sup2;, Wadi Tessa is an ephemeral river that falls within the semi-arid Mediterranean bioclimatic class. It is one of the major tributaries of the Medjerda River on its right bank (Hamed and Dhahri \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). It belongs to the semi-arid, Mediterranean bioclimatic level. Rainfall averages are 480 mm per year (1985\u0026ndash;2020). According to topographic data, the watershed's highest point is 1240 m southwest of it, while its lowest point is 280 m at its outlet to the northeast (Abidi et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The selection of the study region was based on the watershed's changing hydrological regime, which has seen periods of flooding and drought.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"3. Data used and methodological approach","content":"\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Data used\u003c/h2\u003e \u003cp\u003eIn this study, Landsat 5TM image acquired on 13-3-1993 and Landsat 8 OLI image acquired on 15-3-2023, images belonging to the joint NASA/USGS program, are used. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the feature characteristics of these two types of images. These images were extracted from the USGS data portal (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://earthexplorer.usgs.gov\u003c/span\u003e\u003cspan address=\"https://earthexplorer.usgs.gov\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to assess LULC change over the three decades between 1993 and 2023.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eFeature characteristics of Landsat 5 TM and Landsat 8 OLI images\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBands\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWavelength\u003c/p\u003e \u003cp\u003e(\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eResolution\u003c/p\u003e \u003cp\u003e(m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003eLandsat 5TM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 1: Blue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.45\u0026ndash;0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 2: Green\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.52\u0026ndash;0.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 3: Red\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.63\u0026ndash;0.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 4: Near Infrared\u003c/p\u003e \u003cp\u003e(NIR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.76\u0026ndash;0.90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 5: Short-wave\u003c/p\u003e \u003cp\u003eInfrared (SWIR) 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.55\u0026ndash;1.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 6: Thermal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10.40\u0026ndash;12.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e120\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 7: Short-wave\u003c/p\u003e \u003cp\u003eInfrared (SWIR) 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.08\u0026ndash;2.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"8\" rowspan=\"9\"\u003e \u003cp\u003eLandsat 8OLI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 1: Ultra Blue\u003c/p\u003e \u003cp\u003e(coastal/aerosol)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.435\u0026ndash;0.451\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 2: Blue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.452\u0026ndash;0.512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 3: Green\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.533\u0026ndash;0.590\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 4: Red\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.636\u0026ndash;0.673\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 5: Near Infrared\u003c/p\u003e \u003cp\u003e(NIR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.851\u0026ndash;0.879\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 6: Shortwave Infrared (SWIR) 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.566\u0026ndash;1.651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 7: Shortwave Infrared (SWIR) 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.107\u0026ndash;2.294\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 8: Panchromatic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.503\u0026ndash;0.676\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBand 9: Cirrus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.363\u0026ndash;1.384\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Methodological approach\u003c/h2\u003e \u003cp\u003eThe methodological approach used for the detection of changes is essentially based on the pre-classification, classification, and post-classification steps (Peiman \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). This study focuses on the use of the semi-automatic classification SCP plugin which is a Python utility for downloading and manipulating remote sensing images. This utility was built in the QGIS environment. Many research studies have applied the SCP to land cover classification for an array of purposes (Arekhi et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Furukawa et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Operating semi-automatically, the SCP allows the user to download satellite images, preprocess them, and then carry out unsupervised or supervised classification automatically by selecting specified parameters in the user interface (Tempa et al. 2022). An overview of the SCP interface's functions is provided following: (i) Platform for uploading remote sensing images (ii) Pre-processing including atmospheric correction, conversion to reflectance and clipping (iii) Processing: classification and analysis; (iv) Post-processing: refinement of the classification and data interpretation; and (v) Creating training sites inputs (ROI), calculation spectral signatures and accuracy assessment. An overview of the different steps taken in the classification of images and the detection of changes is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1 Pre-processing\u003c/h2\u003e \u003cp\u003ePre-processing techniques are applied aiming to improve classification accuracy (Sree Sharmila et al. 2013; Okolie and Smit \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This consists of a series of operations necessary to organize the data before any classification. The preprocessing techniques help to improve the image's contrast and the edges across each zone of the image (Rajendran et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). These operations allow in particular the standardization of the data when the data are from different acquisition dates, but also the possibility of generating additional information useful for the classification. These operations include the conversion of images into surface reflectance and then the derivation of different spectral indices, making it possible to associate the information to reveal the discriminating properties of the surfaces (Kamusoko \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Landsat imagery is known to be distorted. Therefore, preprocessing techniques such as radiometric, atmospheric, and geometric corrections are made to provide a more direct link between data and biophysical phenomena. Radiometric, atmospheric correction, and geometric corrections of the two images were also performed using QGIS software. All image data were geometrically corrected to the WGS 84 / UTM zone 32N coordinate reference system. In addition, image enhancement, and reduction processing were performed. We had also performed preliminary image interpretation using Red, Green, and Blue (RGB) pseudocolor composites.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Classification\u003c/h2\u003e \u003cp\u003eThe supervised classification is related to the extrapolation of the training sites or Region Of Interest (ROI) previously identified, and for which a thematic class of LULC can be granted via a classification algorithm (model). All of the image's pixels are classified by the algorithm based on a comparison of its spectral properties with reference items in the training database. Conventional methods for automatically classifying land use using data from remote sensing rely on pattern recognition algorithms (Richards, 1992). Machine learning algorithms are used in remote sensing to generate and explore Land Use classification from multi-source remote sensing data (Zhang et al. \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Based on moderate-resolution data, such as those from Landsat, several efficient techniques and classifiers have been developed to enhance the classification of land use and land cover. Classical approaches for categorizing remote sensing data include logistic regression, maximum likelihood, distance measure, and clustering. There were later more sophisticated techniques for LULC mapping such as neural networks, decision trees, Support Vector Machine, Random Forest and k-nearest neighbours (Rana et al. 2020).\u003c/p\u003e \u003cp\u003eRandom Forest (RF) and Support Vector Machine (SVM), the two machine learning (ML) methods that are applied in this work, are algorithms widely used today for classifying satellite images (Akar and Oguz \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Dabija et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; John and Varghese \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Avci et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Before the selection of ROI training sites, Google Earth images and topographic maps related to the Tessa watershed were carefully analyzed. To improve discrimination when selecting training sites from image classes, the results of an unsupervised algorithm were used to define the training sites. In addition, RGB synthesis and visual interpretation of the reference map were considered to better select the training sites. The minimum number of training sites was 20 across all classes. 1993 Landsat 5 TM and 2020 Landsat 8 OLI images classified using SCP-QGIS supervised semi-automatic classification, made it possible to map the LULC of the Tessa watershed. Six classes have been identified and classified according to the classification adopted by the Food and Agriculture Organization of the United Nations, the FAO (Latham et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2014\u003c/span\u003e): Forest, Scrubland, Rangeland, Water bodies, Agricultural areas, and Bare ground (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStudy area\u0026rsquo;s LULC classes description (Latham et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2014\u003c/span\u003e)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLand Use/Land Cover Class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAt least 10% overlap of wooded area, a minimum crown width of 15 m, and a minimum surface area of 4 ha or more than 250 young trees per hectare\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow, loose woody formations, often found on limestone soils in semi-arid and arid regions of Tunisia (such as Romari Scrubland)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eA natural herbaceous landscape (steppes, meadows) with a coverage of 10% or more. Composed of native vegetation, as opposed to species introduced by humans, sets them apart from pasture areas.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNatural or artificial surface water.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCrop farming (land dedicated to growing crops, cereals), Herbaceous and/or woody crops (olive).\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThe area is not covered with vegetation on at least 90% of its surface or it is covered with lichens and mosses.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e3.2.3 Post-classification\u003c/h2\u003e \u003cp\u003ePost-classification is a very important step to ensure the accuracy of the classification of each image (Congalton and Green \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). It is a method in classification-based change detection (Wu et al. \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). A thorough tabulation of the changes between two classified images is created using this statistical technique. By recognizing the classes in which pixels have changed in the final state, it is possible to identify the pixels that have changed. By using this technique, we can understand the transformation of pixels and the change of class in depth, which gives us a better understanding of how LULC changes over time. This study used the SCP-QGIS plugin to run change detection statistics. QGIS visualizes LULC statistics in a set of confusion and transition matrices. The confusion matrix reflects the agreement between the current LULC raster and the ground truth using a set of accuracy measures (Olofsson et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Seyam et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Among these accuracy metrics, we have the Producer\u0026rsquo;s Accuracy (PA), the User\u0026rsquo;s Accuracy (UA), the Overall Accuracy (OA), the Kappa Coefficient (Ka), and the Percent Correct (PC) skill measure (Esch et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Yang et al. 2021). The term \"Producer's Accuracy\" describes a map's accuracy from the producer's perception. This accuracy type equals 100% minus Omission Error (Franquesa et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The Overall Accuracy indicates how many cases are correctly classified (T\u0026uuml;rk \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e1979\u003c/span\u003e). The User's Accuracy can be evaluated by comparing the number of correctly classified cases of a class to the number of cases assigned to that class in the classification (Stehman and Foody \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Stated differently, this process involves dividing the number of classified points corresponding to the reference data by the total number of points classified in this class (Zhang et al. \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). To measure the degree of agreement among classes, the Kappa Coefficient is commonly used (Warrens \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Vanbelle et al. \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Yang and Zhou 2015; Ramadhani et al. \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Moreover, the Percent Correct is one of the most popular skill metrics (Wang et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In the present study, User, Producer, and Overall accuracies, Kappa Coefficient, and Percent Correct skill measures were mathematically calculated as follows:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(User{\\prime }s Accuracy= \\frac{{n}_{ii}}{{n}_{i+}}\\)\u003c/span\u003e \u003c/span\u003e (Eq.\u0026nbsp;1)\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(Producer{\\prime }s Accuray= \\frac{{n}_{ii}}{{n}_{+i}}\\)\u003c/span\u003e \u003c/span\u003e (Eq.\u0026nbsp;2)\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(Overall Accuracy = \\frac{1}{N}\\sum _{i=1}^{r}{n}_{i}\\)\u003c/span\u003e \u003c/span\u003e (Eq.\u0026nbsp;3)\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(Kappa Cofficient= \\frac{N\\sum _{i=1}^{r}{n}_{ii}-\\sum _{i=1}^{r}{n}_{i+ *}{n}_{+i}}{{N}^{2}-\\sum _{i=1}^{r}{n}_{i+}* {n}_{+i}}\\)\u003c/span\u003e \u003c/span\u003e (Eq.\u0026nbsp;4)\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(Percent Correct= \\frac{\\sum _{i=1}^{r}{n}_{ii}*100}{n}\\)\u003c/span\u003e \u003c/span\u003e (Eq.\u0026nbsp;5)\u003c/p\u003e \u003cp\u003ewhere\u003c/p\u003e \u003cp\u003eN: stands for the pixels in total, r for the class number, and n\u003csub\u003e\u003cem\u003eii\u003c/em\u003e\u003c/sub\u003e for the sum of the pixels in rows \"i\" and columns \"i,\" respectively. In the error matrix, the total samples in column \"i\" are represented by subscription n\u003csub\u003e\u003cem\u003e+\u0026thinsp;i\u003c/em\u003e\u003c/sub\u003e, while the total samples in row \"i\" are represented by n \u003csub\u003e\u003cem\u003ei+\u003c/em\u003e\u003c/sub\u003e\u003c/p\u003e \u003cp\u003eEvaluating the accuracy of the results is an important step in the analysis. The LULC evolution is analyzed by comparing the states of the landscape in 1993 and 2023 and then by evaluating the changes that have occurred. Cross-referencing the 1993 and 2023 LULC maps made it possible to extract the evolution statistics and the matrices reflecting the conversions of the different landscape units. 0.5 m orthophotos and high-resolution satellite images were superimposed on the dataset to analyze it visually at local scales. The map shows a comparison between the pixels of different LULC classes and the actual LULC pixels observed in the field (ground truth data).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"4. Results and discussion","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Application of the SVM classification\u003c/h2\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e4.1.1 SVM 1993 and 2023\u0026rsquo;s LULC maps\u003c/h2\u003e \u003cp\u003eThe map of LULC extracted using the SVM classification for 1993 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e5\u003c/span\u003ea) shows 6 LULC classes: Agricultural areas, Rangeland, Scrubland, Bare ground, Forest, and Water bodies. It is distinguished that the Agricultural area (625 km\u0026sup2;) and Rangeland (1159 km\u0026sup2;) have the highest superficies (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Agricultural areas are mainly scattered along the hydrographic network and the watershed plains, while Rangelands are primarily located on the watershed's boundary. The smallest superficies are occupied by Bare ground (102 km\u0026sup2;) and Scrubland (112 km\u0026sup2;) which are mainly concentrated in the south and northwest of the study area. It should be noted that during 1993, the rains were very weak and consequently there were almost no Water bodies. The second map related to 2023 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) shows that Forests occupy 101 km\u0026sup2;, Scrubland 323 km\u0026sup2;, Bare ground 25 km\u0026sup2;, Rangelands 853 km\u0026sup2;, Agricultural areas 732 km\u0026sup2;, and Water bodies only 0.5 km\u003cb\u003e\u0026sup2;\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e4.1.2 Accuracy assessment and LULC change statistics\u003c/h2\u003e \u003cdiv id=\"Sec13\" class=\"Section4\"\u003e \u003ch2\u003e4.1.2.1 SVM LULC classification in Tessa watershed (1993 and 2023)\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e represents the area in km\u0026sup2; of each LULC class of Tessa watershed in 1993 and 2023, as well as their percentages. For the year 1993, Rangelands and Agricultural areas respectively cover the highest areas, 57% and 31% of the total area of the watershed. The rest of the basin is occupied by Forest (2%), Scrubland (6%), and Bare ground (5%). For the year 2023, there is an increase in Agricultural areas, Scrubland, and Forest reaching 36%, 16%, and 5% respectively. On the other hand, there is a regression for Rangeland (42%) and Bare ground (1%). In addition, it should be noted that the years 1989 and 1993 were years of drought, which explains the absence of Water bodies on the LULC 1993 map (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAreas of LULC classes (SVM classification) observed in 1993 and 2023\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClasses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea in1993\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArea in 1993 (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea in 2023\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eArea in 2023 (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRegression trangression (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e63\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e323\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e211\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e625\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e107\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1159\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e853\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-306\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-77\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003ePerformance metrics for classification models are often determined by confusion matrices (Luque et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). A confusion matrix is created to test the reliability of the classified images. The column component of the confusion matrix represents the classified data, whereas the row section contains ground truth or reference data collected during field trips. Comparing these data points with the study area's current LULC facilitates the validation of the classified results (Aryal et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The confusion matrix is critical in analyzing key metrics based on reference or ground truth data, such as overall accuracy, producer accuracy, user accuracy, and the Kappa Coefficient of the classified maps. Using the SCP-QGIS plugin, both the error matrix and Kappa coefficient were generated for the classified images of 1993 and 2023. The error matrix illustrates the accuracy of categorization, with columns representing classes described as \"user value\" derived from the reference value, and rows corresponding to classes denoted as \"producer value\" from the classified image. The matrix's sidebars display the total number of correctly detected points for each class in both the classified and reference data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section4\"\u003e \u003ch2\u003e4.1.2.2 SVM confusion matrix and accuracy assessment for 1993\u003c/h2\u003e \u003cp\u003eIn Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the confusion matrix (in pixels) is shown for the 1993 Landsat 5 TM image. Rangelands and Scrubland are the most confusing classes, while Bare ground is the least confusing. The confusion matrix's diagonal line displays the degree of agreement between ground truth points and LULC. The overall accuracy of the classified image was calculated to be 88.24%, which is considered satisfactory for confirming the 1993 image classification. The Kappa coefficient of 80.08% indicates that the classes are in high agreement. The producer accuracy (PA) was calculated by dividing the values along the major diagonal (chord) by the total number of sample points inside the specified class on the map. This calculation was conducted according to the ground truth data compared to the result obtained by SVM image processing. The process for calculating the producer accuracy of each class is illustrated in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. The computation of producer accuracy (PA) for various classes revealed that Bare ground had the highest producer accuracy among all classified classes (100%). This was followed by Rangeland (99.77%), Scrubland (98.98%), and Forest (78.82%). The lowest value of producer accuracy was observed for Agricultural areas (73.38%). Additionally, user accuracy (UA) was calculated for all generated classes of the classified image (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The highest user accuracy was recorded for Bare ground (100%), followed by Agricultural areas (99.93%), Forest (98.2%), and Rangeland (81.66%). The lowest user accuracy value was associated with Scrubland (75.43%).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSVM LULC confusion matrix related to the 1993 Landsat TM image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClass\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e873\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9382\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e131\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e588\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAccuracy assessment related to the 1993 Landsat 5 TM image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClass\u003c/p\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003cp\u003eassessment\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePA [%]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e78.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e98.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e73.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e99.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUA [%]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e98.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e99.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e81.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKappa coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOA [%]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e88.24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e0.80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section4\"\u003e \u003ch2\u003e4.1.2.3 SVM confusion matrix and accuracy assessment for 2023\u003c/h2\u003e \u003cp\u003eIn Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, the confusion matrix (in pixels) is presented for the 2023 Landsat 8 OLI image. The Scrublands have the most confusion compared with the other classes. There is an overall accuracy of 99.4% for this classification. The Kappa coefficient of 0.99 indicates substantial concordance between the classes. Among all the classified classes, Bare ground producer accuracy (PA) was the highest (100%). It was followed by Agricultural areas (99.97%), Rangeland (99.92%), Scrubland (99.82%), Water bodies (25.61%) and Forest (95.16%). User accuracy (UA) was also calculated for all classes of the classified image (Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). It was high enough for all LULC classes. We obtained an overall accuracy (OA) of 88.24% for the 1993 image and 99.4% for the 2023 image. In terms of spatial occupation of landscape elements, Agricultural areas, and Rangeland dominate, whereas Forest, Scrubland, Bare ground, and Water bodies are the least dominant.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStatistical metrics used to analyze the SVM classification accuracy for the 2023 Landsat 8 OLI image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClasses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e425\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e718\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e559\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e310\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e272\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAccuracy assessment related to the 2023 Landsat 8 OLI image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccuracy assessment\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePA [%]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e95.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e99.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e99.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e99.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUA [%]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e99.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e96.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e99.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e100.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e99.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverall accuracy (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e99.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe changes detected by using the SVM classification were determined by highlighting the areas of the different LULC units between 1993 and 2023. The overall classification accuracies obtained were 88.24% and 99.01% and the Kappa coefficients were 0.8 and 0.99, respectively for the 1993\u0026rsquo;s and the 2023\u0026rsquo;s images. According to Lea and Curtis (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), the ratio accuracy assessment requires an overall classification accuracy above 80% and a Kappa coefficient above 0.7 (De Souza et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), something that was successfully achieved in the present study.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Application of the RF Classification\u003c/h2\u003e \u003cdiv id=\"Sec17\" class=\"Section3\"\u003e \u003ch2\u003e4.2.1 RF 1993\u0026rsquo;s and 2023\u0026rsquo;s LULC maps\u003c/h2\u003e \u003cp\u003eThe map of LULC extracted using the RF classification for the year 1993 shows the same 6 LULC classes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). The study area's smallest superficies are composed of Bare ground (59 km2) (Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The Agricultural areas and Rangeland have the highest superficies (932 and 933 km\u0026sup2; respectively). Forest and Scrubland occupy respectively 37 and 72 km\u0026sup2;. There were almost no Water bodies in 1993. The second map related to 2023 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e7\u003c/span\u003eb) shows that Forests occupy 118 km\u0026sup2;, Scrubland 324 km\u0026sup2;, Bare ground 25 km\u0026sup2;, Rangelands 716 km\u0026sup2;, Agricultural areas 851 km\u0026sup2;, and Water bodies only 0.5 km\u0026sup2; (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e8\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDetermining the classification accuracy is essential after image classification. Using the stratified sampling technique, the QGIS SCP plugin accuracy assessment tool randomly generated 500 reference points for each of the 1993 and 2023 classified images (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e8\u003c/span\u003e). A color and a pixel value are allocated to each point. The user will then have to manually select the right class when the generated points have been detected. Random positions are selected for field verification. It should be noted that field trips, as well as the use of high-resolution Google Earth Pro images, are essential for regional mapping of land use and cover. To identify the changes detected using RF classification, the areas of the various LULC units between 1993 and 2023 were highlighted. The Kappa coefficients for the 1993 and 2023 images were 0.82 and 0.77, respectively, while the overall classification accuracy was 86% and 81%. To evaluate the reliability of the classified images, a confusion matrix and an accuracy evaluation table were generated for each classified image (1993 and 2023) (Tables\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e and \u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAreas of LULC classes (RF classification) observed in 1993 and 2023\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClasses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea in1993\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArea in 1993 (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea in 2023\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eArea in 2023\u003c/p\u003e \u003cp\u003e(Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRegression/\u003c/p\u003e \u003cp\u003eProgression\u003c/p\u003e \u003cp\u003e(Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e118\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e81\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e324\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e252\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e932\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e851\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-81\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e933\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e716\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-217\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-34\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section3\"\u003e \u003ch2\u003e4.2.2 RF LULC confusion matrix and accuracy assessment in 1993\u003c/h2\u003e \u003cp\u003eTables\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e and \u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e10\u003c/span\u003e present the confusion matrix (in pixels) for the Landsat 5 TM image from 1993. The overall accuracy (OA) of the classified image is 86%, which is deemed satisfactory for validating the classification. The Kappa coefficient is 0.82, indicating a high level of agreement among the classes. The computation of the producer accuracy (PA) and the user accuracy (UA) were performed. Producer accuracy (PA) computation for the different RF classes indicated that Bare ground exhibited the highest PA among all the classes (93%), followed by Forest and Scrubland (86% for each of them), and Agricultural areas (85%). The lowest PA was observed for Water bodies (80%). Furthermore, UA was calculated for all generated classes of the classified image. The highest UA was recorded for Scrubland (90%), followed by Agricultural areas (88%), Bare ground (87.5%), Forest (85.71%), Rangeland (80%), and Water bodies (80%).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRF LULC confusion matrix for the Landsat 5 TM 1993 image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater\u003c/p\u003e \u003cp\u003ebodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare\u003c/p\u003e \u003cp\u003eground\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e130\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e500\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab10\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 10\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAccuracy assessment for the Landsat TM 1993 image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e93\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e85.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e80.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e80.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e88.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e87.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOA (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003e4.2.3 RF LULC confusion matrix and accuracy assessment in 2023\u003c/h2\u003e \u003cp\u003eFor the Landsat 8 OLI image from 2023, a confusion matrix (in pixels) is provided in Table\u0026nbsp;\u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e11\u003c/span\u003e. Compared to the other classes, the Scrubland class is the most accurate. In this classification, the OA is 0.81 and the Kappa coefficient is 0.77, indicating substantial agreement between classes (Table\u0026nbsp;\u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e12\u003c/span\u003e). Among all the extracted classes, Bare ground has the highest PA at 100%, followed by Agricultural areas (91%), Forest (83%), Rangeland (74%), and Scrublands (73%). Water bodies recorded the lowest PA (56%). UA was also calculated for all classes of the RF image (Table\u0026nbsp;\u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e12\u003c/span\u003e), showing high values for all LULC classes. OA is 81%. Agricultural area and Rangeland classes are predominant, while Forest, Scrubland, Bare ground, and Water bodies are less prevalent. The high value of OA (0.81) of our mapping indicates its reliability for LULC analyses. Regarding the spatial arrangement of landscape elements, our mapping reveals that Agricultural areas and Rangelands dominate, while Forest, Scrubland, Bare ground, and Water bodies constitute a smaller proportion.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab11\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 11\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRF LULC confusion matrix for the Landsat 8 OLI 2023 image\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e35\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e125\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e500\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab12\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 12\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRF accuracy assessment for the Landsat 8 OLI 2023\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAgricultural areas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e71.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e84.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e87.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e71.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e80.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e84.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOA (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Transition matrix and change detection of LULC from 1993 and 2023\u003c/h2\u003e \u003cp\u003eOne of the different ways for detecting land cover changes is the transition matrix, which contributes to revealing the spatiotemporal land cover transformation in table form (Bagwan and Sopan Gavali \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The transition matrix's primary purpose is to show the historical and current state of the various classes. These statistics can offer real assistance to decision-makers in tracking LULC changes and formulating suitable strategies for national development and natural resource preservation. Many Remote Sensing and GIS studies utilize the LULC transition matrix to assess LULC change patterns quantitatively (Takada et al. \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). The area that has undergone a transition from one LULC class to another between two times, t\u003csub\u003e0\u003c/sub\u003e and t\u003csub\u003e1\u003c/sub\u003e, is visible thanks to the change matrix comparison technique (Daba and You \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). For each LULC class, the transition matrix shows the areas of changes in LULC, and the diagonal values indicate the areas of LULC that persist between the initial and final times, with rows indicating the transitions from the initial time and columns indicating the transitions from the final time (Viana et al. 2020).\u003c/p\u003e \u003cp\u003eThe developed transition matrix related to the SVM classification depicts the various transition routes for each LULC class between 1993 and 2023 (Table\u0026nbsp;\u003cspan refid=\"Tab13\" class=\"InternalRef\"\u003e13\u003c/span\u003e). Regarding Forest class (FR), approximately 6 km\u0026sup2; have been converted into Scrubland (SL), 1 km\u0026sup2; into Rangeland (RL), and 1 km\u0026sup2; into Agricultural areas (AA). 30km\u0026sup2; of Forest areas (FR) remained unchanged. 40 km\u0026sup2; of Scrubland areas (SL) were transformed into Forests (FR). Overall, the Forest class (FR) has gained 63km\u0026sup2; of surface area between 1993 and 2023. For the Scrubland class 51 km\u0026sup2; are unchanged, around 31 km\u0026sup2; have been converted to Forest (FR), 4 km\u0026sup2; to Rangeland (RL), and 40 km\u0026sup2; to Forest (FR), while 134 km\u0026sup2; of Agricultural areas (AA) have gone to Scrubland areas (SL). The Scrubland areas (SL) increased from 112 km\u0026sup2; in 1993 to 323 km\u0026sup2; in 2023 with a gain of 211 km\u0026sup2;. For Agricultural areas (AA), 326 km\u0026sup2; remains unchanged, about 134 km\u0026sup2; has been converted to Scrubland class (SL), 133 km\u0026sup2; to Rangeland (RL), and 1 km\u0026sup2; to Water bodies (WB). For the Rangeland class (RL), 646 km\u0026sup2; remain unchanged, about 7 km\u0026sup2; have been converted into Bare ground (BG) and 1 km\u0026sup2; into Water bodies (WB). At the same time, 374 km\u0026sup2; of Rangeland class (RL) went to Agricultural areas (AA) and 131 to Scrubland class (SL). For Bare ground (BG) 18 km\u0026sup2; remain unchanged and around 14 km\u0026sup2; have been converted into Agricultural areas (AA), 69 km\u0026sup2; into Rangeland (RL), and 1 km\u0026sup2; have gone into Scrubland class (SL). This implies a decrease of 77 km\u0026sup2; for Bare ground (BG) during the period 1993\u0026ndash;2023. Regarding the Water bodies class (WB), it did not exist in 1993 since it was a dry year where the quantities of rain were very low with a rate of 230 mm (DGRE, 1994). It can be concluded that for the period 1993\u0026ndash;2023, major changes were observed in the Agricultural areas (AA) and Rangeland (RL) classes. Agricultural areas (AA) increased from 625 km\u0026sup2; in 1993 to reach an area of 732 km\u0026sup2; in 2023. This transgression was mainly at the expense of the Rangeland class (RL) with an increase of around 107 km\u0026sup2;. Rangeland (RL) went from 1159 km\u0026sup2; in 1993 to 853 km\u0026sup2; in 2023, with a regression of 306 km\u0026sup2; for 30 years.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab13\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 13\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTransition matrix of LULC related to SVM classification, observed between 1993 and 2023\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c9\" namest=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRegression/\u003c/p\u003e \u003cp\u003etransgression\u003c/p\u003e \u003cp\u003e(km\u0026sup2;)\u003c/p\u003e \u003cp\u003e(1993\u0026ndash;2023)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"7\" rowspan=\"8\"\u003e \u003cp\u003e1993\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLULC Class\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAgriculture\u003c/p\u003e \u003cp\u003eareas\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare\u003c/p\u003e \u003cp\u003eground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eWater\u003c/p\u003e \u003cp\u003ebodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003cp\u003e1993\u003c/p\u003e \u003cp\u003e(km\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eForest\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+\u0026thinsp;63\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eScrubland\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+\u0026thinsp;211\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eAgricultural\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eareas\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e134\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e326\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e133\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e625\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+ 107\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRangeland\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e131\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e374\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e646\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1159\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;306\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eBare ground\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;77\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003ewater bodies\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+\u0026thinsp;0.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTotal 2023\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(km\u0026sup2;)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e323\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e853\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab14\" class=\"InternalRef\"\u003e14\u003c/span\u003e related to the RF classifier shows that concerning the Forest class (FR), around 6 km\u0026sup2; were converted into Scrubland (SL) and 1 km\u0026sup2; into Rangeland (RL), and around 30 km\u0026sup2; remained unchanged. For Scrubland class (SL), 18 km\u0026sup2; remained unchanged but around 42 km\u0026sup2; were converted to Forest (FR), 2 km\u0026sup2; to Rangeland (RL), and 10 km\u0026sup2; to Agricultural areas (AA). For Agricultural areas (AA), 439 km\u0026sup2; remained unchanged but about 165 km\u0026sup2; were converted to Scrubland class (SL), 294 km\u0026sup2; to Rangeland (RL), 33 km\u0026sup2; to Forest (FR), and 1 km\u0026sup2; to Bare ground (BG). For Rangeland (RL), 481 km\u0026sup2; remained unchanged but about 8 km\u0026sup2; were converted to Bare ground (BG), 14 km\u0026sup2; to Forest (FR), 134 km\u0026sup2; to Scrubland (SL), while 296 km\u0026sup2; were converted to Agricultural areas (AA). For Bare ground (BG), 15 km\u0026sup2; were unchanged. But around 6 km\u0026sup2; have been converted into Agricultural areas (AA), and 38 km\u0026sup2; into Rangeland (RL). The Water Bodies (WB) did not exist in 1993 because it was a dry year. For the period 1993\u0026ndash;2023, major changes were observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e9\u003c/span\u003e): a decrease in the areas of the three classes, namely RL (\u0026minus;\u0026thinsp;117 km\u0026sup2;), AA (\u0026minus;\u0026thinsp;81 km\u0026sup2;), and BG (\u0026minus;\u0026thinsp;34 km\u0026sup2;). On the other hand, increases in the areas of the three other LULC classes, namely SL (+\u0026thinsp;252 km\u0026sup2;), FR (+\u0026thinsp;81km\u0026sup2;), and WB (+\u0026thinsp;0.5 km\u0026sup2;). We can see in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e9\u003c/span\u003e the RF and SVM changes between the different LULC classes from 1993 and 2023.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab14\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 14\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTransition matrix of LULC related to RF classification, observed between 1993 and 2023\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYear\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"8\" nameend=\"c9\" namest=\"c2\"\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRegression/\u003c/p\u003e \u003cp\u003etransgression\u003c/p\u003e \u003cp\u003e(km\u0026sup2;)\u003c/p\u003e \u003cp\u003e(1993\u0026ndash;2023\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003e1993\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLULC Class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAgriculture\u003c/p\u003e \u003cp\u003eareas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBare\u003c/p\u003e \u003cp\u003eground\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eWater\u003c/p\u003e \u003cp\u003ebodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003cp\u003e1993\u003c/p\u003e \u003cp\u003e(km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+\u0026thinsp;81\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+\u0026thinsp;252\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAgricultural\u003c/p\u003e \u003cp\u003eareas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e165\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e439\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e294\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e932\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;81\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e134\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e296\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e481\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e933\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;117\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e59\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;34\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ewater bodies\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003e+ 0.5\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal 2023\u003c/p\u003e \u003cp\u003e(km\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e118\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e324\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e751\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e816\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the field, the total surface of Forest (FR) and Agricultural areas (AA) has significantly increased in the study area according to the registered 1993 data and the predicted data for 2023 by the National Agricultural Observatory (ONAGRI) service of the Ministry of Agriculture, Hydraulic Resources and Fisheries in collaboration with the Food and Agriculture Organization of the United Nations (FAO) (ONAGRI 2022; Di Gregorio 2022). The surface area in km\u0026sup2; of each LULC class in the Tessa watershed in 1993 and 2023, along with their corresponding percentages, are shown in Table\u0026nbsp;\u003cspan refid=\"Tab15\" class=\"InternalRef\"\u003e15\u003c/span\u003e extracted from the ONAGRI report (ONAGRI 2022). This report explains the food security policy, and strategies for the preservation and development of agricultural resources followed by the Ministry of Agriculture, Hydraulic Resources, and Fisheries. According to this report detailed in Table\u0026nbsp;\u003cspan refid=\"Tab15\" class=\"InternalRef\"\u003e15\u003c/span\u003e, Rangeland (RL) and Agricultural areas (AA) together accounted for 31% and 54% of the watershed's total area in 1993, respectively. The remaining area of the watershed is made up of Bare ground (7%), Scrubland (5%) and Forest (2%). It should be noted that for the year 1993, a very dry year, water bodies were absent. In 2023, the predicted percentage of agricultural, scrubland, and forest areas is expected to rise to 37%, 15%, and 5%, respectively. However, a regression for Bare ground (BR) (3%), and Rangelands (RL) (40%) was expected. All of these ground truth data are consistent with the results obtained by the SVM classifier and conflict with the results of the RF classifier.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab15\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 15\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLULC classes and change observed for 1993 and predicted for 2023 (ONAGRI 2022)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClass\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eArea in1993\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eArea in 1993 (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eArea in 2023\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eArea in 2023 (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRegression/ transgression (Km\u0026sup2;)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eForest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eScrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e306\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e207\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAgricultural area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e640\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e761\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e121\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRangeland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e810\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-295\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBare ground\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-90\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eBased on the results of Table\u0026nbsp;16 the areas relating to the SVM classifier are very close to those of ground truth (GT) data provided by the Ministry of Agriculture, Hydraulic Resources and Fisheries (ONAGRI 2022). On the other hand, the areas relating to the RF classifier, are less similar. It was clear from the SVM results that the Forests FR increased by 56 km\u0026sup2;, the Scrublands SL by 207 km\u0026sup2;, and the Agricultural areas AA by 121 km\u0026sup2;. Furthermore, the Rangeland RL and the Bare ground BG classes decreased by 295 and 90 km\u0026sup2; respectively. By subtracting the results of the ground truth from those of the SVM classifier, we found that for the same period, ranging from 1993 to 2023, there is only a very slight increase for classes AA, SL, and WB (water bodies) as well as for classes FR, SL, and BG. It is clearer from Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e10\u003c/span\u003e that there is a higher degree of conformity between the ground truth and SVM results than between the ground truth and RF results. In Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e11\u003c/span\u003e, we compared the change predictions made by SVM and RF classifiers from 1993 to 2023 with the ground truth GT change data. This comparison shows that the SVM conforms to the GT ground truth more closely than the RF classification.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTableau 16\u003c/b\u003e Analysis of 1993\u0026ndash;2023 change results (transgression or regression) related to SVM and RF classifiers compared to ground truth GT change\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabb\" border=\"1\"\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSVM change\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRF change\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGT change\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eGT change - SVM change\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGT change - RF change\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+\u0026thinsp;63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u0026thinsp;81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u0026thinsp;56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+\u0026thinsp;211\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u0026thinsp;252\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u0026thinsp;207\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;45\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+ 107\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u0026thinsp;121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e202\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;306\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;217\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;295\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;78\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+ 0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u0026thinsp;0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u0026thinsp;1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4.4 SVM and RF Algorithms Accuracy Comparison\u003c/h2\u003e \u003cp\u003eParticularly in the area of Remote Sensing image analysis, where dimensionality is very high, the SVM is considered as a useful classifier (Chi and Bruzzone \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Furthermore, only one subset of training data is needed to make decisions. SVM is one of the most efficient techniques in memory since just this fraction of training data needs to be kept there. On the other hand, since Breiman introduced the RF classifier in 2001, it has gained popularity for use in classification, prediction, analysis, selection, and external detection. Due to its good classification results and simple, comprehensible decision-making process (Bassa et al \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), as well as its ease of implementation in a parallel structure to accelerate geo-big data processing (Wright and Ziegler \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), RF has become more and more popular in the land cover classification field.\u003c/p\u003e \u003cp\u003eIn our study, when the Landsat 5 TM image from 1993 is put through to the SVM method, the overall accuracy obtained is 88.24% with a Kappa value of 0.8, whereas the Landsat 8 OLI image from 2023 yields an overall accuracy of 99.4% with a Kappa value of 0.99. Using the RF algorithm, the 2023 Landsat 8 OLI image yielded an overall accuracy of 81% with a Kappa value of 0.77, while the 1993 Landsat 5 TM image yielded an overall accuracy (OA) of 86% with a Kappa value of 0.8. Change detection analysis between 1993 and 2023 proved that compared to the ground truth data, the SVM results are more accurate. Sheykhmousa et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) examined and compared RF and SVM applications in 251 peer-reviewed journal papers and found that the SVM classifier is more effective when fewer classes are used. Statistically speaking, input data with fewer than or equal to six classes is with higher accuracy. In this instance, the mean number of classes is roughly 5.5, while the mean number of classes where RF performs better is 8.4. This is consistent with the current study, in which the number of classes is 6 and the SVM method was found to be the most accurate, according to the results of the different skill measures.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn this study, the main objective was to test changes in land use/land cover (LULC) in Tessa watershed (Northwestern Tunisia) between 1993 and 2023. Remote sensing, Geographic Information Systems (GIS), and statistical analysis were employed to achieve this goal. Our work focused on evaluating the accuracy of two machine learning (ML) methods: Support Vector Machine (SVM) and Random Forest (RF). These algorithms are widely employed to classify satellite images to map the Earth's surface and, therefore, LULC. Supervised classifications were carried out on Landsat images from 1993 and 2023 using the SVM and RF Machine Learning classifiers. Several accuracy metrics were employed to assess the effectiveness of the image classification technique, including Producer Accuracy, User Accuracy, Overall Accuracy, and Kappa Coefficient. On the other hand, changes in the LULC coverage were identified using transition matrices.\u003c/p\u003e \u003cp\u003eApplying the SVM classifier on the Landsat 5 TM image from 1993 yielded an overall accuracy of 88.24% with a Kappa value of 0.8, whereas the Landsat 8 OLI image from 2023 yielded an overall accuracy of 99.4% with a Kappa value of 0.99. Using the RF classifier, the overall accuracy acquired for the Landsat 5 TM image from 1993 is 86%, with a Kappa value of 0.8, and the overall accuracy obtained for the Landsat 8 OLI image from 2023 is 81%, with a Kappa value of 0.77. The results indicated that the Tessa Watershed underwent many changes between 1993 and 2003. By comparing these classification results with field reality data, it was concluded that the SVM algorithm produced the most efficient classification result and was thus more realistic in the study area compared to the RF. Particularly in the Agricultural Areas AA class, the SVM classifier's excellent performance is clear. The SVM method assessed a transgression of 107 km\u0026sup2; for the same class between 1993 and 2023; this is strongly closer to the ground truth data, where there was a transgression of 121 km\u0026sup2;. According to the RF classifier results, there was a regression of 81 km2 for the Agricultural Areas AA class, which is completely contrary to the ground truth. For the other LULC classes, the results of the SVM classifier follow also the ground reality with transgressions in the Forest FR, Scrubland SL, and Water bodies WB classes, and regressions in the Rangeland RL and BG Bare ground classes.\u003c/p\u003e \u003cp\u003eSVM is widely used to conduct remote sensing classification studies, and it has been noted that it can handle complex LULC classification well. Meanwhile, the Random Forest (RF) classifier can handle high dimensionality and multicollinearity, and it is fast and without overfitting. Many studies have shown that the effectiveness of the SVM classifier is most visible when there are less than or equal to six classes, while RF performs better on average when there are nine or more classes. This is in accordance with the findings of the present study, which uses six LULC classes and where the SVM classifier results are more valid compared to the ground truth. Further studies may be carried out in other study areas in Tunisia to validate the results obtained.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e There is no Funding for this research.\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eData availability:\u0026nbsp;\u003c/strong\u003eThe data used in this study is either open or public. The datasets generated or analyzed during the current study are available from the corresponding author upon request.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eEthics approval\u003c/strong\u003e\u003cstrong\u003e:\u0026nbsp;\u003c/strong\u003eThere is no approval committee.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eConflict of interest:\u0026nbsp;\u003c/strong\u003eOn behalf of all authors, the corresponding author declares that there are no conflicts of interest.\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003e\u003cspan\u003eAbdu HA (2019) Classification accuracy and trend assessments of land cover-land use changes from principal components of land satellite images. Int J Remote Sens 40(4):1275\u0026ndash;1300. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2018.1524587\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAbebe G, Getachew D, Ewunetu A (2021) Analysing land use/land cover changes and its dynamics using remote sensing and GIS in Gubalafito district, Northeastern Ethiopia. SN Applied Sciences\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s42452-021-04915-8\u003c/span\u003e\u003c/span\u003e. 4\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAbidi S, Hajji O, Habaieb H (2017) Study of Rainfall Variations In Tessa Subwatershed of Medjerda River in Tunisia. Water Resources in Arid Areas: The Way Forward, Springer Water Book. pp. 59\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-51856-5\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAdam E, Mut anga O, Odindi J, Abdel-Rahman EM (2014) Land-use/cover classification in a heterogeneous coastal landscape using RapidEye imagery: evaluating the performance of random forest and support vector machines classifiers. Int J Remote Sens 35(10):3440\u0026ndash;3458. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2014.903435\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAfifi HA (2011) Evaluation of change detection techniques for monitoring land-cover changes: a case study in new Burg El-Arab area. Alexandria Eng J 50(2):187\u0026ndash;195. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.aej.2011.06.001\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAkar \u0026Ouml;, Oguz G (2012) Classification of multispectral images using Random Forest algorithm. J Geodesy Geoinf 1(2):105\u0026ndash;112. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.9733/jgg.241212.1\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eArekhi M, Goksel C, Sanli BF, Senel G (2019) Comparative evaluation of the spectral and spatial consistency of Sentinel-2 and Landsat-8 OLI data for Igneada Longos Forest. Int J Geo-Inf 8:1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijgi8020056\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAryal J, Sitaula C, Frery A (2023) Land use and land cover (LULC) performance modeling using machine learning algorithms: a case study of the city of Melbourne, Australia. Sci Rep 13:13510. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-023-40564-0\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAvci C, Budak M, Yağmur N, Bal\u0026ccedil;ik F (2023) Comparison between random forest and support vector machine algorithms for LULC classification. Int J Eng Geosci 8(1):1\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26833/ijeg.987605\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBaccari N, Boussema MR, Lamachere JM, Nasri S (2008) Efficiency of contour benches, filling-in and silting-up of a hillside reservoir in a semi-arid climate in Tunisia. CR Geosci 340(1):38\u0026ndash;48. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.crte.2007.09.020\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBagwan WA, Sopan Gavali R (2021) Dam-triggered Land Use Land Cover change detection and comparison (transition matrix method) of Urmodi River Watershed of Maharashtra, India: a Remote Sensing and GIS approach. Geology, Ecology, and Landscapes, 7, 189\u0026ndash;197. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/24749508.2021.1952762\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBassa Z, Bob U, Szantoi Z, Ismail R (2016) Land cover and land use mapping of the iSimangaliso Wetland Park, South Africa: Comparison of oblique and orthogonal random Forest algorithms. J Appl Remote Sens 10(1):015017\u0026ndash;015017. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1117/1.JRS.10.015017\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBazgeera S, Sharma PK, Maheya RK, Hundala SS, Sood A (2008) Assessment of land use changes using remote sensing and GIS and their implications on climatic variability for Balachaur watershed in Punjab, India. Desert 12, 139\u0026ndash;147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.22059/JDESERT.2008.27113\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBelgiu M, Drăguţ L (2016) Random Forest in remote sensing: A review of applications and future directions. ISPRS J photogrammetry remote Sens 114:24\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2016.01.011\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBreiman L (2001) ST4 Method Random Forest, Mach. Learn., vol. 45, no. 1, pp. 5\u0026ndash;32\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eBorra S, Thanki R, Dey N (2019) Satellite image analysis: clustering and classification. Springer, Singapore\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eChebbi HE, Pellissier JP, Khechimi W, Rolland JP (2019) Rapport de synth\u0026egrave;se sur l\u0026rsquo;agriculture en Tunisie. Rapport de recherche CIHEAM-IAMM, pp.99, hal-02137636. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://hal.science/hal-02137636\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eChi M, Bruzzone L (2007) Semisupervised classification of hyperspectral images by SVMs optimized in the primal. IEEE Trans Geosci Remote Sens 45(6):1870\u0026ndash;1880. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TGRS.2007.894550\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eChuvieco E (2020) Fundamentals of Satellite Remote Sensing An Environmental Approach. CRC Press Taylor \u0026amp; Francis Group, London\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCoppin P, Jonckheere I, Nackaerts K, Muys B, Lambin E (2004) Review ArticleDigital change detection methods in ecosystem monitoring: a review. Int J Remote Sens 25(9):1565\u0026ndash;1596. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0143116031000101675\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20:273\u0026ndash;297. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/BF00994018\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCongalton RG, Green K (2019) Assessing the Accuracy of Remotely Sensed Data: Principles and Practices, Third Edition, 3rd edn. CRC Press Taylor \u0026amp; Francis Group. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1201/9780429052729\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCongedo L (2016) Semi-automatic classification plugin documentation. Release 4 (0.1), 29, 2016\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCongedo L (2021) Semi-Automatic Classification Plugin: A Python tool for the download and processing of remote sensing images in QGIS. Journal of Open Source Software 6, n\u003csup\u003eo\u003c/sup\u003e 64 3172. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21105/joss.03172\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDaba MH, You S (2022) Quantitatively Assessing the Future Land-Use/Land-Cover Changes and Their Driving Factors in the Upper Stream of the Awash River Based on the CA\u0026ndash;Markov Model and Their Implications for Water Resources Management. Sustainability 14:1538. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su14031538\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDabija A, Kluczek M, Zagajewski B, Raczko E, Kycko M, Al-Sulttani AH, Tard\u0026agrave; A, Pineda L, Corbera J (2021) Comparison of Support Vector Machines and Random Forest s for Corine Land Cover Mapping. Remote Sens 13:777. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs13040777\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDe Souza CH, Mercante E, Johann JA, Lamparelli RA, Uribe-Opazo MA (2015) Mapping and discrimination of soya bean and corn crops using spectro-temporal profiles of vegetation indices. Int J Remote Sens 36:1809\u0026ndash;1824. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2015.1026956\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDGRE Direction G\u0026eacute;n\u0026eacute;rale des Ressources en Eau (1994) Annuaire pluviom\u0026eacute;trique de Tunisie\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDi Gregorio A, Essifi B, Henry M, Selmi K, Belhaj Salah S, Salis A, Aloui K, Cherif I, Saad K, Dannunzio R, Khemakhem F, Karray B (2022) Un syst\u0026egrave;me national de r\u0026eacute;f\u0026eacute;rence de l\u0026apos;occupation du sol pour la Tunisie. Rome, FAO et Tunis, DGF. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4060/cc0834fr\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eDing X, Liu J, Yang F, Cao J (2021) Random radial basis function kernel-based support vector machine. J Franklin Inst 358(18):10121\u0026ndash;10140. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jfranklin.2021.10.005\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eEsch T, Zeidler J, Palacios-Lopez D, Marconcini M, Roth A, M\u0026ouml;nks M, Leutner B, Brzoska E, Metz-Marconcini A, Bachofer F et al (2020) Towards a Large-Scale 3D Modeling of the Built Environment-Joint Analysis of TanDEM-X, Sentinel-2 and. 12:2391. Open Street Map Data. Remote Sens.\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs12152391\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eFisher JR, Acosta EA, Dennedy-Frank P, Kroeger T, Boucher T (2018) Impact of satellite imagery spatial resolution on land use classification accuracy and modeled water quality. Remote Sensing in Ecology and Conservation\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/rse2.61\u003c/span\u003e\u003c/span\u003e. 4\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eFlood N, Watson F, Collett L (2019) Using a U-net convolutional neural network to map woody vegetation extent from high resolution satellite imagery across Queensland, Australia. Int J Appl Earth Obs Geoinf 82. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jag.2019.101897\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eFranquesa M, Stehman SV, Chuvieco E (2022) Assessment and characterization of sources of error impacting the accuracy of global burned area products. Remote Sens Environ. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2022.113214\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eFurukawa F, Morimoto J, Yoshimura N, Kaneko M (2020) Comparison of conventional change detection methodologies using high-resolution imagery to find forest damage caused by typhoons. Remote Sens 12:1\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs12193242\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eGajbhiye S, Sharma SK, Tignath S, Mishra SK (2015) Development of geomorphological erosion index for Shakker watershed. J. Geol. Soc. India. 2015 a; 86(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://doi.org/10.1007/s12594-015-0323-3\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHamed Y, Dhahri F (2013) Hydro-geochemical and isotopic composition of groundwater, with emphasis on sources of salinity, in the aquifer system in Northwestern Tunisia. J Afr Earth Sc 83:10\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jafrearsci.2013.02.004\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHamza MH, Chmit M (2022) GIS-Based Planning and Web/3D Web GIS Applications for the Analysis and Management of MV/LV Electrical Networks (A Case Study in Tunisia). Appl Sci 12:2554. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/app12052554\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHamza MH, Saegh AM (2023) Flash Flood Risk Assessment Due to a Possible Dam Break in Urban Arid Environment, the New Um Al-Khair Dam Case Study, Jeddah, Saudi Arabia. Sustainability 15:1074. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su15021074\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHoang KH (2007) Les changements de l\u0026apos;occupation du sol et ses impacts sur les eaux de surface du bassin versant. Le cas du bassin versant de la rivi\u0026egrave;re C\u0026acirc;u (Vi\u0026ecirc;t-nam). M\u0026eacute;moire de Ma\u0026icirc;tre \u0026egrave;s Science (M.Sc.) en Sciences de l\u0026apos;eau. Universit\u0026eacute; du Qu\u0026eacute;bec INRS. Terre et Environnement 127 pages.\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eH\u0026uuml;tt C, Koppe W, Miao Y, Bareth G (2016) Best Accuracy Land Use/Land Cover (LULC) Classification to Derive Crop Types Using Multitemporal, Multisensor, and Multi-Polarization SAR Satellite Images. Remote Sens 8:684. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs8080684\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLatham J, Cumani R, Rosati I, Bloise M (2014) FAO global land cover (GLC-SHARE) beta-release 1.0 database. Division LaW\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eJames G, Witten D, Hastie T, Tibshirani R (2013) An introduction to statistical learning, vol 112. Springer, New York, p 18\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eJia K, Wei XQ, Gu XF, Yao YJ, Xie XH, Li B (2014) Land cover classification using Landsat 8 Operational Land Imager data in Beijing. China Geocarto Int 29(8):941\u0026ndash;951. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10106049.2014.894586\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eJohn S, Varghese AO (2022) Analysis of support vector machine and maximum likelihood classifiers in (LULC) classification using Sentinel-2 images. Proc.Indian Natl. Sci. Acad. 88, 213\u0026ndash;227. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s43538-022-00078-1\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eJozdani SE, Johnson BA, Chen D (2019) Comparing deep neural networks, ensemble classifiers, and support vector machine algorithms for object-based urban land use/land cover classification. Remote Sens 11(14):1713. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs11141713\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKamusoko C (2019) Pre-processing. Remote Sensing Image Classification in R. Springer Geography. Springer, Singapore. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-981-13-8012-9_2\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKraemer HC (2014) Kappa Coefficient. Wiley StatsRef: Statistics Reference Online \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/9781118445112.stat003\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKung SY (2014) Kernel methods and machine learning. Cambridge University Press\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLea C, Curtis AC (2010) Thematic accuracy assessment procedures: National Park Service Vegetation Inventory, version 2.0. Natural Resource Report NPS/2010/NRR\u0026ndash;\u0026ndash;2010/204, National Park Service, Fort Collins, Colorado, USA\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLillesand TM, Kiefer RW, Chipman JW (2015) Remote sensing and image interpretation. 7th ed, Hoboken, USA: John Wiley\u0026amp;Sons, Inc.; 2015\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLu D, Mausel P, Brond\u0026iacute;zio E, Moran E (2004) Change detection techniques. Int J Remote Sens 25(12):2365\u0026ndash;2401. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0143116031000139863\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLuque A, Carrasco A, Mart\u0026iacute;n A, Heras AD (2019) The impact of class imbalance in classification performance metrics based on the binary confusion matrix. Pattern Recognit 91:216\u0026ndash;231. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.patcog.2019.02.023\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eMaxwell AE, Warner TA, Fang F (2018) Implementation of machine learning classification in remote sensing: an applied review. Int J Remote Sens 39(9):2784\u0026ndash;2817. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2018.1433343\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eMountrakis G, Im J, Ogole C (2011) Support vector machines in remote sensing: A review. ISPRS J Photogramm Remote Sens 66(3):247\u0026ndash;259. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2010.11.001\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eOkolie CJ, Smit JL (2022) A systematic review and meta-analysis of Digital Elevation Model (DEM) fusion: pre-processing, methods and applications. ISPRS J Photogramm Remote Sens 188:1\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2022.03.016\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eOlofsson P, Foody GM, Herold M, Stehman SV, Woodcock CE, Wulder MA (2014) Good ractices for Estimating Area and Assessing Accuracy of Land Change. Remote Sens Environ 148:42\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2014.02.015\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eONAGRI Observatoire National de l\u0026apos;Agriculture (2023) Tableau de bord de l\u0026rsquo;agriculture en tunisie. Minist\u0026egrave;re de l\u0026rsquo;Agriculture, des Ressources Hydrauliques et de la P\u0026ecirc;che, Tunisie\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003ePeiman R (2011) Pre-classification and post-classification change-detection techniques to monitor land-cover and land-use change using multi-temporal Landsat imagery: a case study on Pisa Province in Italy. Int J Remote Sens 32:4365\u0026ndash;4381. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2010.486806\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eQGIS Development Team (2021) QGIS geographic information system. QGIS Association. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.qgis.org\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eRajendran GB, Kumarasamy UM, Zarro C, Divakarachari PB, Ullo SL (2020) Land-Use and Land-Cover Classification Using a Human Group-Based Particle Swarm Optimization Algorithm with an LSTM Classifier on Hybrid Pre-Processing Remote-Sensing Images. Remote Sens 12:4135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs12244135\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eRamadhani F, Pullanagari R, Kereszturi G, Procter J (2020) Mapping of rice growth phases and Bare land using Landsat-8 OLI with machine learning. Int J Remote Sens 41:8428\u0026ndash;8452. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431161.2020.1779378\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eRana VK, Venkata Suryanarayana TM (2020) Performance evaluation of MLE, RF and SVM classification algorithms for watershed scale land use/land cover mapping using sentinel 2 bands. Remote Sens Applications: Soc Environ 19:100351. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rsase.2020.100351\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSeyam MMH, Haque MR, Rahman MM (2023) Identifying the land use land cover (LULC) changes using remote sensing and GIS approach: A case study at Bhaluka in Mymensingh, Bangladesh. 7:, Article 100293. Case Studies in Chemical and Environmental Engineering\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cscee.2022.100293\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSheykhmousa M, Mahdianpari M, Ghanbari H, Mohammadimanesh F, Ghamisi P, Homayouni S (2020) Support vector machine versus random Forest for remote sensing image classification: A meta-analysis and systematic review. IEEE J Sel Top Appl Earth Observations Remote Sens 13:6308\u0026ndash;6325. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/JSTARS.2020.3026724\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSingh P, Gupta A, Singh M (2014) Hydrological inferences from watershed analysis for water resource management using remote sensing and GIS techniques. Egypt J Remote Sens Space Sci 17:111\u0026ndash;121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejrs.2014.09.003\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSingh G, Pandey A (2021) Evaluation of classification algorithms for land use land cover mapping in the snow-fed Alaknanda River Basin of the Northwest Himalayan Region. Appl Geomat 13:863\u0026ndash;875. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12518-021-00401-3\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSlama T, Sebei A (2020) Spatial and temporal analysis of shallow groundwater quality using GIS, Grombalia aquifer, Northern Tunisia. J Afr Earth Sc 170:103915. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jafrearsci.2020.103915\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSmits PC, Dellepiane SG, Schowengerdt RA (1999) Quality Assessment of Image Classification Algorithms for Land-Cover Mapping: A Review and a Proposal for a Cost-Based Approach. Int J Remote Sens 20 no 8:1461\u0026ndash;1486. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/014311699212560\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSouissi D, Souie A, Sebei A, Mahfoudhi R, Zghibi A, Zouhri L, Amiri W, Ghanmi G (2022) Flood hazard mapping and assessment using fuzzy analytic hierarchy process and GIS techniques in Takelsa, Northeast Tunisia. Arab J Geosci 15:1405. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12517-022-10541-4\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSree Sharmila T, Ramar K, Sree Renga Raja T (2014) Impact of applying pre-processing techniques for improving classification accuracy. SIViP 8:149\u0026ndash;157. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11760-013-0505-7\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eStehman SV, Foody GM (2009) Accuracy assessment. The SAGE handbook of remote sensing, 297\u0026ndash;309. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4135/9780857021052\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eTakada T, Miyamoto A, Hasegawa SF (2010) Derivation of a yearly transition probability matrix for land-use dynamics and its applications. Landsc Ecol 25:561\u0026ndash;572. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10980-009-9433-x\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eTempa K, Aryal KR (2022) Semi-automatic classification for rapid delineation of the geohazard-prone areas using Sentinel-2 satellite imagery. SN Appl Sci 4:141. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s42452-022-05028-6\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eThanh Noi P, Kappas M (2018) Comparison of Random Forest, k-Nearest Neighbor, and Support Vector Machine Classifiers for Land Cover Classification Using Sentinel-2 Imagery. Sensors 18(1):18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s18010018\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eTian H, Pei J, Huang J, Li X, Wang J, Zhou B, Qin Y, Wang L (2020) Garlic and Winter Wheat Identification Based on Active and Passive Satellite Imagery and the Google Earth Engine in Northern China. Remote Sens 12:3539. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs12213539\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eT\u0026uuml;rk G (1979) Gt index: A measure of the success of prediction. Remote Sens Environ 8:65\u0026ndash;75. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0034-4257(79)90024-5\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eVanbelle S, Mutsvari T, Declerck D, Lesaffre E (2012) Hierarchical modeling of agreement, statistics in medicine. 31:3667\u0026ndash;3680. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/sim.5424\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eViana CM, Rocha J (2020) Evaluating Dominant Land Use/Land Cover Changes and Predicting Future Scenario in a Rural Region Using a Memoryless Stochastic Method. Sustainability 2020, 12, 4332. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su12104332\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWang WX, Watkins AB, Jones DA (2019) A user-oriented forecast verification metric: Weighted Percent Correct. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1127/metz/2019/0882\u003c/span\u003e\u003c/span\u003e. Meteorologische Zeitschrift\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWarrens MJ (2010) Inequalities between multi-rater kappas. Adv Data Anal Classif 4:271\u0026ndash;286. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11634-010-0073-4\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWiatkowska B, Słodczyk J, Stokowska A (2021) Spatial-Temporal Land Use and Land Cover Changes in Urban Areas Using Remote Sensing Images and GIS Analysis: The Case Study of Opole, Poland. Geosciences 11(8):312. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/geosciences11080312\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWright MN, Ziegler A (2015) ranger: A fast implementation of random Forest s for high dimensional data in C\u0026thinsp;+\u0026thinsp;+\u0026thinsp;and R. J Stat Softw 77(1):1\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.1508.04409\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWu C, Du B, Cui X, Zhang L (2017) A post-classification change detection method based on iterative slow feature analysis and Bayesian soft fusion. Remote Sens Environ 199:241\u0026ndash;255. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2017.07.009\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWu T, Luo J, Fang J, Ma J, Song X (2018) Unsupervised object-based change detection via a Weibull mixture model-based binarization for high-resolution remote sensing images. IEEE Geosci Remote Sens Lett 15:63\u0026ndash;67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/LGRS.2017.2773118\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eYang Z, Zhou M (2014) Kappa statistic for clustered matched-pair data. Statistics in Medicine, 33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.csda.2014.08.004\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eYang J, Xu J, Zhai D-L (2023) Integrating Phenological and Geographical Information with Artificial Intelligence Algorithm to Map Rubber Plantations in Xishuangbanna. Remote Sens 13:2793. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs13142793\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eYuan F, Sawaya KE, Loeffelholz BC, Bauer ME (2005) LULC classification and change analysis of the Twin Cities (Minnesota) metropolitan area by multitemporal Landsat remote sensing. Remote Sens Environ 98:317\u0026ndash;328. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/j.rse.2005.08.006\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eZhang C, Sargent I, Pan X, Li H, Gardiner A, Hare J, Atkinson PM (2019) Joint Deep Learning for (LULC) Classification. Rem Sens Environ 221:173\u0026ndash;187. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2018.11.014\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eZhang W, Qi J, Wan P, Wang H, Xie D, Wang X, Yan G (2016) An Easy-to-Use Airborne LiDAR Data Filtering Method Based on Cloth Simulation. Remote Sens 8:501. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs8060501\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eZoungrana LE, Barbouchi M, Toukabri W, Ben Khatra N, Annabi M, Bahri H (2024) Sentinel SAR-optical fusion for improving in-season wheat crop mapping at a large scale using machine learning and the Google Earth engine platform. Appl Geomat 16:147\u0026ndash;160 \u0026nbsp; \u0026nbsp;\u003cspan\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12518-023-00545-4\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"LULC changes. SVM and RF algorithms. Accuracy metrics. Transition matrix","lastPublishedDoi":"10.21203/rs.3.rs-4359112/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4359112/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study aims to highlight the changes in LULC (land use and Land cover) in the Tessa watershed (Northwest of Tunisia) between 1993 and 2023. Remote sensing coupled with geographic information systems (GIS) and statistical analysis, are used. Accuracy metrics make it possible to evaluate the performance of the image classification method, using the calculation of the producer\u0026rsquo;s accuracy, the user\u0026rsquo;s accuracy, overall accuracy, and the Kappa coefficient. Two Machine Learning (ML) algorithms related to the supervised classification are used for two Landsat images related to 1993 and 2023: the Support Vector Machine (SVM) and the Random Forest (RF). These algorithms are integrated into the SCP plugin of the QGIS software used in this study. The overall accuracy achieved by applying the SVM algorithm to the Landsat 5 TM image from 1993 is 88.24% with a Kappa value of 0.8, whereas the overall accuracy obtained for the Landsat 8 OLI image from 2023 is 99.4% with a Kappa value of 0.99. By applying the RF algorithm, the overall accuracy obtained for the 1993 Landsat 5 TM image is 86% with a Kappa value of 0.8, while for the 2023 Landsat 8 OLI image, the overall accuracy obtained is 81% with a Kappa value of 0.77. Using the transition matrix, it was possible to detect LULC changes spatiotemporally. A comparison of the classification results obtained from SVM and RF algorithms with ground truth showed that the SVM classifier was more accurate in the study area.\u003c/p\u003e","manuscriptTitle":"Evaluation of SVM and RF Machine Learning Algorithms in Land Use/Land Cover Change Assessment: Tessa Watershed Case Study (Northwest of Tunisia)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-30 19:12:03","doi":"10.21203/rs.3.rs-4359112/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2db3cf9a-6d41-4b4a-97c0-9d0536a6f098","owner":[],"postedDate":"May 30th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-01-08T17:01:10+00:00","versionOfRecord":{"articleIdentity":"rs-4359112","link":"https://doi.org/10.1007/s41748-024-00562-2","journal":{"identity":"earth-systems-and-environment","isVorOnly":false,"title":"Earth Systems and Environment"},"publishedOn":"2025-01-02 00:00:00","publishedOnDateReadable":"January 2nd, 2025"},"versionCreatedAt":"2024-05-30 19:12:03","video":"","vorDoi":"10.1007/s41748-024-00562-2","vorDoiUrl":"https://doi.org/10.1007/s41748-024-00562-2","workflowStages":[]},"version":"v1","identity":"rs-4359112","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4359112","identity":"rs-4359112","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}