Investigation of the Effect of Spectral Bands and Vegetation Index Selection on Agricultural Crop Type Classification (Especially for Double Crops)

Investigation of the Effect of Spectral Bands and Vegetation Index Selection on Agricultural Crop Type Classification (Especially for Double Crops) | 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 Investigation of the Effect of Spectral Bands and Vegetation Index Selection on Agricultural Crop Type Classification (Especially for Double Crops) Fatih Fehmi Şimşek This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3910868/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Satellite imagery and remote sensing technology allow the identification, observation and assessment of dynamic agricultural areas. Image classification is one of the most widely used methods to determine the pattern of agricultural crops. The accuracy of the agricultural crops to be classified depends on many parameters such as the classification method used, satellite image resolution, number of images used, bands, indices and training data. In this study, a classification study was carried out using multi-temporal Sentinel-2 imagery and datasets generated from different vegetation and spectral indices, and the effects on the classification result were investigated. As the study area has very fertile soils, suitable climate and temperature conditions and irrigated land, it is possible to grow more than one crop on the same plot during a production season. Wheat_maize (winter_wheat + summer_maize), wheat_cotton (winter_wheat + summer_cotton), lentil_cotton (winter_lentil + summer_cotton), lentil_maize (winter_lentil + summer_maize) are the crops included in the classification study, except for single crops; maize, cotton, wheat and lentils are also included. Time series of vegetation indices can be used to capture information on plant phenology and can be used as reference information in crop classification. Time series curves of different vegetation indices were constructed and compared for all crops, especially for double crops with the same phenological periods. In addition to the vegetation indices, the variation of the time series reflectance values of each spectral band was also observed for all crops and the effect of different indices and bands on the classification result was investigated. The study generated 16 different data sets using conventional vegetation indices, NDVI, SAVI, EVI and NDRE vegetation indices and all other bands of the Sentinel-2 satellite except the 60m bands. While single crops with different time series (maize, cotton, lentil, wheat) had an accuracy of over 90% in each dataset, double crops could not exceed 81% accuracy by mixing with each other in the DS-5 (R-G-B-NIR) dataset. In the DS-1 (NDVI time series) dataset, the overall accuracy for double crops is in the range of 84–85%. Classification with DS-2 (NDRE time series) increased the overall accuracy for double crops to 90%. When comparing the time series reflectance values of each spectral band for all crop types, except the crop indices, it was observed that the B6 (Red Edge-2) and B11 (SWIR-1) bands were separated from the other bands and increased the classification result by 2% when included in the dataset. Especially in the classification studies carried out on products with close phenological periods, the Red Edge band (especially Red Edge-2) and the indices (NDRE) generated from these bands will improve the classification result by preventing confusion between classes, and the B11 (SWIR-1) band also has a positive effect on classification. This study has fully demonstrated the application potential of red edge bands and the indices constructed from them. It also promotes the use of red edge band optical satellite data in agricultural remote sensing. crop classification random forest sentinel spectral band vegetation index Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1. Introduction Remote sensing has a wide range of applications, from environmental monitoring and climate change studies to agricultural and geological applications. It has the advantage of providing data at different temporal, spatial and spectral resolutions. Recent technological developments in Earth observation satellites and the increasing number of satellites have made access to information on land cover/use, agricultural crop patterns and changes at local and global scales faster, cheaper and more accessible (Khatami et al. 2016 ). A notable example of this progress is the Copernicus programme, which includes Sentinel satellites with different characteristics. Among these, the Sentinel-2 satellite stands out as a passive sensor satellite. Its wide range of bands, its ability to provide images with different spatial resolutions and its frequent acquisition intervals are a significant advantage for monitoring areas with a continuous dynamic structure, such as agriculture. Satellite imagery allows the observation, identification, mapping and evaluation of dynamic agricultural areas with at different temporal and spatial resolutions. The most common method used in agricultural crop type detection with satellite imagery is image classification. Classification accuracy depends on the classification method used (pixel or object based) and the characteristics of the satellite image (low, medium or high spatial resolution; multispectral or hyperspectral), as well as the design and characteristics of the training/test data (number of pixels, statistical distribution of selected samples, etc.) and the selection of the appropriate number of images and bands (Lu and Weng 2018;Kavzoglu 2009 ). Due to the complexity and diversity of crop types and the small spectral differences between different crops, crop classification using a single time-phased remote sensing image is prone to the phenomena of "same object with different spectra" and "different objects with the same spectrum", resulting in misclassification and mixed classification, and the classification accuracy is difficult to improve (Conese and Maselli 1991 ;Gomez et al. 2016). Agricultural crops are grown in different phenological periods according to crop variety, topography and climatic conditions, as well as in similar or very close phenological periods. For this reason, it is necessary to use multi-temporal images in classification studies to detect crops in close or different phenological periods. Crop type classification studies can be performed from a single image or multiple images, but when applied to time series images, they have been shown to perform better than single date mapping methods (Gomez et al. 2016). Time series remote sensing data are widely used in the field of agricultural remote sensing, as they can reflect differences in the growth status of different crops, show different phenological characteristics, and improve separability and classification accuracy (Murty et al. 2003;Zhong et al. 2019 ). Additional data (such as texture filters, vegetation indices and digital elevation models) are sometimes used to improve the distinguishability of the products to be classified, and vegetation indices are one of these additional data (Song et al. 2015 ;Kim and Yeom 2015 ). Specifically, Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), and Enhanced Vegetation Index (EVI) are used to monitor vegetation systems or ecological responses to environmental change (Song et al. 2015 ). Normalized Difference Red Edge (NDRE), The NDRE index includes a red edge band and plays a very important role in vegetation monitoring, providing valuable information on plant health, species differentiation, stress detection and other factors. Its sensitivity to chlorophyll content and reduced susceptibility to atmospheric interference make it a key component in remote sensing applications for agriculture, forestry and ecosystem monitoring. The use of multi-temporal remote sensing data to construct Normalized Difference Vegetation Index (NDVI) and other vegetation index time series, combined with the seasonal rhythms and phenological differences of different crops, has been widely used in crop classification, which has improved the accuracy of crop classification (Kang et al. 2021 ). In addition to the characteristics of remote sensing data, classification algorithms are important to improve the classification accuracy of crop maps. Recently, Random Forest (RF) is a widely used machine learning algorithm consisting of an ensemble of decision trees, and it has been a highly successful machine learning algorithm for classification and regression methods (Biau and Scornet 2016 ). Random forest algorithms have been used to map land cover over large areas using high-resolution satellite imagery time series, with successful results (Pelletier et al. 2016 ). There are many studies in the literature on crop type classification using different satellites, different indices and bands. In these studies using different algorithms, the effects of different indices and bands on the classification results have also been investigated and compared Table 1 . Table 1 Literature review of crop type classification Ref. Study area Satellite Feature Method Classses Subject Kobayashi et al. 2020 Hokkaido, Japan Sentinel-2 Spectral bands 91 Vegetation indices RF Beans,beetroot,grass,maize,potato,wheat Crop classification using spectral indices derived from Sentinel-2A imagery Kang et al. 2021 Hebei, China Sentinel-2 10 Red edge Indices RF Wood, orchard,minor crop, cotton,spring maize, winter wheat-summer maize, greenhouses,water body, cities Land Cover and Crop Classification Based on Red Edge Indices Features Stern et al. 2023 Iowa, USA Landsat NDI5,NDI7,NDTI, STI,NDSVI SVM MINDIST MAXLI RANDTR SAM Corn, soybean, other Comparison of Five Spectral Indices and Six Imagery Classification Techniques for Assessment of Crop Residue Cover Using Four Years of Landsat Pasternak and Filipiak 2002 Lower Silesian, Poland Sentinel-2 12 Vegetation indices PCA RF Beetrooot,maize,wheat,canola sunflower,potato,rye The Evaluation of Spectral Vegetation Indexes and Redundancy Reduction on the Accuracy of Crop Type Detection Sun et al. 2019 Yangzi, China Sentine-1 Sentinel-2 Landsat-8 NDVI,EVI,TVI, NDWI,NDTI Texture ANN RF SVM Forest,maize,rape, urban,water,wheat Using of Multi-Source and Multi-Temporal Remote Sensing Data Improves Crop-Type Mapping in the Subtropical Agriculture Region Zhang et al. 2020 Heilongjiang China Sentinel-2 NDVI,PMI,NDSVI NDRI,NDTI SVM RF CART Rice,corn,water, soybean,potato,beet, forest,building Accessing the temporal and spectral features in crop type mapping using multi-temporal Sentinel-2 imagery Vuolo 2018 Marchfeld lower, Austria Sentinel-2 Spectral bands RF Carrot,maize,onion, pumpkin,soybean, sugarbeet,sunflower, winter cereal How much does multi-temporal Sentinel-2 data improve crop type classification? Most of the studies in the literature are on single crops. In this study, not only were double crops with very close phenological periods studied, but also the changes in the temporal curves of these crops in different vegetation and spectral bands were monitored and compared. The changes in the time series curves of these crops with close phenological periods and the time curves of different vegetation indices were studied and the effects of the index data causing these changes on the classification result were investigated. In addition to the index data, the effect of the spectral bands on the classification accuracy was also investigated. 2. Study Area and Materials 2.1. Study Area The study area was located in the Harran Plain, Şanlıurfa, Turkey (36° 47'-39° 15' E, 36° 40' 37° 41' N) at an altitude between 350 and 500m (Fig. 1 ). The Harran Plain, with a total area of 225,000 ha, is the third largest plain in Turkey and has great agricultural potential. The Harran Plain has a continental climate with mild winters and high summer temperatures. These climate and temperature conditions, together with the increase in irrigated areas in recent years, have led to the cultivation of double crops (two types of crops: wheat_maize (winter_wheat + summer_maize), wheat_cotton (winter_wheat + summer_cotton), lentil_cotton (winter_lentil + summer_cotton), lentil_maize (winter_lentil + summer_maize), during the production season (Bozkurt and Aybek 2016 ). The main crops grown in the region, including wheat, barley, lentil, cotton and maize, cover 95% of the Harran Plain. As these products are not homogeneously distributed within the boundaries of the plain, and considering the plain as a whole, the study area includes the boundaries of the plain. 2.2. Sentinel-2 Sentinel-2A and Sentinel-2B are constellation satellites launched by the European Space Agency (ESA) under the European Commission's (EC) Copernicus programme. Each identical satellite is equipped with a multispectral sensor covering 13 spectral bands with a spatial resolution of 10 m to 60 m and a radiometric resolution of 12 bits. Sentinel-2A was launched in June 2015, followed by Sentinel-2B in March 2017. Information on the spectral bands and reflectance values of the Sentinel-2 satellite images is given in Table 2 . Table 2 Spectral bands of Sentnel-2 images Description Bands Wavelength (µm) Resolution (m) Aerosol B1 458–523 60 Blue B2 458–523 10 Green B3 543–578 10 Red B4 650–680 10 Red-Edge-1 B5 698–713 20 Red-Edge-2 B6 733–748 20 Red-Edge-3 B7 773–793 20 NIR B8 785–900 10 NIR B8a 885 − 875 10 Water Vapour B9 935–955 60 Cirrus B10 1360–1390 60 SWIR-1 B11 1565–1655 20 SWIR-2 B12 2100–2280 20 All bands except B1-B9-B10 were used in this study. However, the three atmospheric bands were not used in this study as they are mainly used for atmospheric corrections and cloud screening. (Drusch et al. 2012 ). The fact that Sentinel-2 imagery contains a large number of spectral bands, and in particular has a temporal resolution of 5 days, provides a significant advantage for monitoring agricultural areas, detecting crop patterns and analyzing changes. (Şimşek and Durduran 2023 ). As the study area is predominantly covered by crops with closely aligned phenological periods, 23 observations were made by Sentinel-2A between March and September 2016. After June, the selected images were cloud-free, while before June, images with a low cloud fraction were selected and cloudy areas were detected and masked. The acquisition dates of the Sentinel-2 satellite images are shown in Table 3 . Table 3 Acquisition time of Sentinel-2 Day of Year (DOY) Acquisition Time Day of Year (DOY) Acquisition Time 37 29 March 2022 212 29 March 2022 57 5 April 2022 220 5 April 2022 87 5 May 2022 230 5 May 2022 102 20 May 2022 240 20 May 2022 117 6 June 2022 252 6 June 2022 132 14 July 2022 262 14 July 2022 152 11 August 2022 272 11 August 2022 167 7 September 2022 282 7 September 2022 177 2 October 2022 292 2 October 2022 192 17 October 2022 317 17 October 2022 202 27 October 2022 202 27 October 2022 2.3. Ground Truth Data Two field studies were carried out in April for wheat, barley and lentil crops and in August for maize and cotton crops. Given the large area of the plain (225,000 ha), the number of samples could not reach the desired level due to time and cost constraints. In addition to field data, parcels declared by farmers were used as ground truth data in the study. 3145 agricultural parcels (cotton: 610, maize: 105, lentil: 340, wheat: 520, lentil_maize: 250, lentil_cotton: 380, wheat_maize: 413, wheat_cotton: 527) were used in the classification study. The Farmer Declaration Parcels (FDP) in Turkey, also known as the Farmer Registration System (FRS), is a government initiative aimed at registering and tracking agricultural activities and farmers in the country (Aydoğdu et al. 2011 ). This system was introduced to improve the efficiency and transparency of agricultural practices and to provide various benefits to registered farmers. The FRS requires farmers to register themselves and their farming activities. This registration process involves providing personal information and details of the land they farm. The parcels of agricultural activity registered by farmers in the system are called FDPs. In addition to the geometric information of the parcels, the system also contains information on the province, district, parcel number, agricultural parcel number, information on the products grown, area, surface, cadastral area, date of cultivation and date of harvest of each parcel. (Şimşek and Durduran 2023 ). When the FDPs were examined, it was found that there were systematic and non-systematic differences between the geometry and attribute information of the parcels and the parcels in the field. Because of these differences, a number of processing steps were applied to the FDPs, and at the end of these steps, ground truth data were produced from the FDPs and used in the classification study together with the data collected in the field. The agricultural calendar for the crops grown in the plain was obtained and is presented graphically. (Fig. 3 .). The calendar provides information on the sowing, growing, dense vegetation and harvesting periods of crops grown in the region. The exact dates of sowing and harvesting vary between plots and farmers. In some cases, there may be a difference of up to 1–3 weeks between the categorized crops; wheat and barley crops are included in the wheat class. Looking at the phenological period of the crops, it is observed that maize and cotton are separated from each other with a dense vegetation stage, while wheat and lentil are in close phenological stages. It was observed that double crops (wheat_maize, lentil_maize, wheat_cotton, lentil_cotton) were in close phenological stages. The calendar shows that the phenological periods of both single and double crops are very close and different indices and bands should be used to distinguish these crops in the classification study. 3. Methodology 3.1. General Architecture Overview After acquiring 23 Sentinel-2 images, the cloudy areas and the shadow areas caused by the clouds were masked. The bands were resampled from 20 m to 10 m. Four different spectral indices were calculated for each Sentinel-2 image, and spectral curves were created for each crop by overlapped these indices with ground truth data. The spectral index curves and spectral bands were then compared for each crop type. Classification studies using the RF algorithm were carried out with different datasets generated from different indices and bands, and the effect of bands and indices on the classification results was compared. Figure 4 shows the flowchart used in this study. 3.2. Satellite Image Processing In data processing, no further geometric correction of the L1C products is required, only atmospheric correction and spatial resolution resampling. (Zheng et al. 2007 ). Sentinel-2 images are provided in Level 1C format and contain above-atmospheric reflectance values. To calculate the actual reflectance values of plants in a classification study, top-of-atmosphere (TOA) values should be converted to bottom-of-atmosphere (BOA) reflectance values (Wilm 2017 ). Atmospheric effects were eliminated by converting TOA values to BOA values using the Sen2cor plugin. After Sen2Cor processing, the L1C TOA reflectance values were converted to Level 2A (L2A) BOA reflectance values (Wilm et al. 2013 ). The conversion of Sentinel-2 data from L1C to L2A improves the accuracy of derived vegetation indices such as NDVI by accounting for atmospheric effects, ensuring that observed changes in NDVI and other indices are more closely related to actual variations in vegetation health. Clouds and cloud shadows in satellite imagery are the main sources of noise that cause problems in image analysis (Kalkan and Maktav 2016 ). Brightness caused by clouds and shadows can affect data analysis and lead to changes in NDVI and other indices. (Zhu and Woodcook 2012). Cloud-covered areas cause anomalies in the image bands as well as in the pixel values of the indices created from these bands, which adversely affects the classification results (Karslen et al. 2021). To eliminate this situation, clouds and shadow areas caused by clouds have been detected and masked by the Sen2core software. The cloud screening and classification part of Sen2Cor is available as source code in the distributed (Skakun et al. 2022 ). Potentially cloudy pixels are subjected to a series of filters based on spectral band thresholds, ratios and index calculations - Normalized Difference Snow Index (NDSI), (NDVI). After atmospheric correction and cloud masking, B5-B7-B8a-B11-B12 were resampled from 20m to 10m. 3.3. Vegetation Indices and Spectral Band Analysis As can be seen from the phenological calendar in Fig. 3 , the single and double crops in the study area have very close phenological periods. The seasonal rhythm and phenological characteristics of different crops can be reflected by the difference in the spectrum or vegetation index of multi-temporal remote sensing data ( Gomez et al. 2016). NDVI is widely used for crop type identification ( Gumma et al. 2020 ). The NDVI index provides information on plant health and development ( Morsy and Hadı 2022 ). In order to see this phenological closeness on a crop by crop basis, time series NDVI plots were first created. Each plot was overlapped with multi-temporal NDVI images and the median NDVI values of each plot in the time series were calculated. After this process, the vegetative development and change of each plot was determined and the characteristics of the NDVI curves showing the variability over time were revealed. The characteristic NDVI curves for each crop were checked against the phenological calendar and the spectral reflectance values of each crop collected during the field study. When analysing the NDVI time series curves of each crop, it can be seen that lentils and wheat, which are winter crops, are separated from each other, while maize and cotton, which are summer crops, show new different time curves (Fig. 6 ). With the NDVI index, these four crops (lentil, wheat, maize, cotton) have different reflectance values on the same dates and their temporal curves are separated from each other. Therefore, no studies have been carried out with different vegetation indices in single crops. When analyzing the NDVI time series curves, wheat_maize-wheat_cotton crops and lentil_maize-lentil_cotton crops differ from each other between day 57. and day 132.However due to the later sowing of lentil, but it can be seen that the reflectance values for all crop types are very close to each other between day 220. and day 317. Figure 7 . It was observed that the NDVI time series curves were similar for both crops, especially in the second period of the year. The effect of different vegetation indices on the time series curves of the crops and the classification results were investigated in this study. There are nearly 100 vegetation indices in the literature. In this study, NDVI, NDRE, SAVI and NDRE indices, which are traditional vegetation indices, were used Table 4 . Table 4 Acquisition time of Sentinel-2 Spectral Index Full Name Formula Description Reference NDVI Normalized Difference Vegetation Index (NIR - R)/(NIR + R) Commonly used for assessing overall vegetation health and density Tucker 1979 EVI Enhanced vegetation index 2.5*(NIR - R)/(NIR + 6*R-7.5*B + 1) Similar to NDVI but designed to minimize atmospheric influences and improve sensitivity in high biomass regions Huete et al. 2002 SAVI Soil-Adjusted Vegetation Index (NIR - R)/(NIR + R + L)*(1 + L) Designed to minimize soil background influences in the vegetation index Huete 1988 NDRE Normalized Difference Red Edge (NIR – Redge2)/(NIR + Redge2) Particularly sensitive to changes in chlorophyll content, making it valuable for detecting early signs of stress Barnes 2000 Comparing the NDVI, SAVI and NDRE time series curves of lentil_maize and lentil_cotton, it is observed that the curves are close to each other as in the case of NDVI, this is also the case for wheat_maize and wheat_cotton. It was observed that the SAVI index did not change in relation to the NDVI index. As a result of the EVI index, the time series curves for four crops show some separation compared to the SAVI and NDVI indices. When analyzing the time series curves of the NDRE index, it can be seen that the curves for the two crops wheat_maize-wheat_cotton and lentil_maize-lentil_cotton are more separated compared to the other indices Fig. 7 . In summary, the NDRE index values performed better than other indices in separating the time series curves of crops. In addition to observing the changes in the time series curves of crops with different vegetation indices, the study also were analysed the time series reflectance values of each spectral band for all crop types. Comparing the time series reflectance values of each spectral band for all crop types, it is seen that the B6 (Red Edge-2) and B11 (SWIR-1) bands are separated from the other bands Fig. 8 . Different datasets were generated using 4 different vegetation indices and 10 spectral bands Table 5 . Classification studies were performed on these datasets and the results were compared. It was also investigated if the indices and bands that create time series curve differences in the generated datasets affect the classification result. Table 5 Data sets and number of bands Data Set Features Number of bands Data Set Features Number of bands Data Set (DS-1) NDVI 21 Data Set (DS-9) (4BAND)-EVI 105 Data Set (DS-2) NDRE 21 Data Set (DS-10) (4 BAND)-B6-B11 126 Data Set (DS-3) SAVI 21 Data Set (DS-11) (4 BAND)-B6-B11-NDRE 147 Data Set (DS-4) EVI 21 Data Set (DS-12) B2-B3-B4-B5-B6-B7-B8- B8A-B11-B12 (10 BAND) 210 Data Set (DS-5) B2-B3-B4-B8 (4BAND) 84 Data Set (DS-13) (10 BAND)-NDVI 231 Data Set (DS-6) (4BAND)-NDVI 105 Data Set (DS-14) (10 BAND)-NDRE 231 Data Set (DS-7) (4BAND)-SAVI 105 Data Set (DS-15) (10 BAND)-SAVI 231 Data Set (DS-8) (4BAND)-NDRE 105 Data Set (DS-16) (10 BAND)-EVI 231 3.4 Classification Using Random Forest Random Forest (RF) is a combinatorial ensemble learning classification technique. RF is an improved algorithm based on an ensemble learning technique that builds multiple CARTs. (Breiman 2001 ) In fact, RF has been very successful as a general purpose classification and regression method (Biau and Scornet 2016 ). RF fits many classification trees to training data sets and then combines the predictions of all the trees to make a final decision. RF is an ensemble classifier that is currently widely used in remote sensing studies due to its classification accuracy (Belgiu and Dragut 2016 ). Higher accuracies have been achieved with RF compared to other machine learning algorithms in many crop mapping studies (Tatsumi et al. 2015 ). RF is known to work efficiently on large datasets with a large number of input variables to estimate which variables are significant in the classification process, and is relatively robust to noise and outliers. 44 Many examples of the use of this algorithm can be found in the literature (Feng et al. 2019 ). Hyperparameter optimisation is the process of finding the optimal combination of parameters for a machine learning algorithm according to specified success criteria. Hyperparameter optimisation aims to achieve a balance between overlearning and under learning by balancing high model success and model complexity. The original RF has two hyperparameters including the number of trees (ntree) and the number of variables used to partition the nodes (mtry). Several studies have shown that satisfactory results can be achieved with the default parameters (Zhang and Roy 2017 ). In this study, to find the optimal RF model for classification, a range of values for both parameters were tested and evaluated using the grid search method for each dataset: ntree = 100, 200, 400, 600,800,1000; mtry = 1:20. After determining the optimal parameters of the algorithm to be used in the classification process, the k-fold cross-validation method is used. K-fold cross-validation allows to see whether the high performance of the model is random or not. In this method, the data set is divided into k parts and k-1 subset is used to train the model and the remaining subset is used to calculate the accuracy of the model. The process is repeated k times, each time using different pieces of training and test data. The average of the accuracy values obtained represents the accuracy of the model, and in this study the k value is taken as 5 (Kohavi 1995 ). Table 6 Hyper parameter values for different data sets Data Set Number of bands Ntree Mtry Data Set (DS: 1–4) 21 100 4 Data Set (DS-5) 84 200 9 Data Set (DS: 6–9) 105 200 10 Data Set (DS-10) 126 400 11 Data Set (DS-11) 147 400 12 Data Set (DS-12) 210 800 14 Data Set (DS: 13–16) 231 800 15 4. Results and Discussion In order to assess the accuracy of classification performance, there are many metrics available in the literature. In this work, PA, UA, OA and F1-score have been used to assessment class accuracy Table 7 . PA (Producer Accuracy): The ratio of correctly predicted positive observations to the total predicted positives. It measures the model's ability to correctly identify positive instances among the instances it predicted as positive. UA (User Accuracy): The ratio of correctly predicted positive observations to the total actual positives. It measures the model's ability to correctly identify positive instances. OA (Overall Accuray): The ratio of correctly predicted instances (both positive and negative) to the total instances. It provides a general measure of the model's correctness. F1 Score: The harmonic mean of precision and recall. It provides a balance between precision and recall, especially useful when dealing with imbalanced datasets. These classification accuracy indicators can reflect the overall classification accuracy and specific type identification accuracy of remote sensing images from different aspects (Congatol, 1991). Table 7. Accuracy assessment for all data sets For wheat and lentil crops it is observed that the accuracy is 90% and above in 16 datasets. While the lowest accuracy value belongs to DS-3 dataset, the highest accuracy values belong to DS-7 and DS-14 datasets. It was observed that the classification results with DS-5 and DS-10 datasets were not different. It was observed that wheat and lentil products, which have different spectral separation curves, achieved high accuracy as a result of classification with each dataset. For maize and cotton crops, it can be seen that the accuracy is 90% and above in 16 datasets. While the lowest accuracy value for cotton is DS-5, the highest accuracy value belongs to DS-14. While the lowest accuracy value for maize is DS-3 dataset, the highest accuracy value belongs to DS-14 dataset. Since the time series curves of maize and cotton crops are different from each other, like wheat and lentil crops, the accuracy values are high in each dataset. For these classes, it was found that increasing the number of bands did not increase the accuracy too much. The DS-5 dataset (81–82%) gives the lowest accuracy value for all double crops and the DS-14 dataset (94.8%) gives the highest accuracy value. It can be observed that the DS-2 dataset with NDRE values gives a higher accuracy compared to the datasets with other indices (DS-1-3-4). This high accuracy value is due to the NDRE index, which performs better than other indices in separating the time series curves of crops. In the DS-5 (81–82%) dataset, double crops remained with low accuracy, in the DS-12 (84–85%) dataset, double crops remained with low accuracy and in the DS-12 (84–85%) dataset, the accuracy increased by about 3% but remained insufficient Fig. 10 . When evaluating Fig. 10 , the lowest overall accuracy value belongs to DS-5 (86.4%) and DS-3 (87.02%) datasets. The highest accuracy value belongs to DS-14 (94.77%), as in the product base. Comparing the DS-7 (91.24%) and DS-14 (94.77%) datasets, there is an increase of approximately 3%. Comparing DS-7 (91.24%) and DS-11 (93.39%) data sets, band 6 and band 11, which are separated from other bands in all products, increase the accuracy value by 2%. When comparing the DS-11 (93.39%) and DS-14 (94.77%) data sets, it is also seen that the results are close to each other. It can be seen that the datasets containing the NDRE index (DS-2-7-11-14) and the datasets containing 10 bands (DS-12-13-14-15-16) have an accuracy value above 90%. The comparison of the DS-5 (81–82%) dataset with the DS-2 (89–90%) dataset shows that the accuracy of double crops increased by about 8% and the NDRE index improved the classification result of crops with close phenological stages. When DS-2 was compared with the DS-7 dataset, it was observed that the + 4 band in the time series NDRE dataset did not have much effect on the classification result both for double crops and for the overall accuracy. Comparing DS-7 (91%) with DS-11 (92%), it can be seen that band 6 and band 11 slightly increase the classification accuracy of double crops. For crops with similar time series, data sets with only 4 bands (DS-5) and only 10 bands (DS-12) did not reach a high accuracy for the classification of crops with close phenological stages. It is observed that the data sets containing the NDRE index, which performs better than other indices in separating the time series curves of crops, increase the overall accuracy by 2–3% compared to other data sets, and the accuracy increases by 2% with the addition of the 6th and 11th bands to these data set. The most important factor in increasing this accuracy is Band 6 (Red Edge-2), which is also used in the NDRE index. 5. Conclusion This study investigated the effect of vegetation indices and spectral bands on the classification of agricultural crops. Time series curves of classes with different vegetation indices were generated and compared, and time series reflectance values of each spectral band were also observed for all crops. The study shows the importance of time series curves (phenological period) generated from multi-temporal images of crops. In the time series curves generated with different vegetation indices, it was found that the crops that differ (with a unique curve) have a high accuracy value with optimal data sets, while increasing the number of features and bands in the input data set has almost no effect on the accuracy value of these crops. For crops with very close phenological periods, indices that reveal the difference between the time series curves (NDRE-EVI) were found to increase the classification result. When comparing the time series reflectance values of each spectral band for all crop types, except the crop indices, it was observed that the B6 (Red Edge-2) and B11 (SWIR-1) bands were separated from the other bands and increased the classification result by 2% when included in the dataset. This result showed that B6 (Red Edge-2) and B11 (SWIR-1) bands should be used in agricultural crop type classification studies, especially for crops covering close phenological stages. In particular, for classification studies carried out in large working areas such as this one, by determining the bands with indices that separate the time series curves of the crops, the bands and indices that will give the highest accuracy as a result of the classification can be determined and time can be saved by obtaining maximum accuracy with minimum data in classification studies. Future studies will extend the scope of the study by using different algorithms for different crop indices, different crop types and different study areas. Declarations Author Contribution a) Fatih Fehmi ŞİMŞEKThe article is the work of a single author, and the content of the article is entirely written by him. References Aydoğdu M, Akçar HT, Çullu MA (2011) Coğrafi bilgi sistemleri CBS ve uzaktan algılama UA kullanılarak çiftçi kayıt sistemi çks verilerinin analizi ile pamuk ve mısır primlerinin ödenmesi Şanlıurfa-Harran İlçesi örneği. Jeodezi ve Jeoinformasyon Dergisi 104(–S):47–52 Barnes EM, Clarke TR, Richards SE, Colaizzi PD, Haberland J, Kostrzewski M, Waller P, Choi C, Riley E, Thompson T, Lascano RJ, Li H, Moran MS (2000) Coincident detection of crop water stress, nitrogen status, and canopy density using ground-based multispectral data. 5th International Conference on Precision Agriculture, Bloomington, 16–19, 1–15 Belgiu M, Dragut 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 Biau G, Scornet E (2016) A random forest guided tour. Test 25:197–227. https://doi.org/10.1007/s11749-016-0481-7 Bozkurt M, Aybek A (2016) Şanlıurfa İli Harran Ovasının tarımsal yapı ve mekanizasyon özellikleri, KSÜ Doğa Bilimleri Dergisi 19(3),2016, pp. 319–331 Breiman L (2001) Random Forests. Mach Learn 45:5–32. http://dx.doi.org/10.1023/A:1010933404324 Conese C, Maselli F (1991) Use of multi-temporal information to improve classification performance of TM scenes in complex terrain. ISPRS J photogrammetry Remote Sens 46(4):187–197. https://doi.org/10.1016/0924-2716(91)90052-W Congalton RG (1991) A review of assessing the accuracy of classifications of remotely sensed data. Remote Sens 37(1):35–46. https://doi.org/10.1016/0034-4257(91)90048-B Drusch M, Bello UD, Carlier S, Colin O, Fernandez V, Gascon F, Hoersch B, Isola C, Laberinti P, Martimort P, Meygret A, Spoto F, Sy O, Marchese F, Bargellini PL (2012) Sentinel-2: ESA's optical high-resolution mission for GMES operational services. Remote Sens Environ 120:25–36. https://doi.org/10.1016/j.rse.2011.11.026 Feng S, Zhao J, Liu T, Zhang H, Zhang Z, Guo X (2019) Crop type identification and mapping using machine learning algorithms and Sentinel-2 time series data. IEEE J Sel Top Appl Earth Observations Remote Sens 12:3295–3306. https://doi.org/10.1109/JSTARS.2019.2922469 Gómez C, White JC, Wulder MA (2016) Optical remotely sensed time series data for land cover classification: A review. ISPRS J Photogrammetry Remote Sens 116:55–72. https://doi.org/10.1016/j.isprsjprs.2016.03.008 Gumma MK, Tummala K, Dixit S, Collivignarelli F, Holecz F, Kolli RN (2020) Whitbread AM. Crop type identification and spatial mapping using Sentinel-2 satellite data with focus on field-level information. Geocarto Int 37(11):1–17. https://doi.org/10.1080/10106049.2020.1805029 Huete A, Didan K, Miura T, Rodriguez EP, Gao X, Ferreira LG (2002) Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens Environ 83:195–213. https://doi.org/10.1016/S0034-4257(02)00096-2 Huete AR (1988) A soil adjusted vegetation index (SAVI). Remote Sens Environ 25:295–309. http://dx.doi.org/10.1016/0034-4257(88)90106-X Kalkan K, Maktav D (2016) Landsat-8 görüntülerinden gölge ve bulut belirleme. VI. Uzaktan Algılama ve Coğrafi. Bilgi Sistemleri Sempozyumu, Adana, Türkiye Kang Y, Meng Q, Liu M, Zou Y, Wang X (2021) Crop classification based on red edge features analysis of GF-6 WFV data. Sensors 21(13). https://doi.org/10.3390/s21134328 Karlsen SR, Stendardi L, Tømmervik H, Nilsen L, Arntzen I, Cooper EJ (2021) Time-series of cloud-free Sentinel-2 NDVI data used in mapping the onset of growth of Central Spitsbergen. Svalbard Remote Sens 13(15):3031. https://doi.org/10.3390/rs13153031 Kavzoglu T (2009) Increasing the accuracy of neural network classification using refined training data. Environ Model Softw 24(7):850–858. https://doi.org/10.1016/j.envsoft.2008.11.012 Kobayashi N, Tani H, Wang X, Sonobe R (2020) Crop classification using spectral indices derived from Sentinel-2A imagery. J Inform Telecommunication 4(1):67–90. https://doi.org/10.1080/24751839.2019.1694765 Khatami R, Mountrakis G, Stehman SV (2016) A meta-analysis of remote sensing research on supervised pixel-based land-cover image classification processes: General guidelines for practitioners and future research. Remote Sens Environ 177:89–100. https://doi.org/10.1016/j.rse.2016.02.028 Kim HO, Yeom JM (2015) Sensitivity of vegetation indices to spatial degradation of Rapid Eye imagery for paddy rice detection: a case study of South Korea. GIScience Remote Sens 52(1):1–17. https://doi.org/10.1080/15481603.2014.1001666 Kohavi R (1995) A Study of cross-validation and bootstrap for accuracy estimation and model selection. Proceedings of the 14th international joint conference on Artificial intelligence 2, 1137–1143 Lu D, Weng Q (2007) A survey of image classification methods and techniques for improving classification performance. Int J Remote Sens 28:823–870 https://doi.org/10.1080/01431160600746456 Morsy S, Hadı M (2022) Impact of land use/land cover on land surface temperature and its relationship with spectral indices in Dakahlia Governorate, Egypt. Int J Eng Geosci 7(3):272–282. https://doi.org/10.26833/ijeg.978961 Murthy CS, Raju PV, Badrinath KV (2003) Classification of wheat crop with multi-temporal images: performance of maximum likelihood and artificial neural networks. Int J Remote Sens 24:4871–4890. https://doi.org/10.1080/0143116031000070490 Pasternak M, Filipiak KP (2022) The evaluation of spectral vegetation indexes and redundancy reduction on the accuracy of crop type detection. Appl Sci 12(10):5067. https://doi.org/10.3390/app12105067 Pelletier C, Valeroa S, Ingladaa C, Championb N, Dedieua G (2016) Assessing the robustness of random forests to map land cover with high resolution satellite image time series over large areas. Remote Sens Environ 187:156–168. http://dx.doi.org/10.1016/j.rse.2016.10.010 Skakun S, Wevers J, Brockmann C, Doxani G, Aleksandrov M, Batič M, Frantz D, Gascon F, Chova LG, Hagolle O, Puigdollers OL, Louis J, Lubej M, García GM, Osman J, Peressutti D, Pflug B, Puc J, Richter R, Roger JC, Scaramuzza P, Vermote E, Vesel N, Zupanc A, Žust L (2022) Cloud mask intercomparison exercise (CMIX): An evaluation of cloud masking algorithms for Landsat 8 and Sentinel-2. Remote Sens Environ 274:112990. https://doi.org/10.1016/j.rse.2022.112990 Sun C, Bian Y, Zhou T, Pan J (2019) Using of multi-source and multi-temporal remote sensing data improves crop-type mapping in the subtropical agriculture region. Sensors 19(10):2401. https://doi.org/10.3390/s19102401 Stern AJ, Daughtry CST, Hunt ERJr, Gao F (2023) Comparison of five spectral indices and six imagery classification techniques for assessment of crop residue cover using four years of Landsat imagery. Remote Sens 15:4596. https://doi.org/10.3390/rs15184596 Song D, Huang C, Sexton JO, Channan S, Feng M, Townshend JR (2015) Case studies from the Eastern United States and Central Brazil. ISPRS J Photogrammetry Remote Sens 103:81–92. https://doi.org/10.1016/j.isprsjprs.2014.09.005 . Use of Landsat and corona data for mapping forest cover change from the mid-1960s to 2000s Sonobe R, Yamaya Y, Tani H, Wang X, Kobayashi N, Mochizuki K (2018) Crop classification from Sentinel-2 derived vegetation indices using ensemble learning. J Appl Remote Sens 12(2). https://doi.org/10.1117/1.JRS.12.026019 Şimşek FF, Durduran SS (2023) Açık kaynak kodlu Eo-learn kütüphanesi ve çok zamanlı Sentinel-2 görüntüleri ile tarımsal ürün sınıflandırması. Jeodezi ve Jeoinformasyon Dergisi 10(1):45–62. https://doi.org/10.9733/JGG.2023R0004.T Tatsumi M, Yamashiki Y, Torres MAC, Taipe CLR (2015) Crop classification of upland fields using random forest of time-series Landsat 7 ETM + data. Comput Electron Agric 115:171–179. https://doi.org/10.1016/j.compag.2015.05.001 Tucker CJ (1979) Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens Environ 8:127–150. http://dx.doi.org/10.1016/0034-4257(79)90013-0 Vuolo F, Neuwirth M, Immitzer M, Atzberger C, Ng W (2018) How much does multi-temporal Sentinel-2 data improve crop type classification? Int J Appl Earth Obs Geoinf 72:122–130. https://doi.org/10.1016/j.jag.2018.06.007 Wilm UM (2017) Sen2Cor configuration and user manual. 2017, 9–12 Wilm UM, Louis J, Richter R, Gascon F, Niezette M (2013) Sentinel-2 Level-2A prototype processor: Architecture, algorithms and first results. ESA, Living Planet Symposium ESA-SP-722 Zhang H, Kang J, Xu X, Zhang L (2020) Accessing the temporal and spectral features in crop type mapping using multi-temporal Sentinel-2 imagery: A case study of Yi'an County, Heilongjiang province, China. Comput Electron Agric 176:105618. https://doi.org/10.1016/j.compag.2020.105618 Zhang HK, Roy DP (2017) Using the 500 m MODIS land cover product to derive a consistent continental scale 30 m Landsat land cover classification. Remote Sens Environ 197:15–34. https://doi.org/10.1016/j.rse.2017.05.024 Zheng H, Du P, Chen J, Xia J, Li E, Xu Z, Li X, Yokoya N (2007) Performance evaluation of downscaling sentinel-2 imagery for land use and land cover classification by spectral-spatial features. Remote Sens 9(12):1274. https://doi.org/10.3390/rs9121274 Zhong L, Hu L, Zhou H (2019) Deep learning based multi-temporal crop classification. Remote Sens Environ 221:430–443. https://doi.org/10.1016/j.rse.2018.11.032 Zhu Z, Woodcock CE (2012) Object-based cloud and cloud shadow detection in Landsat imagery. Remote Sens Environ 118:83–94. https://doi.org/10.1016/j.rse.2011.10.028 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted 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-3910868","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":270196768,"identity":"d57f610e-a4eb-4b8e-bd9f-9e08aa7cbd66","order_by":0,"name":"Fatih Fehmi Şimşek","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+ElEQVRIiWNgGAWjYBAC9gbGB8wMDBIMbEAOY0MFkGRmbsCrhecAswGSljMgLYxEaYEAxsY2MEVAC9DQx4U5FvZ8YscfPpw5rzaavx2o5UfFNnxamI1nbpNIbJPOMTbcuO147ozDjA2MPWdu49Riz8B/TJp3m0QCm3QOm+TDbcdyG4BamBnbcGsB2sL+G6jFnk06/fnPh3OO5c4nQgsbM1ALY5t0ghnjxoaa3A0EtTAzM4McBvaL5IxjB3I3ArUcxOcXHvZmxs+82+rs5WenP/zYU1OXO+/84YMPflTg1sLAjMo9DCYP4FaPCepIUTwKRsEoGAUjBAAAlndSageWFioAAAAASUVORK5CYII=","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Fatih","middleName":"Fehmi","lastName":"Şimşek","suffix":""}],"badges":[],"createdAt":"2024-01-30 14:30:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3910868/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3910868/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":50517073,"identity":"2277f216-94f2-47b0-b548-3b6c48cced2a","added_by":"auto","created_at":"2024-02-01 17:05:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":8370226,"visible":true,"origin":"","legend":"\u003cp\u003eStudy are\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/63314558bc9476e6bb72f2de.png"},{"id":50517526,"identity":"96b699f3-7973-41c1-a92c-e832f8b1d545","added_by":"auto","created_at":"2024-02-01 17:13:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":282093,"visible":true,"origin":"","legend":"\u003cp\u003eFig.3. Approximate crop calendar in the region.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/fca785f1db5cc2baf6b9090e.png"},{"id":50517528,"identity":"1b4ac0c9-49ba-42b4-b4c9-549998205721","added_by":"auto","created_at":"2024-02-01 17:13:19","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":510249,"visible":true,"origin":"","legend":"\u003cp\u003eFig.4. Workflow of study\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/356b4c50f0bcee21361bc80b.png"},{"id":50517529,"identity":"41083ca0-5ea5-4b34-aedb-556350314cf4","added_by":"auto","created_at":"2024-02-01 17:13:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":6834998,"visible":true,"origin":"","legend":"\u003cp\u003eFig.5. Cloud detection and masking.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/51b6acdc690b16663b89a57c.png"},{"id":50517527,"identity":"76280e36-d1de-46dc-b354-aa00df6f19a2","added_by":"auto","created_at":"2024-02-01 17:13:19","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":171402,"visible":true,"origin":"","legend":"\u003cp\u003eFig.6. NDVI time series curves of single crops\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/5a982f38409c198558e7ad04.png"},{"id":50517066,"identity":"c55e3da6-506a-456c-b9f8-e0622d3a6bae","added_by":"auto","created_at":"2024-02-01 17:05:19","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":817017,"visible":true,"origin":"","legend":"\u003cp\u003eFig.7. NDVI time series curves of double crops\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/dfd558645c829f00b5cb66ff.png"},{"id":50517072,"identity":"ab1b796c-4f0a-4431-9711-d880f63f1e74","added_by":"auto","created_at":"2024-02-01 17:05:19","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":754039,"visible":true,"origin":"","legend":"\u003cp\u003eFig.7. NDVI-EVI-SAVI-NDRE time series curves of double crops\u003c/p\u003e","description":"","filename":"image77.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/1d3d839657d7f788e50ccfc6.png"},{"id":50517070,"identity":"f29489a7-c76f-4e82-9b4c-f8c514160682","added_by":"auto","created_at":"2024-02-01 17:05:19","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":805316,"visible":true,"origin":"","legend":"\u003cp\u003eFig.8. Reflection of different spectral bands for double crop type as time series.\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/29587f76cf4bf9238a52024f.png"},{"id":50517530,"identity":"776a05c5-0291-4b0f-9ffa-4c415f438141","added_by":"auto","created_at":"2024-02-01 17:13:19","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":2931269,"visible":true,"origin":"","legend":"\u003cp\u003eFig.9. Classification result produced with DS-16 dataset\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/0fdfb57314c4a1eb70f07f57.png"},{"id":50517075,"identity":"7c4366ea-5a79-420a-8ef0-1815e29afdf4","added_by":"auto","created_at":"2024-02-01 17:05:19","extension":"jpeg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2472276,"visible":true,"origin":"","legend":"\u003cp\u003eFig.10. Overall accuracy for all data sets\u003c/p\u003e","description":"","filename":"image10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/a85b45dcaf2e53b51e18b21f.jpeg"},{"id":50518483,"identity":"822bd152-8662-4d2b-8da3-7af19ff3d7d1","added_by":"auto","created_at":"2024-02-01 17:45:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9083930,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3910868/v1/069ebb77-d8b8-4d42-8afc-21ed7c9974f0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Investigation of the Effect of Spectral Bands and Vegetation Index Selection on Agricultural Crop Type Classification (Especially for Double Crops)","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eRemote sensing has a wide range of applications, from environmental monitoring and climate change studies to agricultural and geological applications. It has the advantage of providing data at different temporal, spatial and spectral resolutions. Recent technological developments in Earth observation satellites and the increasing number of satellites have made access to information on land cover/use, agricultural crop patterns and changes at local and global scales faster, cheaper and more accessible (Khatami et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). A notable example of this progress is the Copernicus programme, which includes Sentinel satellites with different characteristics. Among these, the Sentinel-2 satellite stands out as a passive sensor satellite. Its wide range of bands, its ability to provide images with different spatial resolutions and its frequent acquisition intervals are a significant advantage for monitoring areas with a continuous dynamic structure, such as agriculture.\u003c/p\u003e \u003cp\u003eSatellite imagery allows the observation, identification, mapping and evaluation of dynamic agricultural areas with at different temporal and spatial resolutions. The most common method used in agricultural crop type detection with satellite imagery is image classification. Classification accuracy depends on the classification method used (pixel or object based) and the characteristics of the satellite image (low, medium or high spatial resolution; multispectral or hyperspectral), as well as the design and characteristics of the training/test data (number of pixels, statistical distribution of selected samples, etc.) and the selection of the appropriate number of images and bands (Lu and Weng 2018;Kavzoglu \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDue to the complexity and diversity of crop types and the small spectral differences between different crops, crop classification using a single time-phased remote sensing image is prone to the phenomena of \"same object with different spectra\" and \"different objects with the same spectrum\", resulting in misclassification and mixed classification, and the classification accuracy is difficult to improve (Conese and Maselli \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e1991\u003c/span\u003e;Gomez et al. 2016). Agricultural crops are grown in different phenological periods according to crop variety, topography and climatic conditions, as well as in similar or very close phenological periods. For this reason, it is necessary to use multi-temporal images in classification studies to detect crops in close or different phenological periods. Crop type classification studies can be performed from a single image or multiple images, but when applied to time series images, they have been shown to perform better than single date mapping methods (Gomez et al. 2016). Time series remote sensing data are widely used in the field of agricultural remote sensing, as they can reflect differences in the growth status of different crops, show different phenological characteristics, and improve separability and classification accuracy (Murty et al. 2003;Zhong et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAdditional data (such as texture filters, vegetation indices and digital elevation models) are sometimes used to improve the distinguishability of the products to be classified, and vegetation indices are one of these additional data (Song et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2015\u003c/span\u003e;Kim and Yeom \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Specifically, Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), and Enhanced Vegetation Index (EVI) are used to monitor vegetation systems or ecological responses to environmental change (Song et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Normalized Difference Red Edge (NDRE), The NDRE index includes a red edge band and plays a very important role in vegetation monitoring, providing valuable information on plant health, species differentiation, stress detection and other factors. Its sensitivity to chlorophyll content and reduced susceptibility to atmospheric interference make it a key component in remote sensing applications for agriculture, forestry and ecosystem monitoring. The use of multi-temporal remote sensing data to construct Normalized Difference Vegetation Index (NDVI) and other vegetation index time series, combined with the seasonal rhythms and phenological differences of different crops, has been widely used in crop classification, which has improved the accuracy of crop classification (Kang et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In addition to the characteristics of remote sensing data, classification algorithms are important to improve the classification accuracy of crop maps. Recently, Random Forest (RF) is a widely used machine learning algorithm consisting of an ensemble of decision trees, and it has been a highly successful machine learning algorithm for classification and regression methods (Biau and Scornet \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Random forest algorithms have been used to map land cover over large areas using high-resolution satellite imagery time series, with successful results (Pelletier et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThere are many studies in the literature on crop type classification using different satellites, different indices and bands. In these studies using different algorithms, the effects of different indices and bands on the classification results have also been investigated and compared Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\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\u003eLiterature review of crop type classification\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\u003eRef.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStudy area\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSatellite\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFeature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMethod\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eClassses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSubject\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKobayashi et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2020\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHokkaido, \u003c/p\u003e \u003cp\u003eJapan\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSpectral bands\u003c/p\u003e \u003cp\u003e91 Vegetation indices\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBeans,beetroot,grass,maize,potato,wheat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCrop classification using spectral indices derived\u003c/p\u003e \u003cp\u003efrom Sentinel-2A imagery\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKang et al.\u003c/p\u003e \u003cp\u003e2021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHebei,\u003c/p\u003e \u003cp\u003eChina\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10 Red edge Indices\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eWood, orchard,minor crop, cotton,spring maize, winter wheat-summer maize, greenhouses,water body, cities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eLand Cover and Crop Classification Based on Red Edge Indices\u003c/p\u003e \u003cp\u003eFeatures\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStern et al.\u003c/p\u003e \u003cp\u003e2023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIowa,\u003c/p\u003e \u003cp\u003eUSA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLandsat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNDI5,NDI7,NDTI,\u003c/p\u003e \u003cp\u003eSTI,NDSVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003cp\u003eMINDIST\u003c/p\u003e \u003cp\u003eMAXLI\u003c/p\u003e \u003cp\u003eRANDTR\u003c/p\u003e \u003cp\u003eSAM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCorn, soybean, other\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eComparison of Five Spectral Indices and Six Imagery Classification Techniques for Assessment of Crop Residue Cover Using Four Years of Landsat\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePasternak and Filipiak 2002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLower Silesian, \u003c/p\u003e \u003cp\u003ePoland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12 Vegetation indices\u003c/p\u003e \u003cp\u003ePCA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBeetrooot,maize,wheat,canola\u003c/p\u003e \u003cp\u003esunflower,potato,rye\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eThe Evaluation of Spectral Vegetation Indexes and Redundancy\u003c/p\u003e \u003cp\u003eReduction on the Accuracy of Crop Type Detection\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSun et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYangzi, \u003c/p\u003e \u003cp\u003eChina\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentine-1\u003c/p\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003cp\u003eLandsat-8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNDVI,EVI,TVI,\u003c/p\u003e \u003cp\u003eNDWI,NDTI\u003c/p\u003e \u003cp\u003eTexture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eANN\u003c/p\u003e \u003cp\u003eRF\u003c/p\u003e \u003cp\u003eSVM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eForest,maize,rape,\u003c/p\u003e \u003cp\u003eurban,water,wheat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eUsing of Multi-Source and Multi-Temporal Remote Sensing Data Improves Crop-Type Mapping in the Subtropical Agriculture Region\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eZhang et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2020\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHeilongjiang\u003c/p\u003e \u003cp\u003eChina\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNDVI,PMI,NDSVI\u003c/p\u003e \u003cp\u003eNDRI,NDTI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003cp\u003eRF\u003c/p\u003e \u003cp\u003eCART\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRice,corn,water,\u003c/p\u003e \u003cp\u003esoybean,potato,beet,\u003c/p\u003e \u003cp\u003eforest,building\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAccessing the temporal and spectral features in crop type mapping using multi-temporal Sentinel-2 imagery\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVuolo 2018\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMarchfeld lower, \u003c/p\u003e \u003cp\u003eAustria\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSentinel-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSpectral bands\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCarrot,maize,onion,\u003c/p\u003e \u003cp\u003epumpkin,soybean,\u003c/p\u003e \u003cp\u003esugarbeet,sunflower, winter cereal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eHow much does multi-temporal Sentinel-2 data improve crop type\u003c/p\u003e \u003cp\u003eclassification?\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\u003eMost of the studies in the literature are on single crops. In this study, not only were double crops with very close phenological periods studied, but also the changes in the temporal curves of these crops in different vegetation and spectral bands were monitored and compared. The changes in the time series curves of these crops with close phenological periods and the time curves of different vegetation indices were studied and the effects of the index data causing these changes on the classification result were investigated. In addition to the index data, the effect of the spectral bands on the classification accuracy was also investigated.\u003c/p\u003e"},{"header":"2. Study Area and Materials","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Study Area\u003c/h2\u003e \u003cp\u003eThe study area was located in the Harran Plain, Şanlıurfa, Turkey (36\u0026deg; 47'-39\u0026deg; 15' E, 36\u0026deg; 40' 37\u0026deg; 41' N) at an altitude between 350 and 500m (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The Harran Plain, with a total area of 225,000 ha, is the third largest plain in Turkey and has great agricultural potential. The Harran Plain has a continental climate with mild winters and high summer temperatures. These climate and temperature conditions, together with the increase in irrigated areas in recent years, have led to the cultivation of double crops (two types of crops: wheat_maize (winter_wheat\u0026thinsp;+\u0026thinsp;summer_maize), wheat_cotton (winter_wheat\u0026thinsp;+\u0026thinsp;summer_cotton), lentil_cotton (winter_lentil\u0026thinsp;+\u0026thinsp;summer_cotton), lentil_maize (winter_lentil\u0026thinsp;+\u0026thinsp;summer_maize), during the production season (Bozkurt and Aybek \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The main crops grown in the region, including wheat, barley, lentil, cotton and maize, cover 95% of the Harran Plain. As these products are not homogeneously distributed within the boundaries of the plain, and considering the plain as a whole, the study area includes the boundaries of the plain.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Sentinel-2\u003c/h2\u003e \u003cp\u003eSentinel-2A and Sentinel-2B are constellation satellites launched by the European Space Agency (ESA) under the European Commission's (EC) Copernicus programme. Each identical satellite is equipped with a multispectral sensor covering 13 spectral bands with a spatial resolution of 10 m to 60 m and a radiometric resolution of 12 bits. Sentinel-2A was launched in June 2015, followed by Sentinel-2B in March 2017. Information on the spectral bands and reflectance values of the Sentinel-2 satellite images is given in 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\u003eSpectral bands of Sentnel-2 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 \u003cp\u003eDescription\u003c/p\u003e \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 (\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eResolution (m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAerosol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e458\u0026ndash;523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e458\u0026ndash;523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGreen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e543\u0026ndash;578\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e650\u0026ndash;680\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed-Edge-1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e698\u0026ndash;713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed-Edge-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e733\u0026ndash;748\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed-Edge-3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e773\u0026ndash;793\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e785\u0026ndash;900\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB8a\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e885\u0026thinsp;\u0026minus;\u0026thinsp;875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater Vapour\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e935\u0026ndash;955\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCirrus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1360\u0026ndash;1390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSWIR-1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1565\u0026ndash;1655\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSWIR-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2100\u0026ndash;2280\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20\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\u003eAll bands except B1-B9-B10 were used in this study. However, the three atmospheric bands were not used in this study as they are mainly used for atmospheric corrections and cloud screening. (Drusch et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The fact that Sentinel-2 imagery contains a large number of spectral bands, and in particular has a temporal resolution of 5 days, provides a significant advantage for monitoring agricultural areas, detecting crop patterns and analyzing changes. (Şimşek and Durduran \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). As the study area is predominantly covered by crops with closely aligned phenological periods, 23 observations were made by Sentinel-2A between March and September 2016. After June, the selected images were cloud-free, while before June, images with a low cloud fraction were selected and cloudy areas were detected and masked. The acquisition dates of the Sentinel-2 satellite images are shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\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\u003eAcquisition time of Sentinel-2\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=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay of Year (DOY)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAcquisition Time\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDay of Year (DOY)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAcquisition Time\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 March 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e212\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e29 March 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 April 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e220\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5 April 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 May 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e230\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5 May 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20 May 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20 May 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e117\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 June 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e252\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6 June 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e132\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 July 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e262\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14 July 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e152\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 August 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e272\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11 August 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e167\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 September 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e282\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7 September 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e177\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 October 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e292\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2 October 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e192\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17 October 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17 October 2022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e202\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27 October 2022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e202\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27 October 2022\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\u003e2.3. Ground Truth Data\u003c/h2\u003e \u003cp\u003eTwo field studies were carried out in April for wheat, barley and lentil crops and in August for maize and cotton crops. Given the large area of the plain (225,000 ha), the number of samples could not reach the desired level due to time and cost constraints. In addition to field data, parcels declared by farmers were used as ground truth data in the study. 3145 agricultural parcels (cotton: 610, maize: 105, lentil: 340, wheat: 520, lentil_maize: 250, lentil_cotton: 380, wheat_maize: 413, wheat_cotton: 527) were used in the classification study.\u003c/p\u003e \u003cp\u003eThe Farmer Declaration Parcels (FDP) in Turkey, also known as the Farmer Registration System (FRS), is a government initiative aimed at registering and tracking agricultural activities and farmers in the country (Aydoğdu et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). This system was introduced to improve the efficiency and transparency of agricultural practices and to provide various benefits to registered farmers. The FRS requires farmers to register themselves and their farming activities. This registration process involves providing personal information and details of the land they farm. The parcels of agricultural activity registered by farmers in the system are called FDPs. In addition to the geometric information of the parcels, the system also contains information on the province, district, parcel number, agricultural parcel number, information on the products grown, area, surface, cadastral area, date of cultivation and date of harvest of each parcel. (Şimşek and Durduran \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). When the FDPs were examined, it was found that there were systematic and non-systematic differences between the geometry and attribute information of the parcels and the parcels in the field. Because of these differences, a number of processing steps were applied to the FDPs, and at the end of these steps, ground truth data were produced from the FDPs and used in the classification study together with the data collected in the field. The agricultural calendar for the crops grown in the plain was obtained and is presented graphically. (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e.). The calendar provides information on the sowing, growing, dense vegetation and harvesting periods of crops grown in the region. The exact dates of sowing and harvesting vary between plots and farmers. In some cases, there may be a difference of up to 1\u0026ndash;3 weeks between the categorized crops; wheat and barley crops are included in the wheat class.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eLooking at the phenological period of the crops, it is observed that maize and cotton are separated from each other with a dense vegetation stage, while wheat and lentil are in close phenological stages. It was observed that double crops (wheat_maize, lentil_maize, wheat_cotton, lentil_cotton) were in close phenological stages. The calendar shows that the phenological periods of both single and double crops are very close and different indices and bands should be used to distinguish these crops in the classification study.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Methodology","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1. General Architecture Overview\u003c/h2\u003e \u003cp\u003eAfter acquiring 23 Sentinel-2 images, the cloudy areas and the shadow areas caused by the clouds were masked. The bands were resampled from 20 m to 10 m. Four different spectral indices were calculated for each Sentinel-2 image, and spectral curves were created for each crop by overlapped these indices with ground truth data. The spectral index curves and spectral bands were then compared for each crop type. Classification studies using the RF algorithm were carried out with different datasets generated from different indices and bands, and the effect of bands and indices on the classification results was compared. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows the flowchart used in this study.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Satellite Image Processing\u003c/h2\u003e \u003cp\u003eIn data processing, no further geometric correction of the L1C products is required, only atmospheric correction and spatial resolution resampling. (Zheng et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Sentinel-2 images are provided in Level 1C format and contain above-atmospheric reflectance values. To calculate the actual reflectance values of plants in a classification study, top-of-atmosphere (TOA) values should be converted to bottom-of-atmosphere (BOA) reflectance values (Wilm \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Atmospheric effects were eliminated by converting TOA values to BOA values using the Sen2cor plugin. After Sen2Cor processing, the L1C TOA reflectance values were converted to Level 2A (L2A) BOA reflectance values (Wilm et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The conversion of Sentinel-2 data from L1C to L2A improves the accuracy of derived vegetation indices such as NDVI by accounting for atmospheric effects, ensuring that observed changes in NDVI and other indices are more closely related to actual variations in vegetation health.\u003c/p\u003e \u003cp\u003eClouds and cloud shadows in satellite imagery are the main sources of noise that cause problems in image analysis (Kalkan and Maktav \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Brightness caused by clouds and shadows can affect data analysis and lead to changes in NDVI and other indices. (Zhu and Woodcook 2012). Cloud-covered areas cause anomalies in the image bands as well as in the pixel values of the indices created from these bands, which adversely affects the classification results (Karslen et al. 2021). To eliminate this situation, clouds and shadow areas caused by clouds have been detected and masked by the Sen2core software. The cloud screening and classification part of Sen2Cor is available as source code in the distributed (Skakun et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Potentially cloudy pixels are subjected to a series of filters based on spectral band thresholds, ratios and index calculations - Normalized Difference Snow Index (NDSI), (NDVI). After atmospheric correction and cloud masking, B5-B7-B8a-B11-B12 were resampled from 20m to 10m.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Vegetation Indices and Spectral Band Analysis\u003c/h2\u003e \u003cp\u003eAs can be seen from the phenological calendar in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the single and double crops in the study area have very close phenological periods. The seasonal rhythm and phenological characteristics of different crops can be reflected by the difference in the spectrum or vegetation index of multi-temporal remote sensing data \u003cb\u003e(\u003c/b\u003eGomez et al. 2016). NDVI is widely used for crop type identification \u003cb\u003e(\u003c/b\u003eGumma et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The NDVI index provides information on plant health and development \u003cb\u003e(\u003c/b\u003eMorsy and Hadı \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In order to see this phenological closeness on a crop by crop basis, time series NDVI plots were first created. Each plot was overlapped with multi-temporal NDVI images and the median NDVI values of each plot in the time series were calculated. After this process, the vegetative development and change of each plot was determined and the characteristics of the NDVI curves showing the variability over time were revealed. The characteristic NDVI curves for each crop were checked against the phenological calendar and the spectral reflectance values of each crop collected during the field study. When analysing the NDVI time series curves of each crop, it can be seen that lentils and wheat, which are winter crops, are separated from each other, while maize and cotton, which are summer crops, show new different time curves (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e). With the NDVI index, these four crops (lentil, wheat, maize, cotton) have different reflectance values on the same dates and their temporal curves are separated from each other. Therefore, no studies have been carried out with different vegetation indices in single crops.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWhen analyzing the NDVI time series curves, wheat_maize-wheat_cotton crops and lentil_maize-lentil_cotton crops differ from each other between day 57. and day 132.However due to the later sowing of lentil, but it can be seen that the reflectance values for all crop types are very close to each other between day 220. and day 317. Figure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIt was observed that the NDVI time series curves were similar for both crops, especially in the second period of the year. The effect of different vegetation indices on the time series curves of the crops and the classification results were investigated in this study. There are nearly 100 vegetation indices in the literature. In this study, NDVI, NDRE, SAVI and NDRE indices, which are traditional vegetation indices, were used Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\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\u003eAcquisition time of Sentinel-2\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpectral Index\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFull Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFormula\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eReference\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNDVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNormalized Difference\u003c/p\u003e \u003cp\u003eVegetation Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(NIR - R)/(NIR\u0026thinsp;+\u0026thinsp;R)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCommonly used for assessing overall vegetation health and density\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTucker \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1979\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEnhanced vegetation index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.5*(NIR - R)/(NIR\u0026thinsp;+\u0026thinsp;6*R-7.5*B\u0026thinsp;+\u0026thinsp;1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSimilar to NDVI but designed to minimize atmospheric influences and improve sensitivity in high biomass regions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHuete et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2002\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSoil-Adjusted\u003c/p\u003e \u003cp\u003eVegetation Index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(NIR - R)/(NIR\u0026thinsp;+\u0026thinsp;R\u0026thinsp;+\u0026thinsp;L)*(1\u0026thinsp;+\u0026thinsp;L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDesigned to minimize soil background influences in the vegetation index\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHuete \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1988\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNDRE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNormalized\u003c/p\u003e \u003cp\u003eDifference Red Edge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(NIR \u0026ndash; Redge2)/(NIR\u0026thinsp;+\u0026thinsp;Redge2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eParticularly sensitive to changes in chlorophyll content, making it valuable for detecting early signs of stress\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eBarnes 2000\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\u003eComparing the NDVI, SAVI and NDRE time series curves of lentil_maize and lentil_cotton, it is observed that the curves are close to each other as in the case of NDVI, this is also the case for wheat_maize and wheat_cotton. It was observed that the SAVI index did not change in relation to the NDVI index. As a result of the EVI index, the time series curves for four crops show some separation compared to the SAVI and NDVI indices. When analyzing the time series curves of the NDRE index, it can be seen that the curves for the two crops wheat_maize-wheat_cotton and lentil_maize-lentil_cotton are more separated compared to the other indices Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. In summary, the NDRE index values performed better than other indices in separating the time series curves of crops.\u003c/p\u003e \u003cp\u003eIn addition to observing the changes in the time series curves of crops with different vegetation indices, the study also were analysed the time series reflectance values of each spectral band for all crop types. Comparing the time series reflectance values of each spectral band for all crop types, it is seen that the B6 (Red Edge-2) and B11 (SWIR-1) bands are separated from the other bands Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDifferent datasets were generated using 4 different vegetation indices and 10 spectral bands Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. Classification studies were performed on these datasets and the results were compared. It was also investigated if the indices and bands that create time series curve differences in the generated datasets affect the classification result.\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\u003eData sets and number of bands\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=\"char\" char=\".\" 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=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eData Set\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFeatures\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNumber of bands\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eData Set\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFeatures\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eNumber of bands\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-1)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNDVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-9)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4BAND)-EVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-2)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNDRE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-10)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4 BAND)-B6-B11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e126\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-3)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSAVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-11)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(4 BAND)-B6-B11-NDRE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e147\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-4)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-12)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eB2-B3-B4-B5-B6-B7-B8-\u003c/p\u003e \u003cp\u003eB8A-B11-B12 (10 BAND)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e210\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-5)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eB2-B3-B4-B8\u003c/p\u003e \u003cp\u003e(4BAND)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-13)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10 BAND)-NDVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e231\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-6)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(4BAND)-NDVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-14)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10 BAND)-NDRE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e231\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-7)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(4BAND)-SAVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-15)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10 BAND)-SAVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e231\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-8)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(4BAND)-NDRE\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-16)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(10 BAND)-EVI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e231\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=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Classification Using Random Forest\u003c/h2\u003e \u003cp\u003eRandom Forest (RF) is a combinatorial ensemble learning classification technique. RF is an improved algorithm based on an ensemble learning technique that builds multiple CARTs. (Breiman \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2001\u003c/span\u003e) In fact, RF has been very successful as a general purpose classification and regression method (Biau and Scornet \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). RF fits many classification trees to training data sets and then combines the predictions of all the trees to make a final decision. RF is an ensemble classifier that is currently widely used in remote sensing studies due to its classification accuracy (Belgiu and Dragut \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Higher accuracies have been achieved with RF compared to other machine learning algorithms in many crop mapping studies (Tatsumi et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). RF is known to work efficiently on large datasets with a large number of input variables to estimate which variables are significant in the classification process, and is relatively robust to noise and outliers. 44 Many examples of the use of this algorithm can be found in the literature (Feng et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHyperparameter optimisation is the process of finding the optimal combination of parameters for a machine learning algorithm according to specified success criteria. Hyperparameter optimisation aims to achieve a balance between overlearning and under learning by balancing high model success and model complexity. The original RF has two hyperparameters including the number of trees (ntree) and the number of variables used to partition the nodes (mtry). Several studies have shown that satisfactory results can be achieved with the default parameters (Zhang and Roy \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In this study, to find the optimal RF model for classification, a range of values for both parameters were tested and evaluated using the grid search method for each dataset: ntree\u0026thinsp;=\u0026thinsp;100, 200, 400, 600,800,1000; mtry\u0026thinsp;=\u0026thinsp;1:20.\u003c/p\u003e \u003cp\u003eAfter determining the optimal parameters of the algorithm to be used in the classification process, the k-fold cross-validation method is used. K-fold cross-validation allows to see whether the high performance of the model is random or not. In this method, the data set is divided into k parts and k-1 subset is used to train the model and the remaining subset is used to calculate the accuracy of the model. The process is repeated k times, each time using different pieces of training and test data. The average of the accuracy values obtained represents the accuracy of the model, and in this study the k value is taken as 5 (Kohavi \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1995\u003c/span\u003e).\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\u003eHyper parameter values for different data sets\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=\"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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eData Set\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of \u003c/p\u003e \u003cp\u003ebands\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNtree\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMtry\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS: 1\u0026ndash;4)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-5)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS: 6\u0026ndash;9)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-10)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-11)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e147\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS-12)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e210\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eData Set (DS: 13\u0026ndash;16)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e231\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15\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"},{"header":"4. Results and Discussion","content":"\u003cp\u003eIn order to assess the accuracy of classification performance, there are many metrics available in the literature. In this work, PA, UA, OA and F1-score have been used to assessment class accuracy Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003ePA (Producer Accuracy): The ratio of correctly predicted positive observations to the total predicted positives. It measures the model\u0026apos;s ability to correctly identify positive instances among the instances it predicted as positive.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eUA (User Accuracy): The ratio of correctly predicted positive observations to the total actual positives. It measures the model\u0026apos;s ability to correctly identify positive instances.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eOA (Overall Accuray): The ratio of correctly predicted instances (both positive and negative) to the total instances. It provides a general measure of the model\u0026apos;s correctness.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eF1 Score: The harmonic mean of precision and recall. It provides a balance between precision and recall, especially useful when dealing with imbalanced datasets.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eThese classification accuracy indicators can reflect the overall classification accuracy and specific type identification accuracy of remote sensing images from different aspects (Congatol, 1991).\u003c/p\u003e\n\u003cp\u003eTable 7. Accuracy assessment for all data sets\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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°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\"\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eFor wheat and lentil crops it is observed that the accuracy is 90% and above in 16 datasets. While the lowest accuracy value belongs to DS-3 dataset, the highest accuracy values belong to DS-7 and DS-14 datasets. It was observed that the classification results with DS-5 and DS-10 datasets were not different. It was observed that wheat and lentil products, which have different spectral separation curves, achieved high accuracy as a result of classification with each dataset. For maize and cotton crops, it can be seen that the accuracy is 90% and above in 16 datasets. While the lowest accuracy value for cotton is DS-5, the highest accuracy value belongs to DS-14. While the lowest accuracy value for maize is DS-3 dataset, the highest accuracy value belongs to DS-14 dataset. Since the time series curves of maize and cotton crops are different from each other, like wheat and lentil crops, the accuracy values are high in each dataset. For these classes, it was found that increasing the number of bands did not increase the accuracy too much.\u003c/p\u003e\n\u003cp\u003eThe DS-5 dataset (81\u0026ndash;82%) gives the lowest accuracy value for all double crops and the DS-14 dataset (94.8%) gives the highest accuracy value. It can be observed that the DS-2 dataset with NDRE values gives a higher accuracy compared to the datasets with other indices (DS-1-3-4). This high accuracy value is due to the NDRE index, which performs better than other indices in separating the time series curves of crops. In the DS-5 (81\u0026ndash;82%) dataset, double crops remained with low accuracy, in the DS-12 (84\u0026ndash;85%) dataset, double crops remained with low accuracy and in the DS-12 (84\u0026ndash;85%) dataset, the accuracy increased by about 3% but remained insufficient Fig. \u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003eWhen evaluating Fig. \u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003e, the lowest overall accuracy value belongs to DS-5 (86.4%) and DS-3 (87.02%) datasets. The highest accuracy value belongs to DS-14 (94.77%), as in the product base. Comparing the DS-7 (91.24%) and DS-14 (94.77%) datasets, there is an increase of approximately 3%. Comparing DS-7 (91.24%) and DS-11 (93.39%) data sets, band 6 and band 11, which are separated from other bands in all products, increase the accuracy value by 2%. When comparing the DS-11 (93.39%) and DS-14 (94.77%) data sets, it is also seen that the results are close to each other. It can be seen that the datasets containing the NDRE index (DS-2-7-11-14) and the datasets containing 10 bands (DS-12-13-14-15-16) have an accuracy value above 90%. The comparison of the DS-5 (81\u0026ndash;82%) dataset with the DS-2 (89\u0026ndash;90%) dataset shows that the accuracy of double crops increased by about 8% and the NDRE index improved the classification result of crops with close phenological stages. When DS-2 was compared with the DS-7 dataset, it was observed that the +\u0026thinsp;4 band in the time series NDRE dataset did not have much effect on the classification result both for double crops and for the overall accuracy. Comparing DS-7 (91%) with DS-11 (92%), it can be seen that band 6 and band 11 slightly increase the classification accuracy of double crops. For crops with similar time series, data sets with only 4 bands (DS-5) and only 10 bands (DS-12) did not reach a high accuracy for the classification of crops with close phenological stages.\u003c/p\u003e\n\u003cp\u003eIt is observed that the data sets containing the NDRE index, which performs better than other indices in separating the time series curves of crops, increase the overall accuracy by 2\u0026ndash;3% compared to other data sets, and the accuracy increases by 2% with the addition of the 6th and 11th bands to these data set. The most important factor in increasing this accuracy is Band 6 (Red Edge-2), which is also used in the NDRE index.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study investigated the effect of vegetation indices and spectral bands on the classification of agricultural crops. Time series curves of classes with different vegetation indices were generated and compared, and time series reflectance values of each spectral band were also observed for all crops. The study shows the importance of time series curves (phenological period) generated from multi-temporal images of crops. In the time series curves generated with different vegetation indices, it was found that the crops that differ (with a unique curve) have a high accuracy value with optimal data sets, while increasing the number of features and bands in the input data set has almost no effect on the accuracy value of these crops. For crops with very close phenological periods, indices that reveal the difference between the time series curves (NDRE-EVI) were found to increase the classification result. When comparing the time series reflectance values of each spectral band for all crop types, except the crop indices, it was observed that the B6 (Red Edge-2) and B11 (SWIR-1) bands were separated from the other bands and increased the classification result by 2% when included in the dataset. This result showed that B6 (Red Edge-2) and B11 (SWIR-1) bands should be used in agricultural crop type classification studies, especially for crops covering close phenological stages. In particular, for classification studies carried out in large working areas such as this one, by determining the bands with indices that separate the time series curves of the crops, the bands and indices that will give the highest accuracy as a result of the classification can be determined and time can be saved by obtaining maximum accuracy with minimum data in classification studies. Future studies will extend the scope of the study by using different algorithms for different crop indices, different crop types and different study areas.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003ea) Fatih Fehmi ŞİMŞEKThe article is the work of a single author, and the content of the article is entirely written by him.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAydoğdu M, Ak\u0026ccedil;ar HT, \u0026Ccedil;ullu MA (2011) Coğrafi bilgi sistemleri CBS ve uzaktan algılama UA kullanılarak \u0026ccedil;ift\u0026ccedil;i kayıt sistemi \u0026ccedil;ks verilerinin analizi ile pamuk ve mısır primlerinin \u0026ouml;denmesi Şanlıurfa-Harran İl\u0026ccedil;esi \u0026ouml;rneği. Jeodezi ve Jeoinformasyon Dergisi 104(\u0026ndash;S):47\u0026ndash;52\u003c/li\u003e\n\u003cli\u003eBarnes EM, Clarke TR, Richards SE, Colaizzi PD, Haberland J, Kostrzewski M, Waller P, Choi C, Riley E, Thompson T, Lascano RJ, Li H, Moran MS (2000) Coincident detection of crop water stress, nitrogen status, and canopy density using ground-based multispectral data. 5th International Conference on Precision Agriculture, Bloomington, 16\u0026ndash;19, 1\u0026ndash;15\u003c/li\u003e\n\u003cli\u003eBelgiu M, Dragut 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/li\u003e\n\u003cli\u003eBiau G, Scornet E (2016) A random forest guided tour. Test 25:197\u0026ndash;227. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11749-016-0481-7\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eBozkurt M, Aybek A (2016) Şanlıurfa İli Harran Ovasının tarımsal yapı ve mekanizasyon \u0026ouml;zellikleri, KS\u0026Uuml; Doğa Bilimleri Dergisi 19(3),2016, pp. 319\u0026ndash;331\u003c/li\u003e\n\u003cli\u003eBreiman L (2001) Random Forests. Mach Learn 45:5\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1023/A:1010933404324\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eConese C, Maselli F (1991) Use of multi-temporal information to improve classification performance of TM scenes in complex terrain. ISPRS J photogrammetry Remote Sens 46(4):187\u0026ndash;197. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0924-2716(91)90052-W\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eCongalton RG (1991) A review of assessing the accuracy of classifications of remotely sensed data. Remote Sens 37(1):35\u0026ndash;46. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0034-4257(91)90048-B\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eDrusch M, Bello UD, Carlier S, Colin O, Fernandez V, Gascon F, Hoersch B, Isola C, Laberinti P, Martimort P, Meygret A, Spoto F, Sy O, Marchese F, Bargellini PL (2012) Sentinel-2: ESA's optical high-resolution mission for GMES operational services. Remote Sens Environ 120:25\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2011.11.026\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eFeng S, Zhao J, Liu T, Zhang H, Zhang Z, Guo X (2019) Crop type identification and mapping using machine learning algorithms and Sentinel-2 time series data. IEEE J Sel Top Appl Earth Observations Remote Sens 12:3295\u0026ndash;3306. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/JSTARS.2019.2922469\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eG\u0026oacute;mez C, White JC, Wulder MA (2016) Optical remotely sensed time series data for land cover classification: A review. ISPRS J Photogrammetry Remote Sens 116:55\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2016.03.008\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eGumma MK, Tummala K, Dixit S, Collivignarelli F, Holecz F, Kolli RN (2020) Whitbread AM. Crop type identification and spatial mapping using Sentinel-2 satellite data with focus on field-level information. Geocarto Int 37(11):1\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10106049.2020.1805029\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eHuete A, Didan K, Miura T, Rodriguez EP, Gao X, Ferreira LG (2002) Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens Environ 83:195\u0026ndash;213. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0034-4257(02)00096-2\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eHuete AR (1988) A soil adjusted vegetation index (SAVI). Remote Sens Environ 25:295\u0026ndash;309. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/0034-4257(88)90106-X\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKalkan K, Maktav D (2016) Landsat-8 g\u0026ouml;r\u0026uuml;nt\u0026uuml;lerinden g\u0026ouml;lge ve bulut belirleme. VI. Uzaktan Algılama ve Coğrafi. Bilgi Sistemleri Sempozyumu, Adana, T\u0026uuml;rkiye\u003c/li\u003e\n\u003cli\u003eKang Y, Meng Q, Liu M, Zou Y, Wang X (2021) Crop classification based on red edge features analysis of GF-6 WFV data. Sensors 21(13). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s21134328\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKarlsen SR, Stendardi L, T\u0026oslash;mmervik H, Nilsen L, Arntzen I, Cooper EJ (2021) Time-series of cloud-free Sentinel-2 NDVI data used in mapping the onset of growth of Central Spitsbergen. Svalbard Remote Sens 13(15):3031. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs13153031\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKavzoglu T (2009) Increasing the accuracy of neural network classification using refined training data. Environ Model Softw 24(7):850\u0026ndash;858. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.envsoft.2008.11.012\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKobayashi N, Tani H, Wang X, Sonobe R (2020) Crop classification using spectral indices derived from Sentinel-2A imagery. J Inform Telecommunication 4(1):67\u0026ndash;90. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/24751839.2019.1694765\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKhatami R, Mountrakis G, Stehman SV (2016) A meta-analysis of remote sensing research on supervised pixel-based land-cover image classification processes: General guidelines for practitioners and future research. Remote Sens Environ 177:89\u0026ndash;100. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2016.02.028\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKim HO, Yeom JM (2015) Sensitivity of vegetation indices to spatial degradation of Rapid Eye imagery for paddy rice detection: a case study of South Korea. GIScience Remote Sens 52(1):1\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/15481603.2014.1001666\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eKohavi R (1995) A Study of cross-validation and bootstrap for accuracy estimation and model selection. Proceedings of the 14th international joint conference on Artificial intelligence 2, 1137\u0026ndash;1143\u003c/li\u003e\n\u003cli\u003eLu D, Weng Q (2007) A survey of image classification methods and techniques for improving classification performance. Int J Remote Sens 28:823\u0026ndash;870\u0026nbsp;\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01431160600746456\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eMorsy S, Hadı M (2022) Impact of land use/land cover on land surface temperature and its relationship with spectral indices in Dakahlia Governorate, Egypt. Int J Eng Geosci 7(3):272\u0026ndash;282. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26833/ijeg.978961\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eMurthy CS, Raju PV, Badrinath KV (2003) Classification of wheat crop with multi-temporal images: performance of maximum likelihood and artificial neural networks. Int J Remote Sens 24:4871\u0026ndash;4890. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0143116031000070490\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003ePasternak M, Filipiak KP (2022) The evaluation of spectral vegetation indexes and redundancy reduction on the accuracy of crop type detection. Appl Sci 12(10):5067. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/app12105067\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003ePelletier C, Valeroa S, Ingladaa C, Championb N, Dedieua G (2016) Assessing the robustness of random forests to map land cover with high resolution satellite image time series over large areas. Remote Sens Environ 187:156\u0026ndash;168. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/j.rse.2016.10.010\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eSkakun S, Wevers J, Brockmann C, Doxani G, Aleksandrov M, Batič M, Frantz D, Gascon F, Chova LG, Hagolle O, Puigdollers OL, Louis J, Lubej M, Garc\u0026iacute;a GM, Osman J, Peressutti D, Pflug B, Puc J, Richter R, Roger JC, Scaramuzza P, Vermote E, Vesel N, Zupanc A, Žust L (2022) Cloud mask intercomparison exercise (CMIX): An evaluation of cloud masking algorithms for Landsat 8 and Sentinel-2. Remote Sens Environ 274:112990. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2022.112990\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eSun C, Bian Y, Zhou T, Pan J (2019) Using of multi-source and multi-temporal remote sensing data improves crop-type mapping in the subtropical agriculture region. Sensors 19(10):2401. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s19102401\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eStern AJ, Daughtry CST, Hunt ERJr, Gao F (2023) Comparison of five spectral indices and six imagery classification techniques for assessment of crop residue cover using four years of Landsat imagery. Remote Sens 15:4596. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs15184596\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eSong D, Huang C, Sexton JO, Channan S, Feng M, Townshend JR (2015) Case studies from the Eastern United States and Central Brazil. ISPRS J Photogrammetry Remote Sens 103:81\u0026ndash;92. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2014.09.005\u003c/span\u003e\u003c/span\u003e. Use of Landsat and corona data for mapping forest cover change from the mid-1960s to 2000s\u003c/li\u003e\n\u003cli\u003eSonobe R, Yamaya Y, Tani H, Wang X, Kobayashi N, Mochizuki K (2018) Crop classification from Sentinel-2 derived vegetation indices using ensemble learning. J Appl Remote Sens 12(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1117/1.JRS.12.026019\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eŞimşek FF, Durduran SS (2023) A\u0026ccedil;ık kaynak kodlu Eo-learn k\u0026uuml;t\u0026uuml;phanesi ve \u0026ccedil;ok zamanlı Sentinel-2 g\u0026ouml;r\u0026uuml;nt\u0026uuml;leri ile tarımsal \u0026uuml;r\u0026uuml;n sınıflandırması. Jeodezi ve Jeoinformasyon Dergisi 10(1):45\u0026ndash;62. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.9733/JGG.2023R0004.T\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eTatsumi M, Yamashiki Y, Torres MAC, Taipe CLR (2015) Crop classification of upland fields using random forest of time-series Landsat 7 ETM\u0026thinsp;+\u0026thinsp;data. Comput Electron Agric 115:171\u0026ndash;179. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compag.2015.05.001\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eTucker CJ (1979) Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens Environ 8:127\u0026ndash;150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/0034-4257(79)90013-0\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eVuolo F, Neuwirth M, Immitzer M, Atzberger C, Ng W (2018) How much does multi-temporal Sentinel-2 data improve crop type classification? Int J Appl Earth Obs Geoinf 72:122\u0026ndash;130. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jag.2018.06.007\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eWilm UM (2017) Sen2Cor configuration and user manual. 2017, 9\u0026ndash;12\u003c/li\u003e\n\u003cli\u003eWilm UM, Louis J, Richter R, Gascon F, Niezette M (2013) Sentinel-2 Level-2A prototype processor: Architecture, algorithms and first results. ESA, Living Planet Symposium ESA-SP-722\u003c/li\u003e\n\u003cli\u003eZhang H, Kang J, Xu X, Zhang L (2020) Accessing the temporal and spectral features in crop type mapping using multi-temporal Sentinel-2 imagery: A case study of Yi'an County, Heilongjiang province, China. Comput Electron Agric 176:105618. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.compag.2020.105618\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eZhang HK, Roy DP (2017) Using the 500 m MODIS land cover product to derive a consistent continental scale 30 m Landsat land cover classification. Remote Sens Environ 197:15\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2017.05.024\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eZheng H, Du P, Chen J, Xia J, Li E, Xu Z, Li X, Yokoya N (2007) Performance evaluation of downscaling sentinel-2 imagery for land use and land cover classification by spectral-spatial features. Remote Sens 9(12):1274. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs9121274\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eZhong L, Hu L, Zhou H (2019) Deep learning based multi-temporal crop classification. Remote Sens Environ 221:430\u0026ndash;443. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2018.11.032\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n\u003cli\u003eZhu Z, Woodcock CE (2012) Object-based cloud and cloud shadow detection in Landsat imagery. Remote Sens Environ 118:83\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2011.10.028\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":true,"highlight":"","institution":"","isAcceptedByJournal":false,"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":"crop, classification, random forest, sentinel, spectral band, vegetation index","lastPublishedDoi":"10.21203/rs.3.rs-3910868/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3910868/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSatellite imagery and remote sensing technology allow the identification, observation and assessment of dynamic agricultural areas. Image classification is one of the most widely used methods to determine the pattern of agricultural crops. The accuracy of the agricultural crops to be classified depends on many parameters such as the classification method used, satellite image resolution, number of images used, bands, indices and training data. In this study, a classification study was carried out using multi-temporal Sentinel-2 imagery and datasets generated from different vegetation and spectral indices, and the effects on the classification result were investigated. As the study area has very fertile soils, suitable climate and temperature conditions and irrigated land, it is possible to grow more than one crop on the same plot during a production season. Wheat_maize (winter_wheat\u0026thinsp;+\u0026thinsp;summer_maize), wheat_cotton (winter_wheat\u0026thinsp;+\u0026thinsp;summer_cotton), lentil_cotton (winter_lentil\u0026thinsp;+\u0026thinsp;summer_cotton), lentil_maize (winter_lentil\u0026thinsp;+\u0026thinsp;summer_maize) are the crops included in the classification study, except for single crops; maize, cotton, wheat and lentils are also included. Time series of vegetation indices can be used to capture information on plant phenology and can be used as reference information in crop classification. Time series curves of different vegetation indices were constructed and compared for all crops, especially for double crops with the same phenological periods. In addition to the vegetation indices, the variation of the time series reflectance values of each spectral band was also observed for all crops and the effect of different indices and bands on the classification result was investigated. The study generated 16 different data sets using conventional vegetation indices, NDVI, SAVI, EVI and NDRE vegetation indices and all other bands of the Sentinel-2 satellite except the 60m bands. While single crops with different time series (maize, cotton, lentil, wheat) had an accuracy of over 90% in each dataset, double crops could not exceed 81% accuracy by mixing with each other in the DS-5 (R-G-B-NIR) dataset. In the DS-1 (NDVI time series) dataset, the overall accuracy for double crops is in the range of 84\u0026ndash;85%. Classification with DS-2 (NDRE time series) increased the overall accuracy for double crops to 90%. When comparing the time series reflectance values of each spectral band for all crop types, except the crop indices, it was observed that the B6 (Red Edge-2) and B11 (SWIR-1) bands were separated from the other bands and increased the classification result by 2% when included in the dataset. Especially in the classification studies carried out on products with close phenological periods, the Red Edge band (especially Red Edge-2) and the indices (NDRE) generated from these bands will improve the classification result by preventing confusion between classes, and the B11 (SWIR-1) band also has a positive effect on classification. This study has fully demonstrated the application potential of red edge bands and the indices constructed from them. It also promotes the use of red edge band optical satellite data in agricultural remote sensing.\u003c/p\u003e","manuscriptTitle":"Investigation of the Effect of Spectral Bands and Vegetation Index Selection on Agricultural Crop Type Classification (Especially for Double Crops)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-01 17:05:14","doi":"10.21203/rs.3.rs-3910868/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":"ef8ae144-107f-4bc7-9bb0-87524efa9f57","owner":[],"postedDate":"February 1st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-02-01T17:05:17+00:00","versionOfRecord":[],"versionCreatedAt":"2024-02-01 17:05:14","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3910868","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3910868","identity":"rs-3910868","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}