Machine learning approach for Forest Biomass Modelling with In-Situ and Remote Sensing Data in Narmadapuram central India

Machine learning approach for Forest Biomass Modelling with In-Situ and Remote Sensing Data in Narmadapuram central India | 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 Machine learning approach for Forest Biomass Modelling with In-Situ and Remote Sensing Data in Narmadapuram central India Pradeep Kumar Rajput This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6270236/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 19 Jul, 2025 Read the published version in Modeling Earth Systems and Environment → Version 1 posted 14 You are reading this latest preprint version Abstract The study estimation of forest Biomass using In-Situ and Remote Sensing data presents a comprehensive investigation into the estimation of forest biomass, a pivotal component of forest ecosystems and a key parameter in understanding carbon dynamics. This research merges in-situ field measurements with cutting-edge remote sensing technologies to develop robust and accurate models for predicting forest biomass. The research leverages data acquired from ground-based measurements, including tree diameter, height, and species composition, in tandem with remote sensing data obtained from satellite platforms. Various modelling techniques, including machine learning algorithms and statistical analyses, are applied to establish the relationship between these datasets and forest biomass. The study evaluates the performance of multiple methods, such as Exponential Regression, Linear Regression, Random Forest, and Support Vector Machines (SVM). The results indicate that Random Forest outperformed other methods with an RMSE of 1.61, MAE of 0.84, relRMSE of 0.1046609, and r² of 0.51. In comparison, Exponential Regression achieved an RMSE of 2.26, MAE of 0.97, relRMSE of 0.1471322, and r² of 0.04, Linear Regression produced an RMSE of 2.48, MAE of 1.34, relRMSE of 0.1616262, and r² of -0.16; while SVM recorded an RMSE of 2.00, MAE of 1.06, relRMSE of 0.1301456, and r² of 0.25. The outcomes of this study hold significant implications for forest management, climate change mitigation, and conservation efforts. Accurate forest biomass estimates are crucial for assessing carbon storage, understanding ecosystem health, and designing sustainable forestry practices. Moreover, by integrating in-situ and remote sensing data, this research contributes to the ongoing global efforts to monitor and protect the world's forests in an era of environmental challenges. The findings of this study provide valuable insights for policymakers, environmentalists, and researchers engaged in forestry, ecology, and climate change studies, facilitating more informed decisions and sustainable practices in forest management and conservation. Forest Biomass Allometric conservation ecological modelling ecosystem services Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Introduction The estimation of forest biomass is critical for understanding carbon sequestration, biodiversity, and ecosystem health. Accurate and efficient methods for quantifying forest biomass have been a focal point of research, as these estimates are essential for informed forest management and climate change mitigation efforts (Radtke et al. 2016). This review explores the extensive body of literature related to forest biomass modeling using a combination of in-situ and remote sensing data (Mohren et al. 2012). We delve into the historical development, methodologies, challenges, and recent advancements in this field, highlighting key studies that have contributed to our understanding of forest biomass estimation (Lu et al. 2014). The study of forest biomass estimation dates back several decades, but it has gained increasing importance in the context of climate change mitigation (Brown 1997). Early methods primarily relied on ground-based measurements and allometric equations that related tree variables to biomass. Notable work introduced a widely-used generalized allometric equation for estimating aboveground biomass in forests (Radtke et al. 2016). However, the limited spatial coverage of ground-based data prompted the integration of remote sensing techniques. Remote sensing technologies have revolutionized the field of forest biomass estimation (Kumar et al. 2015). The utilization of satellite and airborne sensors has allowed for the collection of large-scale forest data, making it possible to assess biomass across broad geographic regions (Lu et al. 2014). Demonstrated the potential of Synthetic Aperture Radar (SAR) data for mapping forest biomass (Minh et al. 2016). Their work laid the foundation for subsequent research that leveraged SAR imagery, which is less sensitive to cloud cover and suitable for tropical forests (Ranson et al. 1997). In-situ data, including forest inventory plots and field measurements, remain indispensable for calibrating and validating remote sensing-based biomass models (Sinha et al. 2015). emphasized the importance of plot-based data in estimating forest biomass, highlighting the variability of biomass estimates within different forest types (Schepaschenko et al. 2019). Ground data also aid in addressing issues related to sensor-specific biases, which are critical in refining remote sensing models. Several techniques have emerged for combining in-situ and remote sensing data to estimate forest biomass (Lu 2006). Among these, the integration of optical and LiDAR (Light Detection and Ranging) data has gained considerable attention (Pizaña et al. 2016). Demonstrated the utility of combining optical data from the Landsat satellite series with airborne LiDAR to estimate biomass across large areas (Brovkina et al. 2016). The synergy of these two data sources provides both spatial and vertical information, resulting in more accurate biomass estimates. Despite significant advancements, challenges persist in forest biomass modeling (Sun et al. 2011). Among the primary issues are the spatial and temporal resolution of remote sensing data, which can limit the precision of biomass estimates, especially in areas with rapid land cover change (Knott et al. 2023). Additionally, the accuracy of in-situ data collection is influenced by sampling design, and efforts are ongoing to establish standardized protocols for forest inventory plots (Zolkos et al. 2012). Moreover, model transferability across different forest types and regions remains a challenge, as models trained in one area may not perform well in another due to variations in tree species and environmental conditions (Kilham et al. 2018). Recent research has focused on machine learning approaches, such as Random Forest and Neural Networks, for improved biomass estimation (Gao et al. 2018). used a combination of field data, LiDAR, and satellite data to create a machine learning-based biomass map, achieving higher accuracy compared to traditional methods (Baccini et al. 2004). Furthermore, the integration of multi-temporal remote sensing data has enabled the monitoring of biomass change over time, contributing to our understanding of forest dynamics and carbon fluxes (Tian et al. 2023). Forest biomass estimation is directly relevant to climate change mitigation strategies. Accurate biomass assessments underpin the calculation of carbon stocks in forests, which is essential for carbon offset programs and international agreements like REDD+ (Reducing Emissions from Deforestation and Forest Degradation)(Mo et al. 2023). Research by Researcher demonstrated how remote sensing can be used to estimate carbon stocks in tropical forests, emphasizing the role of accurate biomass data in supporting climate change initiatives(Saatchi et al. 2007). In summary, forest biomass modeling using a combination of in-situ and remote sensing data has witnessed significant developments over the years. The integration of optical, LiDAR, and SAR data, along with machine learning techniques, has improved the accuracy and applicability of biomass estimates (Kumar et al. 2015). Nevertheless, challenges such as data resolution, model transferability, and the need for standardized in-situ data collection persist. As efforts to combat climate change intensify, the accurate quantification of forest biomass remains critical, and further research will likely refine existing techniques and develop new methods for better understanding forest ecosystems and their role in global carbon dynamics. Forest ecosystems play a pivotal role in maintaining ecological balance, serving as critical carbon sinks, and mitigating the impacts of climate change. Forest biomass, which refers to the total mass of living biological organisms within a given forest area, is a fundamental metric for understanding forest productivity, carbon sequestration, and overall ecosystem health (Gleason and Im 2011). Accurate estimation of forest biomass is essential for a variety of applications, including forest management, climate modeling, and carbon accounting under international frameworks such as the United Nations Framework Convention on Climate Change (UNFCCC) (Sundquist et al. 2016). In recent years, the integration of in-situ measurements and remote sensing technologies has revolutionized the field of forest biomass estimation. Traditional methods of biomass estimation, which rely heavily on destructive sampling and allometric equations, are often labour-intensive, time-consuming, and spatially limited (Abbas et al. 2020). Remote sensing, on the other hand, provides a means to estimate biomass over large spatial extents with high temporal and spatial resolution, making it an indispensable tool in forest research and management (Huang et al. 2015). The combination of in-situ and remote sensing data addresses the limitations of each approach and enhances the accuracy of biomass modeling. In-situ data, such as tree diameter, height, and wood density, provide ground truth measurements that are critical for calibrating and validating remote sensing models (Pizaña et al. 2016). Remote sensing platforms, including satellite imagery, airborne LiDAR, and radar sensors, offer the ability to capture spatially explicit information on forest structure and composition, which can be correlated with biomass metrics (Næsset et al. 2011). This introduction explores the significance of forest biomass modeling, the advances in remote sensing technologies, and the synergistic potential of combining in-situ and remote sensing data. It also highlights the key challenges and research gaps in the field, paving the way for more accurate and scalable biomass estimation methodologies (Song et al. 2023). Forest biomass estimation is a cornerstone of global efforts to combat climate change. Forests act as carbon reservoirs, sequestering carbon dioxide from the atmosphere and storing it in vegetation and soil (Duncanson et al. 2019). The quantification of biomass is crucial for assessing carbon stocks, understanding forest dynamics, and evaluating the impact of land-use changes on carbon emissions (Shi and Liu 2017). Accurate biomass estimates are also essential for monitoring deforestation and forest degradation, which account for approximately 10–15% of global greenhouse gas emissions (Mohren et al. 2012). In the context of climate change mitigation, forest biomass data are integral to the implementation of Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiatives (Griscom et al. 2017). These programs incentivize developing countries to conserve forests by providing financial rewards for verified reductions in carbon emissions. However, the success of such programs depends on the availability of reliable and consistent biomass data, which underscores the importance of robust modeling techniques (Haya et al. 2023). Remote sensing technologies have advanced significantly over the past few decades, offering unprecedented capabilities for forest biomass estimation. Optical remote sensing platforms, such as Landsat and Sentinel-2, provide high-resolution imagery that can be used to derive vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), which are proxies for biomass (Eisfelder et al. 2011). Radar sensors, such as those on the Sentinel-1 and ALOS PALSAR satellites, are particularly effective in capturing forest structure and canopy characteristics due to their ability to penetrate vegetation (Lu 2006). LiDAR (Light Detection and Ranging) technology represents a breakthrough in forest biomass modeling, offering detailed three-dimensional information on forest structure (White et al. 2017). Airborne and terrestrial LiDAR systems can measure tree height, canopy density, and vertical forest structure with high precision, making them invaluable for biomass estimation (Simpson et al. 2017). Additionally, the launch of space borne LiDAR missions, such as the Global Ecosystem Dynamics Investigation (GEDI), has expanded the reach of this technology to a global scale (Kellner et al. 2019).The integration of in-situ and remote sensing data represents a paradigm shift in forest biomass modeling. Ground-based measurements provide accurate and detailed information on individual trees, which are essential for developing and validating allometric equations (Laurin et al. 2016). These equations relate easily measurable parameters, such as diameter at breast height (DBH) and tree height, to biomass. However, their applicability is often limited to specific regions, forest types, or species. Remote sensing data address these limitations by providing spatially continuous coverage and capturing forest heterogeneity at multiple scales (Georgopoulos et al. 2023). For example, LiDAR-derived metrics, such as canopy height models (CHMs) and gap fraction, can be combined with in-situ measurements to improve the accuracy and scalability of biomass estimates (Tsui et al. 2013). Similarly, radar backscatter data, which are sensitive to forest density and moisture content, can be calibrated using ground-based measurements to estimate aboveground biomass (Kaasalainen et al. 2015). The use of machine learning algorithms has further enhanced the integration of in-situ and remote sensing data. Techniques such as random forests, support vector machines, and neural networks are increasingly being used to model complex relationships between remote sensing variables and biomass (Cui 2019). These approaches allow for the incorporation of multiple data sources, including spectral indices, LiDAR metrics, and topographic variables, resulting in more robust and accurate models (Kong et al. 2018). Despite significant advancements, several challenges remain in forest biomass modeling. One major issue is the uncertainty associated with remote sensing measurements, which can arise from sensor limitations, atmospheric conditions, or data processing errors (Chen et al. 2016). Additionally, the scalability of allometric equations and remote sensing models is often limited by the lack of ground truth data across diverse forest types and ecological conditions (Pittman et al. 2015). Another challenge is the temporal variability of forest biomass, which is influenced by factors such as seasonal changes, disturbances, and forest growth dynamics. Capturing this variability requires high-frequency remote sensing observations and robust temporal modeling techniques (Naik et al. 2021). Furthermore, the integration of belowground biomass, which accounts for a significant portion of total forest carbon stocks, remains an underexplored area in biomass modelling. Forest biomass modelling is a critical component of efforts to understand and mitigate climate change. The integration of in-situ measurements and remote sensing data offers a powerful approach to overcome the limitations of traditional methods and achieve accurate, scalable, and spatially explicit biomass estimates. Advances in remote sensing technologies, coupled with the use of machine learning algorithms, have opened new avenues for research and applications in this field. However, addressing the challenges of uncertainty, scalability, and temporal variability requires continued innovation and collaboration among researchers, practitioners, and policymakers. By leveraging the strengths of both in-situ and remote sensing approaches, forest biomass modeling can provide the data and insights needed to support sustainable forest management, carbon accounting, and climate change mitigation efforts worldwide. Study area description Narmadapuram district (Fig. 1), situated in the central Indian state of Madhya Pradesh (Latitude: 22.75° N, Longitude: 77.72° E) boasts a rich natural landscape characterized by dense forests, diverse flora, and abundant wildlife (Yadava and Sinha 2020). This region serves as a significant hub for studying forest biomass, offering a plethora of research opportunities and insights into the dynamics of ecosystem functioning (Rajput et al. 2021). The district's topography comprises undulating terrain, interspersed with lush forests, which include both natural and man-made plantations. These forests play a vital role in regulating the local climate, conserving soil fertility, and providing a habitat for numerous species of plants and animals. Research conducted in Narmadapuram district often focuses on quantifying forest biomass, examining carbon sequestration rates, and assessing the impact of anthropogenic activities on forest health. Studies utilize various methodologies, including remote sensing techniques, field surveys, and biomass modeling, to gain a comprehensive understanding of the region's forest ecosystems. Moreover, initiatives such as afforestation programs and community-based conservation efforts further contribute to the sustainable management of forest resources in Narmadapuram district. Assessment of Forest Biomass in Narmadapuram District, Madhya Pradesh, India. provide valuable insights into the dynamics of biomass accumulation and carbon storage in this ecologically significant area, informing conservation strategies and policy interventions aimed at preserving the region's natural heritage (Ahirwar et al. 2020). Data Inventory (Fig. 2) represent the image presents a boxplot depicting the distribution of the number of trees across various forest plots labeled AA to AJ. Each plot corresponds to a different species, represented by their scientific names, and each species is assigned a specific color for easy identification. The boxplots illustrate the frequency distribution of tree counts, with the boxes representing the interquartile range (IQR) and the whiskers indicating the range within which the tree count falls, while black dots signify outliers. The plot reveals significant variation in tree counts across different plots. Some species, like Acacia catechu , Bauhinia racemosa , and Terminalia arjuna , show more clustered values, indicating higher tree frequencies, while others exhibit greater dispersion, with boxes showing more spread-out data. Certain plots have outliers, marked by the black dots, which represent unusually high or low numbers of trees compared to the general distribution. Additionally, there are differences in tree abundance across the plots; for example, plots like AC and AF have relatively higher tree counts, while others like AE and AJ display lower or more variable distributions. This variation could be attributed to differences in plot size, forest density, or environmental conditions affecting the growth of each species. Overall, the boxplot provides a clear visual representation of tree population distributions across the forest plots, highlighting species diversity, tree count variation, and the factors influencing these patterns, which could be useful for forest management, biodiversity studies, and ecological assessments. Methods and Materials Experiment design and Field data The biomass study was conducted across multiple 100m x 100m plots within the Narmadapuram district, Madhya Pradesh, to assess spatial distribution and density. Ten well-defined plots were demarcated with latitude and longitude coordinates for each corner point, ensuring uniform sampling and precise measurements. Biomass data varied across plots, reflecting different vegetation densities. Plot AA recorded 46.34 t/ha, Plot AB had 67.88 t/ha, while Plot AC measured 90.45 t/ha. Plot AD showed a biomass of 61.22 t/ha, and Plot AE had a significant value of 107.31 t/ha. Other plots, such as Plot AF and Plot AG, recorded 101.28 t/ha and 55.11 t/ha, respectively. Plot AH recorded 71.10 t/ha, while Plot AI had the highest biomass value of 294.58 t/ha, indicating dense vegetation. Plot AJ recorded 73.56 t/ha. The total biomass for all plots summed to 968.86 t/ha. This systematic grid-based approach facilitated a comprehensive understanding of biomass variation across the study area, highlighting its importance in environmental monitoring and resource management (Fig. 3). Forest Parameters Estimation Table 1 Summary statistic forest plot data Variable Tree count Mean Std. Dev. Min Pctl. 25 Pctl. 75 Max site: AA(Plot 1) GBH 248 71 28 19 54 80 230 DBH 248 0.23 0.088 0.061 0.17 0.26 0.73 TOTH 248 8.6 2.8 2.1 6.5 11 18 Biomass 248 0.19 0.31 0.00087 0.044 0.19 2.6 site: AB(Plot 2) GBH 258 75 27 29 56 89 203 DBH 258 0.24 0.086 0.092 0.18 0.28 0.65 TOTH 258 9.6 3.4 4 7.7 11 49 Biomass 258 0.26 0.46 0.0042 0.058 0.25 2.8 site: AC(Plot 3) GBH 390 61 31 28 42 68 306 DBH 390 0.2 0.099 0.089 0.13 0.22 0.97 TOTH 390 8.3 9 3.9 6 8.6 112 Biomass 390 0.23 0.44 0.00029 0.027 0.18 3.8 site: AD(Plot 4) GBH 431 60 25 24 41 69 194 DBH 431 0.19 0.08 0.076 0.13 0.22 0.62 TOTH 431 8.2 1.9 3.7 6.9 9.4 15 Biomass 431 0.14 0.23 0.0012 0.032 0.14 2.1 site: AE(Plot 5) GBH 486 64 41 8 39 76 471 DBH 486 0.2 0.13 0.025 0.12 0.24 1.5 TOTH 486 12 3.2 3.8 9.5 14 20 Biomass 486 0.22 0.62 0.00036 0.044 0.2 8.1 site: AF(Plot 6) GBH 315 79 39 16 50 102 230 DBH 315 0.25 0.13 0.051 0.16 0.32 0.73 TOTH 315 9 2.7 3.2 6.9 11 15 Biomass 315 0.32 0.45 0.0029 0.069 0.4 3.5 site: AG(Plot 7) GBH 294 64 30 25 42 80 202 DBH 294 0.2 0.095 0.08 0.13 0.25 0.64 TOTH 294 8.5 2.2 3.8 6.9 10 14 Biomass 294 0.19 0.28 0.0046 0.049 0.22 2.6 site: AH(Plot 8) GBH 315 73 30 24 54 86 366 DBH 315 0.23 0.097 0.076 0.17 0.27 1.2 TOTH 315 10 2.5 1.3 8.6 12 17 Biomass 315 0.23 0.45 0.0032 0.065 0.25 5.4 site: AI(Plot 9) GBH 389 81 51 33 62 89 887 DBH 389 0.26 0.16 0.11 0.2 0.28 2.8 TOTH 389 14 3.3 2.8 12 16 22 Biomass 389 0.76 1.2 0.014 0.14 0.56 5.7 site: AJ(Plot 10) GBH 475 49 23 19 34 57 204 DBH 475 0.16 0.073 0.061 0.11 0.18 0.65 TOTH 475 6.6 1.8 2 5.2 7.8 13 Biomass 475 0.15 0.29 0.0034 0.021 0.09 1.4 The Table 1 summary statistics highlight significant variability in forest structure across the surveyed sites (AA to AJ). The average number of Sub Plots per site ranges from 4.4 to 5.2, indicating consistent sampling density, with sites AE and AI showing slightly higher averages. Tree size, represented by GBH and DBH, varies considerably. Site AI has the largest trees, with an average GBH of 81 cm and DBH of 0.26 m, while site AJ has the smallest, with a GBH of 49 cm and DBH of 0.16 m. This trend is mirrored in total tree height (TOTH), where AI trees are the tallest (mean: 14 m) and AJ the shortest (mean: 6.6 m), reflecting a positive relationship between girth and height. The number of stems (No stem) remains relatively consistent across sites, averaging around 1.1, with slight increases in sites AC and AJ. Tree volume and biomass exhibit the most notable variations, emphasizing differences in forest composition. Site AI has the highest mean volume (1.0 m³) and biomass (0.76 tons), indicating mature forest characteristics. In contrast, site AJ records the lowest values, with a mean volume of 0.2 m³ and biomass of 0.15 tons, reflecting a younger or less dense forest. The standard deviations in GBH, DBH, TOTH, and Volume reveal substantial heterogeneity, with sites like AE and AI showing the greatest variability, likely due to a mix of small and large trees. Maximum values across variables further highlight the diversity, with AI having the tallest tree (22 m) and the largest GBH (887 cm), while AJ’s metrics remain the smallest. Overall, the analysis underscores significant structural differences among sites. Mature forests like AI exhibit larger trees, higher volumes, and greater biomass, contributing more to carbon storage, while younger forests like AJ showcase smaller trees and lower biomass. These findings provide a comprehensive understanding of forestdynamics and are critical for resource management and carbon stock assessment. Remote sensing and Data processing SAR Sentinel Data processing : SAR (Synthetic Aperture Radar) data processing is a vital component of remote sensing and Earth observation, providing a wealth of information for a wide range of applications, including agriculture, forestry, geology, disaster monitoring, and urban planning. This complex process involves the acquisition of raw SAR data, its conversion into meaningful information, and subsequent analysis. In this method for processing the SAR data, we will understand the various stages of SAR data processing, from data process to interpretation. Data Preprocessing : The Sentinel-1A SAR data processing chain typically begins with preprocessing steps to improve the quality and utility of the raw imagery data. The process involves radiometric geometric calibration, terrain correction, and removal of speckle noise, which is inherent to SAR imagery. Additionally, geometric correction is performed to ensure accurate georeferencing. These preprocessing steps are essential for generating accurate and reliable results from SAR data. Raw SAR data is acquired, a line of preprocessing steps are requiring to correct for geometric and radiometric distortions. These corrections ensure that the data accurately represent the Earth's surface. Geometric corrections involve accounting for the sensor's position, attitude, and the topography of the observed area. Radiometric corrections adjust for variations in signal strength caused by factors like atmospheric conditions, sensor settings, and terrain roughness. Key preprocessing steps include radiometric calibration, terrain correction, and geocoding. Radiometric calibration normalizes the radar signal to an absolute scale, allowing for consistent analysis. Terrain correction compensates for topographic effects, ensuring that features at different elevations are accurately represented. Geocoding involves transforming the data into geographic coordinates, enabling easy integration with other spatial datasets. Image Formation and Focusing The core principle of SAR data processing is the creation of high-resolution images through a synthetic aperture. In this process, SAR sends microwave pulses towards the Earth's surface and records the echoes reflected back. These echoes are then used to construct a high-resolution image of the area, with finer details obtained through the synthetic aperture's motion. Focusing techniques, including range compression and azimuth compression, are applied to sharpen the image and reduce blurring caused by satellite motion. Optical Data : sentinel 2 A data used for extraction of vegetation index, such as NDVI, SVAI and other vegetation index. sentinel data has 10 m spatial resolution. Sentinel-2A optical data is a valuable resource for extracting various vegetation indices crucial for environmental monitoring and agricultural analysis. With its 10-meter spatial resolution, Sentinel-2A imagery provides detailed insights into land cover dynamics and vegetation health. One of the most widely used vegetation indices derived from Sentinel-2A data is the Normalized Difference Vegetation Index (NDVI), which quantifies the presence and vigor of vegetation based on the contrast between near-infrared and red reflectance. Additionally, Sentinel-2A data enables the calculation of other vegetation indices such as the Soil-adjusted Vegetation Index (SAVI), which accounts for soil brightness in vegetation assessments, and the Enhanced Vegetation Index (EVI), offering improved sensitivity in high biomass regions. These indices play a pivotal role in monitoring ecosystem health, crop growth, and assessing the impact of environmental changes on vegetation dynamics. By leveraging Sentinel-2A optical data and its suite of vegetation indices, researchers and land managers can make informed decisions for sustainable land use and resource management practices. AGB estimation In Central India, where tropical deciduous forests are predominant, region-specific equations for Above Ground Biomass (AGB) estimation are often developed based on field data. One commonly referenced source is Forest Survey of India (FSI) or studies conducted in the Satpura, Vindhyan ranges, and other dense forest zones of Madhya Pradesh and Chhattisgarh. AGB can be estimated using allometric equations involving tree height DBH and wood density. The allometric equations are of the power function type (eqn. 1) ABG=0.0509*(DBH 2 * H * WD) eqn. 1 where H (m) is the tree height is the DBH(cm), is the Diameter at Breast Height and WD (g/cm³) is Wood Density (species-specific) is are parameters by in situ measurements. Linear regression: Linear regression is a statistical method used to model the relationship between two or more variables by fitting a straight line to the observed data points. The goal is to find the best-fitting line that minimizes the difference between the actual data points and the predicted values generated by the line. This method is commonly used in various fields such as economics, finance, biology, and social sciences to analyze and predict the behavior of dependent variables based on one or more independent variables. By estimating the parameters of the linear equation, such as the slope and intercept, linear regression allows researchers to make predictions and infer relationships between variables in a simple and interpretable manner.Top of Form Exponential: Exponential growth or decay is a fundamental concept in mathematics and science. It refers to a process where a quantity increases or decreases at a rate proportional to its current value. In exponential growth, the rate of increase itself increases over time, leading to rapid growth, while in exponential decay, the rate of decrease decreases over time, resulting in rapid decline. The general form of an exponential function is , where: f(x) is the value of the function at a given point x, a is the initial value or the value at, b is the base of the exponential function, often referred to as the growth or decay factor, x is the exponent, representing the independent variable such as time. Exponential growth is commonly observed in various natural phenomena such as population growth, compound interest, and the spread of diseases in uncontrolled environments. It's also applicable in fields like physics, chemistry, and economics. Understanding exponential growth is crucial because it can lead to significant consequences, both positive and negative, depending on the context. For instance, it can illustrate the potential of technological advancement or highlight the urgency of addressing issues like environmental sustainability. Random forest Modeling : Random forest modeling is a versatile and powerful machine learning technique that leverages the collective wisdom of multiple decision trees to make predictions. Each tree in the forest is trained on a random subset of the data and uses a random subset of the features, which helps to reduce overfitting and increase the model's robustness. During prediction, the results from all the trees are aggregated, typically by averaging for regression tasks or by voting for classification tasks, to produce the final output. This ensemble approach often leads to improved accuracy and generalization compared to individual decision trees. Random forests are widely used across various domains, including finance, healthcare, and marketing, for tasks such as classification, regression, and outlier detection. Additionally, they offer built-in feature importance measures, which can help identify the most relevant variables driving the predictions. Overall, random forest modeling is a valuable tool in the machine learning toolkit, offering a balance of performance, interpretability, and ease of use. Support Vector machine : A Support Vector Machine (SVM) is a powerful supervised learning algorithm used for classification and regression tasks. Its primary objective is to find the optimal hyperplane that best separates the data points into different classes in a high-dimensional space. SVM works by identifying the support vectors, which are the data points closest to the decision boundary, and maximizing the margin between classes. This margin allows the SVM to generalize well to new, unseen data, making it robust against overfitting. Additionally, SVM can handle both linear and non-linear classification tasks through the use of different kernel functions, such as linear, polynomial, radial basis function (RBF), and sigmoid. SVMs are widely used in various fields, including image classification, text classification, and bioinformatics, due to their versatility, effectiveness, and ability to handle high-dimensional data. Model Performance Evaluation Metrics for Forest AGB Estimation The linear regression (LR) model and machine learning techniques, including Random Forest (RF), Artificial Neural Network (ANN), and Support Vector Machine (SVM), were validated using 10 forest inventory plots. To evaluate and compare the performance of these models in estimating forest above-ground biomass (AGB), three statistical metrics were used: the coefficient of determination (R 2 ), mean absolute error (MAE), and root mean square error (RMSE). These metrics are commonly employed to assess the accuracy of predictive models by measuring the differences between observed and predicted values. A better model is indicated by higher R 2 values and lower RMSE and MAE values. The formulas for calculating R 2 , RMSE, and MAE are as follows: 1. Coefficient of Determination (R²) The coefficient of determination evaluates how well the model explains the variability in the dependent variable. R² = 1 - [Σ (yᵢ - ŷᵢ) ² / Σ (yᵢ - ȳ) ²] Where yᵢ = observed value, ŷᵢ = predicted value, ȳ = mean of observed values, n = number of observations 2. Root Mean Square Error (RMSE) The RMSE measures the average magnitude of the error between predicted and observed values. RMSE = √ [1/n * Σ (yᵢ - ŷᵢ) ²] Where yᵢ = observed value, ŷᵢ = predicted value, and n = number of observations 3. Mean Absolute Error (MAE) The MAE calculates the average absolute difference between predicted and observed values. MAE = [1/n * Σ|yᵢ - ŷᵢ|] Where yᵢ = observed value, ŷᵢ = predicted value and n = number of observations Results and Discussion The Table 2 compares four modeling methods Log1, Log2, Weibull, and Michaelis based on their performance metrics: Residual Standard Error (RSE), RSE log, and Average Bias. Log1 (blue) demonstrates relatively low RSE (4.067850) and RSE log (0.3157938), along with a minimal Average Bias of -0.002102274, indicating good overall performance with minimal systematic deviation. Similarly, Log2 (green) slightly outperforms Log1 with the lowest RSE (4.019404) and RSE log (0.3133169) while maintaining a small Average Bias of -0.003154677. This suggests that Log2 provides the most accurate and reliable predictions among the methods compared. In contrast, the Weibull method (orange) exhibits the highest RSE (4.674796) and the largest positive Average Bias (0.242635165), indicating poor performance with significant overestimation. Additionally, the RSE log metric is unavailable for Weibull, suggesting potential limitations in its compatibility with logarithmic transformations. The Michaelis method (purple) performs slightly better than Weibull, with an RSE of 4.476368 and a positive Average Bias of 0.206571212. However, its error and bias levels remain higher than those of Log 1 and Log 2 , and its RSE log metric is also unavailable. In conclusion, Log 2 emerges as the best-performing method due to its low error metrics and minimal bias, making it the most suitable for accurate and unbiased predictions. Log1 is a close second, while the Weibull and Michaelis methods are less reliable due to their higher errors and significant biases. These findings emphasize the importance of selecting modeling approaches with low RSE and minimal bias for robust predictive performance (Fig. 4). Forest Inventory Data The (Fig. 5) represents a bar chart illustrating the tree count distribution across ten forest inventory plots, labeled as AA, AB, AC, AD, AE, AF, AG, AH, AI, and AJ, corresponding to plots 1 through 10. The x-axis represents different forest plots, while the y-axis quantifies the tree count in each plot. The purpose of this figure is to visualize the variation in tree density among different locations. The bar chart employs vertical bars to represent the tree count within each forest plot, with the height of each bar correlating to the number of trees recorded in the respective plot. This visual representation allows for an easy comparison of tree density. The x-axis is labeled as Forest Plots, indicating different sampling locations within the forest, while the y-axis is labelled as Tree Count, representing the number of trees recorded in each plot. The tree count scale ranges from 0 to 500, providing a clear measure of tree density. The bars are shaded in a uniform dark gray color, ensuring clarity and ease of interpretation, while the background grid lines aid in estimating the approximate tree count in each plot.The tree count varies across the ten forest plots, with some plots exhibiting higher densities while others have relatively lower tree counts. Some plots, such as AA and AB, show a relatively low tree count, with values around 250 trees, whereas AG and AH also exhibit lower tree counts, suggesting sparse tree coverage in these areas. Moderate tree count plots include AF, which is slightly above the low-density group but still falls behind the highest-density plots, while AC and AI have intermediate tree densities with moderate variations in tree count. The highest tree counts are observed in AD and AE, with AE reaching the maximum value of approximately 500 trees. AJ also exhibits a high tree count, comparable to AD.Several ecological and environmental factors could contribute to the variation in tree counts across different plots. Soil quality and nutrient availability play a crucial role in tree density, as nutrient-rich soil supports higher tree counts, whereas poor soil conditions may lead to sparse tree coverage. Topographical factors, such as elevation, slope, and aspect, also influence tree growth. For instance, steep slopes may have fewer trees due to soil erosion and water runoff. Additionally, microclimatic conditions like sunlight exposure, temperature, and moisture levels can significantly impact tree density. Forest plots with better moisture retention may support higher tree counts. Human or natural disturbances, including logging activities, land clearance, fire incidents, storms, pest infestations, or disease outbreaks, could also lead to reduced tree densities in certain areas. Furthermore, species composition and growth patterns vary, with some tree species growing in dense clusters while others are sparsely distributed, potentially explaining the differences in tree counts. Understanding tree distribution across forest plots has important ecological implications. Forest plots with higher tree densities may serve as crucial habitats for wildlife, contributing to biodiversity conservation. On the other hand, low-density plots may require conservation efforts such as afforestation and protection from further degradation. High-density plots should be monitored to maintain ecological balance and prevent resource competition among trees. Additionally, forest plots with higher tree densities play a significant role in carbon sequestration, helping mitigate climate change by absorbing atmospheric carbon dioxide. In conclusion, the bar chart effectively illustrates the variation in tree counts across ten forest inventory plots, revealing significant differences in tree density that could be attributed to multiple ecological and environmental factors. Such information is crucial for forest management, conservation planning, and ecological studies. Understanding these patterns helps in making informed decisions regarding sustainable forest resource management and biodiversity conservation. Association between field data of tree The analysis (Fig. 6) of the correlation matrix and pair plot reveals important relationships between key tree characteristics, offering insights into their ecological and practical implications. The correlation between DBH and Volume (0.57) indicates a moderate positive relationship, suggesting that as tree diameter increases, so does its volume, although other factors like tree height, wood density, and shape also play a role. Similarly, DBH shows a moderate correlation with Biomass (0.54), reinforcing its importance in studies of carbon storage and forest inventory. These trends suggest that while DBH is a significant predictor of both Volume and Biomass, it is not the sole determinant. In contrast, the correlations between tree height (TOTH) and Volume (0.30) and Biomass (0.29) are weak, highlighting that height alone does not strongly predict tree volume or biomass. This may be due to variations in diameter, crown shape, branch density, and wood density among trees of similar height. The strongest relationship is observed between Volume and Biomass (0.98), indicating they are almost directly proportional and confirming that Volume is a reliable predictor of Biomass in this dataset. This strong correlation has practical significance for forest inventories, as Biomass can be estimated accurately from Volume, reducing the need for direct measurements. The findings also emphasize the utility of DBH in forest management, though its moderate correlations suggest that combining it with other metrics like height and wood density can improve predictive accuracy. Meanwhile, the limited role of tree height as an independent predictor of Biomass or Volume underscores the need for comprehensive measurements in ecological models. The analysis highlights that correlation does not imply causation and the observed relationships may vary across forest types, species, and environmental conditions. These findings underscore the complexity of tree structure and the importance of using a combination of measurements to accurately assess tree characteristics. Understanding these relationships aids in informed decision-making for forest conservation, resource management, and carbon sequestration, ultimately supporting more sustainable forestry practices. Plot wise correlation between height and DBH The scatter plots depict the relationship between tree height and DBH (Diameter at Breast Height) across ten forest plots (Plot1 to Plot10), with each plot featuring a regression line and confidence intervals (Fig. 7). Most plots exhibit a positive trend, indicating that tree height generally increases with DBH, which aligns with expectations since larger diameters are often associated with taller trees. However, the strength of this correlation varies across the plots. Some, such as Plot1, Plot7, and Plot8, display tighter clusters around the regression line, suggesting a stronger relationship, while others, like Plot3 and Plot4, have more scattered data points, reflecting weaker correlations or greater variability. The slopes of the regression lines also differ, with steeper slopes (e.g., in Plot1 and Plot7) indicating a rapid increase in height with DBH, while flatter slopes (e.g., in Plot3 and Plot4) suggest a slower increase. Additionally, the range of DBH and height values varies across plots, with Plot2 covering a wider DBH range and Plot3 focusing on smaller values. Outliers are evident in several plots, with some trees exhibiting unusually tall heights for small DBH or vice versa, potentially due to measurement errors, unique species traits, or atypical growth conditions. Confidence intervals around the regression lines indicate the level of uncertainty, with wider intervals in plots like Plot3 suggesting less confidence in the model fit due to sparse or variable data. These variations highlight the influence of site-specific factors, such as species composition, soil fertility, and climate, on the height-DBH relationship. This underscores the need for localized data when developing allometric equations or estimating tree characteristics and biomass for forest management or research purposes. Relationship Between Tree Height, DBH, and Biomass in Forest Inventory Data The scatter plot illustrates the relationship between tree height (in meters) and DBH (Diameter at Breast Height, in meters), with bubble size and color representing biomass (Fig. 8). The majority of the data points are concentrated near the lower range of both DBH and height, indicating that most trees in the dataset are relatively small in size. A few outliers are evident, with some trees exhibiting extremely tall heights (over 90 meters) despite having small DBH values, as well as a few trees with unusually large DBH values (around 2 meters). The size and color of the bubbles provide insight into biomass distribution. Larger and darker bubbles, indicating higher biomass, are generally associated with higher DBH values. This is consistent with the expectation that larger trees tend to have greater biomass. However, there are also instances where trees with moderate DBH and height exhibit significant biomass, suggesting variability in wood density or tree form. Overall, the plot highlights the positive relationship between DBH and biomass while showcasing variability in the relationship between height and biomass. The outliers and variability might be attributed to species differences, site-specific factors, or measurement errors. This visualization underscores the complexity of predicting biomass based on structural parameters and the importance of considering additional variables, such as wood density or crown dimensions, in ecological studies and forest management. Multi variable of environmental parameter for Biomass estimation The image displays a correlation matrix that visualizes the relationships between various remote sensing variables and landscape characteristics using a color-coded heat map (Fig. 9). The correlations range from -1 to +1, indicating negative to positive associations. Strong positive correlations are observed among variables such as Sigma0_VH, Sigma0_VH_GLCM Variance, Sigma0_VH_GLCM Mean and Sigma0_VH_GLCM Correlation, with values exceeding 0.8. These relationships suggest that these variables are closely related and likely follow similar patterns, particularly in terms of radar backscatter intensity. Moderate positive correlations are seen between variables like Ratio and Sigma0_VV_Contrast (0.69), as well as between elevation and aspect (0.65), indicating a shared association but with less strength. In contrast, weak or negative correlations are noted with variables such as savi and ndvi which show minimal influence from other landscape metrics, especially radar-based variables. For example, the correlation between savi and other metrics is below 0.2, suggesting that vegetation indices like SAVI and NDVI are less dependent on factors like surface roughness or moisture content (Table 3). Additionally, the variable Diff shows a moderate positive correlation with Sigma0_VH (0.55), while other relationships, such as between slope and elevation, are weak or negative. The heatmap's color gradient, ranging from blue (positive correlations) to red (negative correlations), provides an intuitive way to assess the strength and direction of relationships between the variables. Hence, strong correlations between radar-based variables indicate they might be used interchangeably in models, while weak correlations with vegetation indices suggest these should be treated separately in analyses of environmental features. Table 3 Predictor Variable for Biomass estimation Predictor Description Formula Sigma0_VH Radar backscatter intensity in the vertical-horizontal polarization. Indicates surface texture and structure. σ0,VH=10×log10(RVH) Sigma0_VH_GLCMVariance Measures the variance in the Gray Level Co-occurrence Matrix (GLCM), highlighting texture heterogeneity in images. GLCM Variance = Σi,j p(i,j) * (i - μ)^2 Sigma0_VH_GLCMMean The mean value from the Gray Level Co-occurrence Matrix, indicating average intensity. GLCM Mean = Σi,j p(i,j) * (i + j) Sigma0_VH_GLCMCorrelation Measures the correlation between pixel pairs in the GLCM, reflecting spatial relationship between neighboring pixels. GLCM Correlation = Σi,j p(i,j) * (i - μ_x) * (j - μ_y) / (σ_x σ_y) Sigma0_VV Radar backscatter intensity in vertical-vertical polarization. Indicates surface roughness and structure. σ0_VV = 10 * log10(R_VV) Sigma0_VV_Contrast Contrast in radar backscatter intensity, revealing the roughness or texture of the surface. Contrast = Σi,j p(i,j) * |i - j| Ratio Ratio of the backscatter in different polarization channels, useful for surface characterization. Ratio = σ0_VH / σ0_VV aspect Measures the direction of the slope of the terrain, indicating the orientation of the landscape. Aspect = atan2(∂Z/∂y, ∂Z/∂x) slope Measures the steepness of the terrain, calculated from the elevation data. Slope = arctan(ΔZ / Δdistance) elevation The height of the terrain or surface above a reference point, typically sea level. Elevation = Height above sea level multi Multiplier factor, often used in scaling or transforming certain measurements in data analysis. Typically a scaling constant depending on the dataset or model. Contrast Measures differences in pixel intensities in the image, related to surface texture or heterogeneity. Contrast = Σi,j p(i,j) * |i - j| savi Soil-Adjusted Vegetation Index (SAVI), a vegetation index that accounts for soil background. SAVI = (NIR - RED) * (1 + L) / (NIR + RED + L) ndvi Normalized Difference Vegetation Index (NDVI), a common vegetation index used to detect vegetation health. NDVI = (NIR - RED) / (NIR + RED) Diff Difference between two or more variables or features, often used for change detection or comparative analysis. Diff = X - Y Estimation of AboveGround Biomass Table 4 Model for Aboveground Biomass Prediction Method RMSE MAE relRMSE r 2 1 Exp Regression 2.26 0.97 0.1471322 0.04 2 Linear Regression 2.48 1.34 0.1616262 -0.16 3 Random forest 1.61 0.84 0.1046609 0.51 4 SVM 2.00 1.06 0.1301456 0.25 The Table 4 provides a comparative analysis of four different regression methods: Exponential Regression, Linear Regression, Random Forest, and Support Vector Machine (SVM). Each method's performance is evaluated based on four metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), relative RMSE (relRMSE), and the coefficient of determination (r²). Exponential Regression shows an RMSE of 2.26, an MAE of 0.97, a relRMSE of 0.1471322, and an r² value of 0.04. These results indicate that while the model has moderate error rates, its predictive accuracy is low, as reflected by the low r² value. Linear Regression has an RMSE of 2.48 and an MAE of 1.34, with a relRMSE of 0.1616262 and an r² of -0.16. This method demonstrates higher error rates compared to Exponential Regression and also shows a negative r² value, suggesting that it performs worse than a horizontal line model. Random Forest performs the best among the four methods with an RMSE of 1.61 and an MAE of 0.84, alongside a relRMSE of 0.1046609 and an r² of 0.51. The lower error metrics and a positive r² value indicate a more accurate and reliable predictive performance. Support Vector Machine (SVM) shows an RMSE of 2.00, an MAE of 1.06, a relRMSE of 0.1301456, and an r² of 0.25(Fig. 10). This method performs better than both Exponential and Linear Regression but is outperformed by Random Forest in all evaluated metrics. Hence, Random Forest emerges as the most effective regression method among the ones evaluated, demonstrating the lowest error rates and the highest predictive accuracy. While other methods like support vector machine and expotentional result showed that not good for the present dataset. Aboveground Biomass Predicted Vs Actual by Quantiles The stacked bar chart illustrates the comparison of predicted versus actual Above Ground Biomass (AGB) percentages across four quantiles (0-2, 0-3.6, 0-6.12, and 13-15.4). Each quantile shows the distribution of AGB predictions across four categories (1: 0-0.2, 2: 0-3.6, 3: 6-12, and 4: 13-15.4). In Quantile 1 (0-2), the red category (Predicted Biomass: 1) dominates with 57.69%, followed by orange (2) at 26.32% and green (3) at 13.16%, while blue (4) has the smallest contribution at 2.63%. In Quantile 2 (0-3.6), the red category decreases to 24.32%, and the green category becomes dominant at 43.24%, while orange and blue categories contribute 24.32% and 8.11%, respectively. Quantile 3 (6-12) shows a consistent dominance of the green category at 43.24%, a further decline in the red category to 13.51%, and increased contributions from orange (10.81%) and blue (32.43%). By Quantile 4 (13-15.4), the blue category dominates with 56.76%, indicating improved prediction performance for higher AGB values, while the orange and green categories contribute equally at 18.92%, and the red category accounts for only 5.41%. In conclusion, the chart highlights the model’s progression in prediction accuracy across quantiles. Initially, the red category dominates, reflecting an underestimation of biomass in lower quantiles, but its influence decreases in higher quantiles as the blue category gains dominance. This shift indicates an improvement in predicting higher biomass values in later stages. However, the persistence of red and orange categories across quantiles suggests a need for further refinement of the model to enhance consistency and accuracy across all AGB ranges (Fig. 11). Variable importance AboveGround Biomass The bar chart highlights the importance of various variables in predicting Above Ground Biomass (AGB) using a Random Forest (RF) model. The x-axis represents the percentage increase in Mean Squared Error (% Inc MSE) when a variable is excluded, which indicates the variable's relative importance to the model. Among the variables, Sigma0_VV_GLCM Variance emerges as the most influential, with the highest % Inc MSE, underscoring its critical role in accurate biomass prediction. Following this, Sigma0_VH and Sigma0_VH_GLCM Correlation are identified as the second and third most significant variables, respectively, highlighting their substantial contributions to the model's predictive power. Moderately important variables include Diff and multi, which enhance the model's accuracy but to a lesser extent. Variables such as Ratio, mean, and Sigma0_VV show lower importance but still contribute meaningfully to the predictions. On the other hand, traditional vegetation indices like NDVI and SAVI demonstrate the least impact, with minimal % Inc MSE, suggesting their limited effectiveness compared to radar-based variables. In conclusion, the analysis underscores the dominance of radar-based variables, such as Sigma0_VV_GLCM Variance and Sigma0_VH, in predicting AGB. Their sensitivity to structural and textural properties makes them more effective for biomass estimation compared to traditional vegetation indices. This finding highlights the importance of carefully selecting variables to optimize the performance of biomass prediction models (Fig. 12). Conclusion The study on Forest Biomass Modeling Using In-Situ and Remote Sensing Data reveals the potential of integrating remote sensing data with ground-based measurements for accurate biomass estimation. Various modeling techniques were applied, including Exponential Regression, Linear Regression, Random Forest, and Support Vector Machine (SVM), with each method showing different levels of performance in terms of prediction accuracy. Among the methods tested, Random Forest exhibited the best performance with the lowest RMSE (1.61) and MAE (0.84), along with the highest r 2 value of 0.51, indicating that it is the most reliable model for forest biomass prediction in this study. In contrast, Exponential Regression demonstrated a lower RMSE (2.26) and MAE (0.97) but a relatively low r 2 (0.04), suggesting that while it performed decently in terms of error metrics, its overall explanatory power was limited. Linear Regression showed poorer results, with a higher RMSE (2.48) and r 2 value of -0.16, indicating that it struggled to capture the complexity of the biomass data. SVM also showed moderate performance with a RMSE of 2.00, MAE of 1.06, and an r 2 of 0.25, demonstrating reasonable predictive ability, but not as strong as Random Forest. The integration of remote sensing data, particularly radar-based variables like Sigma0_VV_GLCM Variance and Sigma0_VH, significantly improved the model’s predictive capabilities, especially in forests with complex canopies where traditional vegetation indices like NDVI and SAVI were less effective. This study highlights the importance of combining multiple data sources and modeling techniques for accurate forest biomass estimation. The findings underscore the strength of machine learning models, particularly Random Forest, for biomass prediction and their potential application in large-scale forest management and carbon stock estimation. In conclusion, this research demonstrates the effectiveness of using remote sensing and in-situ data for forest biomass modeling. Future work should focus on further refining these models, incorporating additional data sources, and exploring new remote sensing technologies to enhance the precision and scalability of biomass estimation across different forest types and regions Declarations Acknowledgement We would like to express our sincere gratitude to Forest department of Madhya Pradesh providing for their continuous support over the years. We also extend our thanks to the field staff whose effort and dedication were essential for the data collection throughout the duration of this study. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability Data will be made available on request Conflict of interest The authors have no conflict of interest to declare that are relevant to the contents of this article. References Abbas S, Wong MS, Wu J, Shahzad N, Irteza SM (2020) Approaches of Satellite Remote Sensing for the Assessment of Above-Ground Biomass across Tropical Forests: Pan-tropical to National Scales. Remote Sensing 12:3351 Ahirwar R, Malik MS, Shukla JP (2020) Groundwater vulnerability assessment of Hoshangabad and Budni industrial area, Madhya Pradesh, India, using geospatial techniques. Applied Water Science 10 Baccini A, Friedl MA, Woodcock CE, Warbington R (2004) Forest biomass estimation over regional scales using multisource data. Geophysical Research Letters 31 Brovkina O, Novotný J, Cienciala E, Zemek F, Russ R (2016) Mapping forest aboveground biomass using airborne hyperspectral and LiDAR data in the mountainous conditions of Central Europe. Ecological Engineering 100:219 Brown S (1997) Estimating biomass and biomass change of tropical forests: A primer Chen Q, McRoberts RE, Wang C, Radtke PJ (2016) Forest aboveground biomass mapping and estimation across multiple spatial scales using model-based inference. Remote Sensing of Environment 184:350 Cui H (2019) Sub-urban land classification using GF-2 images and support vector machine method. IOP Conference Series Earth and Environmental Science 351:12028 Duncanson L, Armston J, Disney M, Avitabile V, Barbier N, Calders K, Carter S, Chave J, Herold M, Crowther TW, Falkowski MJ, Kellner JR, Labrière N, Lucas R, MacBean N, McRoberts RE, Meyer V, Næsset E, Nickeson J, Paul KI, Phillips OL, Réjou‐Méchain M, Román MO, Roxburgh SH, Saatchi S, Schepaschenko D, Scipal K, Siqueira P, Whitehurst A, Williams M (2019) The Importance of Consistent Global Forest Aboveground Biomass Product Validation. Surveys in Geophysics 40:979 Eisfelder C, Kuenzer C, Dech S (2011) Derivation of biomass information for semi-arid areas using remote-sensing data. International Journal of Remote Sensing 33:2937 Gao Y, Lu D, Li G, Wang G, Chen Q, Liu L, Li D (2018) Comparative Analysis of Modeling Algorithms for Forest Aboveground Biomass Estimation in a Subtropical Region. Remote Sensing 10:627 Georgopoulos N, Gitas IZ, Korhonen L, Antoniadis K, Stefanidou A (2023) Estimating Crown Biomass in a Multilayered Fir Forest Using Airborne LiDAR Data. Remote Sensing 15:2919 Gleason CJ, Im J (2011) A Review of Remote Sensing of Forest Biomass and Biofuel: Options for Small-Area Applications. GIScience & Remote Sensing 48:141 Griscom BW, Adams J, Ellis PW, Houghton RA, Lomax G, Miteva DA, Schlesinger WH, Shoch D, Siikamäki J, Smith P, Woodbury PB, Zganjar C, Blackman A, Campari JS, Conant RT, Delgado CL, Elias P, Gopalakrishna T, Hamsik MR, Herrero M, Kiesecker JM, Landis E, Laestadius L, Leavitt SM, Minnemeyer S, Polasky S, Potapov P, Putz FE, Sanderman J, Silvius M, Wollenberg E, Fargione J (2017) Natural climate solutions. Proceedings of the National Academy of Sciences 114:11645 Haya B, Evans SL, Brown LB, Bukoski JJ, Butsic V, Cabiyo B, Jacobson R, Kerr A, Potts MD, Sanchez DL (2023) Comprehensive review of carbon quantification by improved forest management offset protocols. Frontiers in Forests and Global Change 6 Huang W, Swatantran A, Johnson K, Duncanson L, Tang H, Dunne JO, Hurtt GC, Dubayah R (2015) Local discrepancies in continental scale biomass maps: a case study over forested and non-forested landscapes in Maryland, USA. Carbon Balance and Management 10 Kaasalainen S, Holopainen M, Karjalainen M, Vastaranta M, Kankare V, Karila K, Osmanoğlu B (2015) Combining Lidar and Synthetic Aperture Radar Data to Estimate Forest Biomass: Status and Prospects. Forests 6:252 Kellner JR, Armston J, Birrer M, Cushman KC, Duncanson L, Eck C, Falleger C, Imbach B, Král K, Krůček M, Trochta J, Vrška T, Zgraggen C (2019) New Opportunities for Forest Remote Sensing Through Ultra-High-Density Drone Lidar. Surveys in Geophysics 40:959 Kilham P, Hartebrodt C, Kändler G (2018) Generating Tree-Level Harvest Predictions from Forest Inventories with Random Forests. Forests 10:20 Knott JA, Liknes GC, Giebink CL, Oh S, Domke GM, McRoberts RE, Quirino VF, Walters BF (2023) Effects of outliers on remote sensing‐assisted forest biomass estimation: A case study from the United States national forest inventory. Methods in Ecology and Evolution 14:1587 Kong B, Yu H, Du R, Wang Q (2018) Quantitative Estimation of Biomass of Alpine Grasslands Using Hyperspectral Remote Sensing. Rangeland Ecology & Management 72:336 Kumar L, Sinha P, Taylor S, Alqurashi AF (2015) Review of the use of remote sensing for biomass estimation to support renewable energy generation. Journal of Applied Remote Sensing 9:97696 Laurin GV, Pirotti F, Callegari M, Chen Q, Cuozzo G, Lingua E, Notarnicola C, Papale D (2016) Potential of ALOS2 and NDVI to Estimate Forest Above-Ground Biomass, and Comparison with Lidar-Derived Estimates. Remote Sensing 9:18 Lu D (2006) The potential and challenge of remote sensing‐based biomass estimation. International Journal of Remote Sensing 27:1297 Lu D, Chen Q, Wang G, Liu L, Li G, Morán EF (2014) A survey of remote sensing-based aboveground biomass estimation methods in forest ecosystems. International Journal of Digital Earth 9:63 Minh DHT, Toan TL, Rocca F, Tebaldini S, Villard L, Réjou‐Méchain M, Phillips OL, Feldpausch TR, Dubois-Fernandez P, Scipal K, Chave J (2016) SAR tomography for the retrieval of forest biomass and height: Cross-validation at two tropical forest sites in French Guiana. Remote Sensing of Environment 175:138 Mo L, Zohner CM, Reich PB, Liang J, de‐Miguel S, Nabuurs GJ, Renner SS, Hoogen J van den, Araza A, Herold M, Mirzagholi L, Ma H, Averill C, Phillips OL, Gamarra JGP, Hordijk I, Routh D, Abegg M, Yao YCA, Alberti G, Zambrano AMA, Alvarado BV, Álvarez‐Dávila E, Álvarez-Loayza P, Alves LF, Amaral IL do, Ammer C, Antón‐Fernández C, Araujo‐Murakami A, Arroyo L, Avitabile V, C. GAA, Baker TR, Bałazy R, Bánki O, Barroso J, Bastian ML, Bastin J, Birigazzi L, Birnbaum P, Bitariho R, Boeckx P, Bongers F, Bouriaud O, Brancalion PHS, Brandl S, Brearley FQ, Brienen R, Broadbent EN, Bruelheide H, Bussotti F, Gatti RC, César RG, Češljar G, Chazdon RL, Chen HYH, Chisholm C, Cho H, Cienciala E, Clark CJ, Clark DB, Colletta GD, Coomes DA, Valverde FC, Corral‐Rivas JJ, Crim PM, Cumming J, Dayanandan S, Gasper AL de, Decuyper M, Derroire G, DeVries B, Djordjević I, Doležal J, Dourdain A, Obiang NLE, Enquist BJ, Eyre TJ, Fandohan AB, Fayle TM, Feldpausch TR, Ferreira LV, Finér L, Fischer M, Fletcher C, Frizzera L, Gianelle D, Glick HB, Harris DJ, Hector A, Hemp A, Hengeveld G, Hérault B, Herbohn J, Hillers A, Coronado ENH, Hui C, Ibanez T, Imai N, Jagodziński AM (2023) Integrated global assessment of the natural forest carbon potential. Nature 624:92 Mohren GMJ, Hasenauer H, Köhl M, Nabuurs G-J (2012) Forest inventories for carbon change assessments. Current Opinion in Environmental Sustainability 4:686 Næsset E, Gobakken T, Solberg S, Grégoire TG, Nelson R, Ståhl G, Weydahl DJ (2011) Model-assisted regional forest biomass estimation using LiDAR and InSAR as auxiliary data: A case study from a boreal forest area. Remote Sensing of Environment 115:3599 Naik P, Dalponte M, Bruzzone L (2021) Prediction of Forest Aboveground Biomass Using Multitemporal Multispectral Remote Sensing Data. Remote Sensing 13:1282 Pittman JJ, Arnall DB, Interrante SM, Moffet CA, Butler TJ (2015) Estimation of Biomass and Canopy Height in Bermudagrass, Alfalfa, and Wheat Using Ultrasonic, Laser, and Spectral Sensors. Sensors 15:2920 Pizaña JMG, Núñez JM, Romero NC (2016) Remote Sensing-Based Biomass Estimation. In: InTech eBooks Radtke PJ, Walker D, Frank J, Weiskittel AR, DeYoung C, MacFarlane DW, Domke GM, Woodall CW, Coulston JW, Westfall JA (2016) Improved accuracy of aboveground biomass and carbon estimates for live trees in forests of the eastern United States. Forestry An International Journal of Forest Research 90:32 Rajput A, Yadav SS, Kori PK (2021) Bird diversity of Narmada wetland area and its adjoining habitat of Mathwad Range Alirajpur Division, Madhya Pradesh, India. International Journal of Applied Research 7:184 Ranson KJ, Sun G, Lang RH, Chauhan NS, Cacciola R, Kıłıc Ö (1997) Mapping of boreal forest biomass from spaceborne synthetic aperture radar. Journal of Geophysical Research Atmospheres 102:29599 Saatchi S, Houghton RA, Alvalá RC dos S, Soares JV, Yu Y (2007) Distribution of aboveground live biomass in the Amazon basin. Global Change Biology 13:816 Schepaschenko D, Chave J, Phillips OL, Lewis SL, Davies SJ, Réjou‐Méchain M, Sist P, Scipal K, Perger C, Hérault B, Labrière N, Hofhansl F, Affum‐Baffoe K, Алейников АА, Alonso A, Amani C, Araujo‐Murakami A, Armston J, Arroyo L, Ascarrunz N, Azevedo CP de, Baker TR, Bałazy R, Bedeau C, Berry N, Bilous A, Bilous S, Bissiengou P, Blanc L, Бобкова КС, Braslavskaya T, Brienen R, Burslem DFRP, Condit R, Cuní‐Sanchez A, Данилина ДМ, Torres DDC, Derroire G, Descroix L, Sotta ED, d’Oliveira MV, Dresel C, Erwin TL, Евдокименко МД, Falck J, Feldpausch TR, Foli EG, Foster RB, Fritz S, García‐Abril A, Горнов АВ, Горнова МВ, Gothard-Bassébé E, Gourlet‐Fleury S, Guedes MC, Hamer KC, Susanty FH, Higuchi N, Coronado ENH, Hubau W, Hubbell SP, Ilstedt U, Иванов ВВ, Kanashiro M, Karlsson A, Karminov V, Killeen TJ, Koffi J-CK, Konovalova ME, Kraxner F, Krejza J, Krisnawati H, Krivobokov L, Kuznetsov MA, Lakyda I, Lakyda P, Licona JC, Lucas R, Лукина НВ, Lussetti D, Malhi Y, Manzanera JA, Marimon BS, Marimon BH, Vásquez R, Мартыненко ОВ, Matsala M, Matyashuk RK, Mazzei L, Memiaghe H, Mendoza C, Mendoza AM, Moroziuk OV, Mukhortova L, Musa S, Назимова ДИ, Okuda T, Oliveira LC de, Ontikov P, Осипов АФ (2019) The Forest Observation System, building a global reference dataset for remote sensing of forest biomass. Scientific Data 6 Shi L, Liu S (2017) Methods of Estimating Forest Biomass: A Review. InTech eBooks Simpson J, Smith TEL, Wooster MJ (2017) Assessment of Errors Caused by Forest Vegetation Structure in Airborne LiDAR-Derived DTMs. Remote Sensing 9:1101 Sinha S, Jeganathan C, Sharma LK, Nathawat MS (2015) A review of radar remote sensing for biomass estimation. International Journal of Environmental Science and Technology 12:1779 Song Q, Albrecht CM, Xiong Z, Zhu XX (2023) Biomass Estimation and Uncertainty Quantification From Tree Height. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 16:4833 Sun G, Ranson KJ, Guo Z, Zhang Z, Montesano P, Kimes DS (2011) Forest biomass mapping from lidar and radar synergies. Remote Sensing of Environment 115:2906 Sundquist ET, Ackerman KV, Bliss N, Kellndorfer JM, Reeves M, Rollins M (2016) Rapid Assessment of U.S. Forest and Soil Organic Carbon Storage and Forest Biomass Carbon Sequestration Capacity Tian L, Wu X, Yu T, Li M, Qian C, Liao L, Fu W (2023) Review of Remote Sensing-Based Methods for Forest Aboveground Biomass Estimation: Progress, Challenges, and Prospects. Forests 14:1086 Tsui OW, Coops NC, Wulder MA, Marshall P (2013) Integrating airborne LiDAR and space-borne radar via multivariate kriging to estimate above-ground biomass. Remote Sensing of Environment 139:340 White JC, Tompalski P, Vastaranta M, Wulder MA, Saarinen N, Stepper C, Coops NC (2017) A model development and application guide for generating an enhanced forest inventory using airborne laser scanning data and an area-based approach Yadava RN, Sinha B (2020) Vulnerability Assessment of Forest Fringe Villages of Madhya Pradesh, India for Planning Adaptation Strategies. Sustainability 12:1253 Zolkos S, Goetz SJ, Dubayah R (2012) A meta-analysis of terrestrial aboveground biomass estimation using lidar remote sensing. Remote Sensing of Environment 128:289 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 19 Jul, 2025 Read the published version in Modeling Earth Systems and Environment → Version 1 posted Editorial decision: Revision requested 21 May, 2025 Reviews received at journal 16 May, 2025 Reviews received at journal 13 May, 2025 Reviews received at journal 23 Apr, 2025 Reviewers agreed at journal 01 Apr, 2025 Reviewers agreed at journal 30 Mar, 2025 Reviewers agreed at journal 29 Mar, 2025 Reviewers agreed at journal 29 Mar, 2025 Reviewers agreed at journal 29 Mar, 2025 Reviewers agreed at journal 29 Mar, 2025 Reviewers invited by journal 29 Mar, 2025 Editor assigned by journal 26 Mar, 2025 Submission checks completed at journal 26 Mar, 2025 First submitted to journal 20 Mar, 2025 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-6270236","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":432531810,"identity":"003cae1c-4e8b-45f3-890b-0d3e5f4a93a4","order_by":0,"name":"Pradeep Kumar Rajput","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYDACHijNDyISCkjRItkA0mJAihaDA2CSCB3mPGcff/jYZpNnfH514ocHBgzy/GIH8Gux7G03k5zZllZsduPtZgmgwwxnzk7Ar8XgPBsbM8+Zw4nbbpzdANKSYHCbsBbmz3/O/E/cPOPs5h/EaTnbxiDNUHEgcQN/7zbibLHsOcYm2VORnDjjBu82iwQDCcJ+MedJY/7ww8Ausb//7OabPyps5PmlCTkMzpIAq5TArxxVC/8BwqpHwSgYBaNgZAIA/9dEv7wCFjcAAAAASUVORK5CYII=","orcid":"","institution":"CSJMU Kanpur","correspondingAuthor":true,"prefix":"","firstName":"Pradeep","middleName":"Kumar","lastName":"Rajput","suffix":""}],"badges":[],"createdAt":"2025-03-20 13:53:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6270236/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6270236/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s40808-025-02527-4","type":"published","date":"2025-07-19T16:05:45+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":79082951,"identity":"a870324b-c235-46c6-94a0-5dd4a1fe3307","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":406012,"visible":true,"origin":"","legend":"\u003cp\u003eStudy area location Map\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/94174768d6fb62c046f88800.png"},{"id":79082950,"identity":"d03389e8-2a8c-4d3f-9d38-2a8986887110","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":154505,"visible":true,"origin":"","legend":"\u003cp\u003eData inventory of forest plots with species scientific name\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/a125eb6c923e651a6a1d377e.png"},{"id":79082952,"identity":"81fd142c-a01c-430b-9b4d-8a23088f3ee7","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":507641,"visible":true,"origin":"","legend":"\u003cp\u003eMethodology flow chart\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/739cc58d96219f38322b170b.png"},{"id":79083462,"identity":"c72eeb31-3240-41e8-8aad-c8e6e9c4610a","added_by":"auto","created_at":"2025-03-24 08:52:25","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":172989,"visible":true,"origin":"","legend":"\u003cp\u003eModel comparison between height and Diameter of tree\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/05ef1b1fb96bc3e218925198.png"},{"id":79083464,"identity":"5e96b102-32d9-4225-90f5-6e55cc32e36e","added_by":"auto","created_at":"2025-03-24 08:52:25","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":44769,"visible":true,"origin":"","legend":"\u003cp\u003eForest inventory plots are AA (Plot 1), AB (Plot 2), AC (Plot 3), AD (Plot 4), AE (Plot 5), AF (Plot 6), AG (Plot 7), AH (Plot 8), AI (Plot 9), and AJ (Plot 10)\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/9be79688374382cf8fcd0dc9.png"},{"id":79082960,"identity":"5f320b41-c548-49dd-8d7b-6b5636073028","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":80767,"visible":true,"origin":"","legend":"\u003cp\u003ecorrelation matrix used to compute correlation for several pairs of variables\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/1d8211100279a59912d7b0ff.png"},{"id":79084371,"identity":"daf2cc9c-9ed5-4b03-a2a6-9a6fa4432707","added_by":"auto","created_at":"2025-03-24 09:00:25","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":224992,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation matrix between two variable heights and DBH\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/d4ad1c4376cec1f365d10507.png"},{"id":79082970,"identity":"e7d9b219-aad5-48ac-9de0-47fd627f72ac","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":84782,"visible":true,"origin":"","legend":"\u003cp\u003eScatterplot of forest biomass between two variable\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/d1b0120c4a93d302a622cae0.png"},{"id":79082957,"identity":"e660e9d0-1541-459a-9f77-ef0c84cb31d2","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":310251,"visible":true,"origin":"","legend":"\u003cp\u003ecorrelation between multi variable of environmental parameter for forest biomass stimulation.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/02dcf2f2c85547aa2231c1cf.png"},{"id":79082963,"identity":"4c9ca177-6e30-4d8c-ba7e-7511beb8fa42","added_by":"auto","created_at":"2025-03-24 08:44:25","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":197288,"visible":true,"origin":"","legend":"\u003cp\u003eAbove forest biomass predicted vs Actual by model\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/369ec208be0f7fe642500db8.png"},{"id":79083467,"identity":"d9dd015e-daa8-445d-9b58-bb15e11de867","added_by":"auto","created_at":"2025-03-24 08:52:25","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":57810,"visible":true,"origin":"","legend":"\u003cp\u003eAbove Ground Biomass predicted vs Actual\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/13b7f534f491935ce2c5cc64.png"},{"id":79082981,"identity":"4d957d98-f52f-4dd0-bea0-e1e4167a92b0","added_by":"auto","created_at":"2025-03-24 08:44:26","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":63566,"visible":true,"origin":"","legend":"\u003cp\u003eVariable Importance Above Ground Biomass\u003c/p\u003e","description":"","filename":"12.png","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/e9d4126a4501581b5b07ed3e.png"},{"id":88506954,"identity":"8740f20c-0f33-41cc-98b3-18ea6bee9fc1","added_by":"auto","created_at":"2025-08-07 07:35:54","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3172168,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6270236/v1/5e410da0-2b29-4426-859b-902bd239972d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine learning approach for Forest Biomass Modelling with In-Situ and Remote Sensing Data in Narmadapuram central India","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe estimation of forest biomass is critical for understanding carbon sequestration, biodiversity, and ecosystem health. Accurate and efficient methods for quantifying forest biomass have been a focal point of research, as these estimates are essential for informed forest management and climate change mitigation efforts (Radtke et al. 2016). This review explores the extensive body of literature related to forest biomass modeling using a combination of in-situ and remote sensing data (Mohren et al. 2012). We delve into the historical development, methodologies, challenges, and recent advancements in this field, highlighting key studies that have contributed to our understanding of forest biomass estimation (Lu et al. 2014). The study of forest biomass estimation dates back several decades, but it has gained increasing importance in the context of climate change mitigation (Brown 1997). Early methods primarily relied on ground-based measurements and allometric equations that related tree variables to biomass. Notable work introduced a widely-used generalized allometric equation for estimating aboveground biomass in forests (Radtke et al. 2016). However, the limited spatial coverage of ground-based data prompted the integration of remote sensing techniques. Remote sensing technologies have revolutionized the field of forest biomass estimation (Kumar et al. 2015). The utilization of satellite and airborne sensors has allowed for the collection of large-scale forest data, making it possible to assess biomass across broad geographic regions (Lu et al. 2014). Demonstrated the potential of Synthetic Aperture Radar (SAR) data for mapping forest biomass (Minh et al. 2016). Their work laid the foundation for subsequent research that leveraged SAR imagery, which is less sensitive to cloud cover and suitable for tropical forests (Ranson et al. 1997). In-situ data, including forest inventory plots and field measurements, remain indispensable for calibrating and validating remote sensing-based biomass models (Sinha et al. 2015). emphasized the importance of plot-based data in estimating forest biomass, highlighting the variability of biomass estimates within different forest types (Schepaschenko et al. 2019). Ground data also aid in addressing issues related to sensor-specific biases, which are critical in refining remote sensing models. Several techniques have emerged for combining in-situ and remote sensing data to estimate forest biomass (Lu 2006). Among these, the integration of optical and LiDAR (Light Detection and Ranging) data has gained considerable attention (Pizaña et al. 2016). Demonstrated the utility of combining optical data from the Landsat satellite series with airborne LiDAR to estimate biomass across large areas (Brovkina et al. 2016). The synergy of these two data sources provides both spatial and vertical information, resulting in more accurate biomass estimates. Despite significant advancements, challenges persist in forest biomass modeling (Sun et al. 2011). Among the primary issues are the spatial and temporal resolution of remote sensing data, which can limit the precision of biomass estimates, especially in areas with rapid land cover change (Knott et al. 2023). Additionally, the accuracy of in-situ data collection is influenced by sampling design, and efforts are ongoing to establish standardized protocols for forest inventory plots (Zolkos et al. 2012). Moreover, model transferability across different forest types and regions remains a challenge, as models trained in one area may not perform well in another due to variations in tree species and environmental conditions (Kilham et al. 2018). Recent research has focused on machine learning approaches, such as Random Forest and Neural Networks, for improved biomass estimation (Gao et al. 2018). used a combination of field data, LiDAR, and satellite data to create a machine learning-based biomass map, achieving higher accuracy compared to traditional methods (Baccini et al. 2004). Furthermore, the integration of multi-temporal remote sensing data has enabled the monitoring of biomass change over time, contributing to our understanding of forest dynamics and carbon fluxes (Tian et al. 2023). Forest biomass estimation is directly relevant to climate change mitigation strategies. Accurate biomass assessments underpin the calculation of carbon stocks in forests, which is essential for carbon offset programs and international agreements like REDD+ (Reducing Emissions from Deforestation and Forest Degradation)(Mo et al. 2023). Research by Researcher demonstrated how remote sensing can be used to estimate carbon stocks in tropical forests, emphasizing the role of accurate biomass data in supporting climate change initiatives(Saatchi et al. 2007). In summary, forest biomass modeling using a combination of in-situ and remote sensing data has witnessed significant developments over the years. The integration of optical, LiDAR, and SAR data, along with machine learning techniques, has improved the accuracy and applicability of biomass estimates (Kumar et al. 2015). Nevertheless, challenges such as data resolution, model transferability, and the need for standardized in-situ data collection persist. As efforts to combat climate change intensify, the accurate quantification of forest biomass remains critical, and further research will likely refine existing techniques and develop new methods for better understanding forest ecosystems and their role in global carbon dynamics.\u003c/p\u003e\n\u003cp\u003eForest ecosystems play a pivotal role in maintaining ecological balance, serving as critical carbon sinks, and mitigating the impacts of climate change. Forest biomass, which refers to the total mass of living biological organisms within a given forest area, is a fundamental metric for understanding forest productivity, carbon sequestration, and overall ecosystem health (Gleason and Im 2011). Accurate estimation of forest biomass is essential for a variety of applications, including forest management, climate modeling, and carbon accounting under international frameworks such as the United Nations Framework Convention on Climate Change (UNFCCC) (Sundquist et al. 2016). In recent years, the integration of in-situ measurements and remote sensing technologies has revolutionized the field of forest biomass estimation. Traditional methods of biomass estimation, which rely heavily on destructive sampling and allometric equations, are often labour-intensive, time-consuming, and spatially limited (Abbas et al. 2020). Remote sensing, on the other hand, provides a means to estimate biomass over large spatial extents with high temporal and spatial resolution, making it an indispensable tool in forest research and management (Huang et al. 2015).\u003c/p\u003e\n\u003cp\u003eThe combination of in-situ and remote sensing data addresses the limitations of each approach and enhances the accuracy of biomass modeling. In-situ data, such as tree diameter, height, and wood density, provide ground truth measurements that are critical for calibrating and validating remote sensing models (Pizaña et al. 2016). Remote sensing platforms, including satellite imagery, airborne LiDAR, and radar sensors, offer the ability to capture spatially explicit information on forest structure and composition, which can be correlated with biomass metrics (Næsset et al. 2011). This introduction explores the significance of forest biomass modeling, the advances in remote sensing technologies, and the synergistic potential of combining in-situ and remote sensing data. It also highlights the key challenges and research gaps in the field, paving the way for more accurate and scalable biomass estimation methodologies (Song et al. 2023). Forest biomass estimation is a cornerstone of global efforts to combat climate change. Forests act as carbon reservoirs, sequestering carbon dioxide from the atmosphere and storing it in vegetation and soil (Duncanson et al. 2019). The quantification of biomass is crucial for assessing carbon stocks, understanding forest dynamics, and evaluating the impact of land-use changes on carbon emissions (Shi and Liu 2017). Accurate biomass estimates are also essential for monitoring deforestation and forest degradation, which account for approximately 10–15% of global greenhouse gas emissions (Mohren et al. 2012). In the context of climate change mitigation, forest biomass data are integral to the implementation of Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiatives (Griscom et al. 2017). These programs incentivize developing countries to conserve forests by providing financial rewards for verified reductions in carbon emissions. However, the success of such programs depends on the availability of reliable and consistent biomass data, which underscores the importance of robust modeling techniques (Haya et al. 2023). Remote sensing technologies have advanced significantly over the past few decades, offering unprecedented capabilities for forest biomass estimation. Optical remote sensing platforms, such as Landsat and Sentinel-2, provide high-resolution imagery that can be used to derive vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), which are proxies for biomass (Eisfelder et al. 2011). Radar sensors, such as those on the Sentinel-1 and ALOS PALSAR satellites, are particularly effective in capturing forest structure and canopy characteristics due to their ability to penetrate vegetation (Lu 2006). LiDAR (Light Detection and Ranging) technology represents a breakthrough in forest biomass modeling, offering detailed three-dimensional information on forest structure (White et al. 2017). Airborne and terrestrial LiDAR systems can measure tree height, canopy density, and vertical forest structure with high precision, making them invaluable for biomass estimation (Simpson et al. 2017). Additionally, the launch of space borne LiDAR missions, such as the Global Ecosystem Dynamics Investigation (GEDI), has expanded the reach of this technology to a global scale (Kellner et al. 2019).The integration of in-situ and remote sensing data represents a paradigm shift in forest biomass modeling. Ground-based measurements provide accurate and detailed information on individual trees, which are essential for developing and validating allometric equations (Laurin et al. 2016). These equations relate easily measurable parameters, such as diameter at breast height (DBH) and tree height, to biomass. However, their applicability is often limited to specific regions, forest types, or species. Remote sensing data address these limitations by providing spatially continuous coverage and capturing forest heterogeneity at multiple scales (Georgopoulos et al. 2023). For example, LiDAR-derived metrics, such as canopy height models (CHMs) and gap fraction, can be combined with in-situ measurements to improve the accuracy and scalability of biomass estimates (Tsui et al. 2013). Similarly, radar backscatter data, which are sensitive to forest density and moisture content, can be calibrated using ground-based measurements to estimate aboveground biomass (Kaasalainen et al. 2015). The use of machine learning algorithms has further enhanced the integration of in-situ and remote sensing data. Techniques such as random forests, support vector machines, and neural networks are increasingly being used to model complex relationships between remote sensing variables and biomass (Cui 2019). These approaches allow for the incorporation of multiple data sources, including spectral indices, LiDAR metrics, and topographic variables, resulting in more robust and accurate models (Kong et al. 2018). Despite significant advancements, several challenges remain in forest biomass modeling. One major issue is the uncertainty associated with remote sensing measurements, which can arise from sensor limitations, atmospheric conditions, or data processing errors (Chen et al. 2016). Additionally, the scalability of allometric equations and remote sensing models is often limited by the lack of ground truth data across diverse forest types and ecological conditions (Pittman et al. 2015). Another challenge is the temporal variability of forest biomass, which is influenced by factors such as seasonal changes, disturbances, and forest growth dynamics. Capturing this variability requires high-frequency remote sensing observations and robust temporal modeling techniques (Naik et al. 2021). Furthermore, the integration of belowground biomass, which accounts for a significant portion of total forest carbon stocks, remains an underexplored area in biomass modelling. Forest biomass modelling is a critical component of efforts to understand and mitigate climate change. The integration of in-situ measurements and remote sensing data offers a powerful approach to overcome the limitations of traditional methods and achieve accurate, scalable, and spatially explicit biomass estimates. Advances in remote sensing technologies, coupled with the use of machine learning algorithms, have opened new avenues for research and applications in this field. However, addressing the challenges of uncertainty, scalability, and temporal variability requires continued innovation and collaboration among researchers, practitioners, and policymakers. By leveraging the strengths of both in-situ and remote sensing approaches, forest biomass modeling can provide the data and insights needed to support sustainable forest management, carbon accounting, and climate change mitigation efforts worldwide.\u003c/p\u003e\n\u003ch3\u003eStudy area description\u003c/h3\u003e\n\u003cp\u003eNarmadapuram district (Fig.\u0026nbsp;1), situated in the central Indian state of Madhya Pradesh (Latitude: 22.75° N, Longitude: 77.72° E) boasts a rich natural landscape characterized by dense forests, diverse flora, and abundant wildlife (Yadava and Sinha 2020). This region serves as a significant hub for studying forest biomass, offering a plethora of research opportunities and insights into the dynamics of ecosystem functioning (Rajput et al. 2021). The district's topography comprises undulating terrain, interspersed with lush forests, which include both natural and man-made plantations. These forests play a vital role in regulating the local climate, conserving soil fertility, and providing a habitat for numerous species of plants and animals. Research conducted in Narmadapuram district often focuses on quantifying forest biomass, examining carbon sequestration rates, and assessing the impact of anthropogenic activities on forest health. Studies utilize various methodologies, including remote sensing techniques, field surveys, and biomass modeling, to gain a comprehensive understanding of the region's forest ecosystems. Moreover, initiatives such as afforestation programs and community-based conservation efforts further contribute to the sustainable management of forest resources in Narmadapuram district. Assessment of Forest Biomass in Narmadapuram District, Madhya Pradesh, India. provide valuable insights into the dynamics of biomass accumulation and carbon storage in this ecologically significant area, informing conservation strategies and policy interventions aimed at preserving the region's natural heritage (Ahirwar et al. 2020).\u003c/p\u003e\n\u003cdiv id=\"Sec3\"\u003e\n \u003ch2\u003eData Inventory\u003c/h2\u003e\n \u003cp\u003e(Fig.\u0026nbsp;2) represent the image presents a boxplot depicting the distribution of the number of trees across various forest plots labeled AA to AJ. Each plot corresponds to a different species, represented by their scientific names, and each species is assigned a specific color for easy identification. The boxplots illustrate the frequency distribution of tree counts, with the boxes representing the interquartile range (IQR) and the whiskers indicating the range within which the tree count falls, while black dots signify outliers. The plot reveals significant variation in tree counts across different plots. Some species, like \u003cem\u003eAcacia catechu\u003c/em\u003e, \u003cem\u003eBauhinia racemosa\u003c/em\u003e, and \u003cem\u003eTerminalia arjuna\u003c/em\u003e, show more clustered values, indicating higher tree frequencies, while others exhibit greater dispersion, with boxes showing more spread-out data. Certain plots have outliers, marked by the black dots, which represent unusually high or low numbers of trees compared to the general distribution. Additionally, there are differences in tree abundance across the plots; for example, plots like AC and AF have relatively higher tree counts, while others like AE and AJ display lower or more variable distributions. This variation could be attributed to differences in plot size, forest density, or environmental conditions affecting the growth of each species. Overall, the boxplot provides a clear visual representation of tree population distributions across the forest plots, highlighting species diversity, tree count variation, and the factors influencing these patterns, which could be useful for forest management, biodiversity studies, and ecological assessments.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Methods and Materials","content":"\u003cp\u003e\u003cstrong\u003eExperiment design and Field data\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe biomass study was conducted across multiple 100m x 100m plots within the Narmadapuram district, Madhya Pradesh, to assess spatial distribution and density. Ten well-defined plots were demarcated with latitude and longitude coordinates for each corner point, ensuring uniform sampling and precise measurements. Biomass data varied across plots, reflecting different vegetation densities. Plot AA recorded 46.34 t/ha, Plot AB had 67.88 t/ha, while Plot AC measured 90.45 t/ha. Plot AD showed a biomass of 61.22 t/ha, and Plot AE had a significant value of 107.31 t/ha. Other plots, such as Plot AF and Plot AG, recorded 101.28 t/ha and 55.11 t/ha, respectively. Plot AH recorded 71.10 t/ha, while Plot AI had the highest biomass value of 294.58 t/ha, indicating dense vegetation. Plot AJ recorded 73.56 t/ha. The total biomass for all plots summed to 968.86 t/ha. This systematic grid-based approach facilitated a comprehensive understanding of biomass variation across the study area, highlighting its importance in environmental monitoring and resource management (Fig. 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eForest Parameters Estimation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 1 Summary statistic forest plot data\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"605\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eTree count\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eStd. Dev.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eMin\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003ePctl. 25\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003ePctl. 75\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003eMax\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AA(Plot 1)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e248\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e230\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e248\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.088\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.061\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e248\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e248\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.00087\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.044\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AB(Plot 2)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e203\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.092\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.65\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e9.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e7.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e49\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0042\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AC(Plot 3)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e390\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e306\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e390\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.099\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.089\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.97\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e390\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e112\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e390\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.00029\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AD(Plot 4)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e431\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e194\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e431\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e431\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e9.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e431\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.032\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AE(Plot 5)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e486\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e471\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e486\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.025\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e486\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e9.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e486\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.00036\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.044\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AF(Plot 6)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e102\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e230\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0029\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.069\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AG(Plot 7)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e294\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e202\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e294\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.095\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.64\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e294\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e294\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0046\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AH(Plot 8)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e366\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.097\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e8.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0032\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.065\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e5.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AI(Plot 9)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e887\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e3.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e5.7\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cstrong\u003esite: AJ(Plot 10)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eGBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e475\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e204\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eDBH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e475\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.073\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.061\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.65\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eTOTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e475\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e6.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e5.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e7.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003eBiomass\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e475\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.0034\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\"\u003e\n \u003cp\u003e1.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe Table 1 summary statistics highlight significant variability in forest structure across the surveyed sites (AA to AJ). The average number of Sub Plots per site ranges from 4.4 to 5.2, indicating consistent sampling density, with sites AE and AI showing slightly higher averages. Tree size, represented by GBH and DBH, varies considerably. Site AI has the largest trees, with an average GBH of 81 cm and DBH of 0.26 m, while site AJ has the smallest, with a GBH of 49 cm and DBH of 0.16 m. This trend is mirrored in total tree height (TOTH), where AI trees are the tallest (mean: 14 m) and AJ the shortest (mean: 6.6 m), reflecting a positive relationship between girth and height. The number of stems (No stem) remains relatively consistent across sites, averaging around 1.1, with slight increases in sites AC and AJ. Tree volume and biomass exhibit the most notable variations, emphasizing differences in forest composition. Site AI has the highest mean volume (1.0 m\u0026sup3;) and biomass (0.76 tons), indicating mature forest characteristics. In contrast, site AJ records the lowest values, with a mean volume of 0.2 m\u0026sup3; and biomass of 0.15 tons, reflecting a younger or less dense forest. The standard deviations in GBH, DBH, TOTH, and Volume reveal substantial heterogeneity, with sites like AE and AI showing the greatest variability, likely due to a mix of small and large trees. Maximum values across variables further highlight the diversity, with AI having the tallest tree (22 m) and the largest GBH (887 cm), while AJ\u0026rsquo;s metrics remain the smallest. Overall, the analysis underscores significant structural differences among sites. Mature forests like AI exhibit larger trees, higher volumes, and greater biomass, contributing more to carbon storage, while younger forests like AJ showcase smaller trees and lower biomass. These findings provide a comprehensive understanding of forestdynamics and are critical for resource management and carbon stock assessment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRemote sensing and Data processing\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSAR Sentinel Data processing\u003c/strong\u003e: SAR (Synthetic Aperture Radar) data processing is a vital component of remote sensing and Earth observation, providing a wealth of information for a wide range of applications, including agriculture, forestry, geology, disaster monitoring, and urban planning. This complex process involves the acquisition of raw SAR data, its conversion into meaningful information, and subsequent analysis. In this method for processing the SAR data, we will understand the various stages of SAR data processing, from data process to interpretation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Preprocessing\u003c/strong\u003e: The Sentinel-1A SAR data processing chain typically begins with preprocessing steps to improve the quality and utility of the raw imagery data. The process involves radiometric geometric calibration, terrain correction, and removal of speckle noise, which is inherent to SAR imagery. Additionally, geometric correction is performed to ensure accurate georeferencing. These preprocessing steps are essential for generating accurate and reliable results from SAR data. Raw SAR data is acquired, a line of preprocessing steps are requiring to correct for geometric and radiometric distortions. These corrections ensure that the data accurately represent the Earth\u0026apos;s surface. Geometric corrections involve accounting for the sensor\u0026apos;s position, attitude, and the topography of the observed area. Radiometric corrections adjust for variations in signal strength caused by factors like atmospheric conditions, sensor settings, and terrain roughness. Key preprocessing steps include radiometric calibration, terrain correction, and geocoding. Radiometric calibration normalizes the radar signal to an absolute scale, allowing for consistent analysis. Terrain correction compensates for topographic effects, ensuring that features at different elevations are accurately represented. Geocoding involves transforming the data into geographic coordinates, enabling easy integration with other spatial datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImage Formation and Focusing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe core principle of SAR data processing is the creation of high-resolution images through a synthetic aperture. In this process, SAR sends microwave pulses towards the Earth\u0026apos;s surface and records the echoes reflected back. These echoes are then used to construct a high-resolution image of the area, with finer details obtained through the synthetic aperture\u0026apos;s motion. Focusing techniques, including range compression and azimuth compression, are applied to sharpen the image and reduce blurring caused by satellite motion.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOptical Data\u003c/strong\u003e: sentinel 2 A data used for extraction of vegetation index, such as NDVI, SVAI and other vegetation index. sentinel data has 10 m spatial resolution. Sentinel-2A optical data is a valuable resource for extracting various vegetation indices crucial for environmental monitoring and agricultural analysis. With its 10-meter spatial resolution, Sentinel-2A imagery provides detailed insights into land cover dynamics and vegetation health. One of the most widely used vegetation indices derived from Sentinel-2A data is the Normalized Difference Vegetation Index (NDVI), which quantifies the presence and vigor of vegetation based on the contrast between near-infrared and red reflectance. Additionally, Sentinel-2A data enables the calculation of other vegetation indices such as the Soil-adjusted Vegetation Index (SAVI), which accounts for soil brightness in vegetation assessments, and the Enhanced Vegetation Index (EVI), offering improved sensitivity in high biomass regions. These indices play a pivotal role in monitoring ecosystem health, crop growth, and assessing the impact of environmental changes on vegetation dynamics. By leveraging Sentinel-2A optical data and its suite of vegetation indices, researchers and land managers can make informed decisions for sustainable land use and resource management practices.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAGB estimation\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn Central India, where tropical deciduous forests are predominant, region-specific equations for Above Ground Biomass (AGB) estimation are often developed based on field data. One commonly referenced source is Forest Survey of India (FSI) or studies conducted in the Satpura, Vindhyan ranges, and other dense forest zones of Madhya Pradesh and Chhattisgarh.\u003c/p\u003e\n\u003cp\u003eAGB can be estimated using allometric equations involving tree height DBH and wood density. The allometric equations are of the power function type (eqn. 1)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eABG=0.0509*(DBH\u003csup\u003e2\u003c/sup\u003e * H * WD) \u0026nbsp;eqn. 1\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ewhere H (m) is the tree height is the DBH(cm), is the Diameter at Breast Height \u0026nbsp; and WD (g/cm\u0026sup3;) is Wood Density (species-specific) is are parameters by in situ measurements.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLinear regression:\u003c/strong\u003e Linear regression is a statistical method used to model the relationship between two or more variables by fitting a straight line to the observed data points. The goal is to find the best-fitting line that minimizes the difference between the actual data points and the predicted values generated by the line. This method is commonly used in various fields such as economics, finance, biology, and social sciences to analyze and predict the behavior of dependent variables based on one or more independent variables. By estimating the parameters of the linear equation, such as the slope and intercept, linear regression allows researchers to make predictions and infer relationships between variables in a simple and interpretable manner.Top of Form\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExponential:\u003c/strong\u003e Exponential growth or decay is a fundamental concept in mathematics and science. It refers to a process where a quantity increases or decreases at a rate proportional to its current value. In exponential growth, the rate of increase itself increases over time, leading to rapid growth, while in exponential decay, the rate of decrease decreases over time, resulting in rapid decline. The general form of an exponential function is , where: f(x) is the value of the function at a given point x, a is the initial value or the value at, b is the base of the exponential function, often referred to as the growth or decay factor, x is the exponent, representing the independent variable such as time. Exponential growth is commonly observed in various natural phenomena such as population growth, compound interest, and the spread of diseases in uncontrolled environments. It\u0026apos;s also applicable in fields like physics, chemistry, and economics. Understanding exponential growth is crucial because it can lead to significant consequences, both positive and negative, depending on the context. For instance, it can illustrate the potential of technological advancement or highlight the urgency of addressing issues like environmental sustainability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRandom forest Modeling\u003c/strong\u003e: Random forest modeling is a versatile and powerful machine learning technique that leverages the collective wisdom of multiple decision trees to make predictions. Each tree in the forest is trained on a random subset of the data and uses a random subset of the features, which helps to reduce overfitting and increase the model\u0026apos;s robustness. During prediction, the results from all the trees are aggregated, typically by averaging for regression tasks or by voting for classification tasks, to produce the final output. This ensemble approach often leads to improved accuracy and generalization compared to individual decision trees. Random forests are widely used across various domains, including finance, healthcare, and marketing, for tasks such as classification, regression, and outlier detection. Additionally, they offer built-in feature importance measures, which can help identify the most relevant variables driving the predictions. Overall, random forest modeling is a valuable tool in the machine learning toolkit, offering a balance of performance, interpretability, and ease of use.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupport Vector machine\u003c/strong\u003e: A Support Vector Machine (SVM) is a powerful supervised learning algorithm used for classification and regression tasks. Its primary objective is to find the optimal hyperplane that best separates the data points into different classes in a high-dimensional space. SVM works by identifying the support vectors, which are the data points closest to the decision boundary, and maximizing the margin between classes. This margin allows the SVM to generalize well to new, unseen data, making it robust against overfitting. Additionally, SVM can handle both linear and non-linear classification tasks through the use of different kernel functions, such as linear, polynomial, radial basis function (RBF), and sigmoid. SVMs are widely used in various fields, including image classification, text classification, and bioinformatics, due to their versatility, effectiveness, and ability to handle high-dimensional data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Performance Evaluation Metrics for Forest AGB Estimation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe linear regression (LR) model and machine learning techniques, including Random Forest (RF), Artificial Neural Network (ANN), and Support Vector Machine (SVM), were validated using 10 forest inventory plots. To evaluate and compare the performance of these models in estimating forest above-ground biomass (AGB), three statistical metrics were used: the coefficient of determination (R\u003csup\u003e2\u003c/sup\u003e), mean absolute error (MAE), and root mean square error (RMSE). These metrics are commonly employed to assess the accuracy of predictive models by measuring the differences between observed and predicted values. A better model is indicated by higher R\u003csup\u003e2\u003c/sup\u003e values and lower RMSE and MAE values. The formulas for calculating R\u003csup\u003e2\u003c/sup\u003e, RMSE, and MAE are as follows:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1. Coefficient of Determination (R\u0026sup2;)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe coefficient of determination evaluates how well the model explains the variability in the dependent variable.\u003c/p\u003e\n\u003cp\u003eR\u0026sup2; = 1 - [\u0026Sigma; (yᵢ - ŷᵢ) \u0026sup2; / \u0026Sigma; (yᵢ - ȳ) \u0026sup2;] \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhere yᵢ = observed value, ŷᵢ = predicted value, ȳ = mean of observed values, n = number of observations\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. Root Mean Square Error (RMSE)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe RMSE measures the average magnitude of the error between predicted and observed values.\u003c/p\u003e\n\u003cp\u003eRMSE = \u0026radic; [1/n * \u0026Sigma; (yᵢ - ŷᵢ) \u0026sup2;] \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhere yᵢ = observed value, ŷᵢ = predicted value, and n = number of observations\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3. Mean Absolute Error (MAE)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe MAE calculates the average absolute difference between predicted and observed values.\u003c/p\u003e\n\u003cp\u003eMAE = [1/n * \u0026Sigma;|yᵢ - ŷᵢ|]\u003c/p\u003e\n\u003cp\u003eWhere yᵢ = observed value, ŷᵢ = predicted value and n = number of observations\u003c/p\u003e"},{"header":"Results and Discussion","content":"\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"660\" height=\"168\"\u003e\u003c/p\u003e\n\u003cp\u003eThe Table 2 compares four modeling methods Log1, Log2, Weibull, and Michaelis based on their performance metrics: Residual Standard Error (RSE), RSE log, and Average Bias. Log1 (blue) demonstrates relatively low RSE (4.067850) and RSE log (0.3157938), along with a minimal Average Bias of -0.002102274, indicating good overall performance with minimal systematic deviation. Similarly, Log2 (green) slightly outperforms Log1 with the lowest RSE (4.019404) and RSE log (0.3133169) while maintaining a small Average Bias of -0.003154677. This suggests that Log2 provides the most accurate and reliable predictions among the methods compared. In contrast, the Weibull method (orange) exhibits the highest RSE (4.674796) and the largest positive Average Bias (0.242635165), indicating poor performance with significant overestimation. Additionally, the RSE log metric is unavailable for Weibull, suggesting potential limitations in its compatibility with logarithmic transformations. The Michaelis method (purple) performs slightly better than Weibull, with an RSE of 4.476368 and a positive Average Bias of 0.206571212. However, its error and bias levels remain higher than those of Log\u003csup\u003e1\u003c/sup\u003e and Log\u003csup\u003e2\u003c/sup\u003e, and its RSE log metric is also unavailable. In conclusion, Log\u003csup\u003e2\u0026nbsp;\u003c/sup\u003eemerges as the best-performing method due to its low error metrics and minimal bias, making it the most suitable for accurate and unbiased predictions. Log1 is a close second, while the Weibull and Michaelis methods are less reliable due to their higher errors and significant biases. These findings emphasize the importance of selecting modeling approaches with low RSE and minimal bias for robust predictive performance (Fig. 4).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eForest Inventory Data\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe (Fig. 5) represents a bar chart illustrating the tree count distribution across ten forest inventory plots, labeled as AA, AB, AC, AD, AE, AF, AG, AH, AI, and AJ, corresponding to plots 1 through 10. The x-axis represents different forest plots, while the y-axis quantifies the tree count in each plot. The purpose of this figure is to visualize the variation in tree density among different locations. The bar chart employs vertical bars to represent the tree count within each forest plot, with the height of each bar correlating to the number of trees recorded in the respective plot. This visual representation allows for an easy comparison of tree density. The x-axis is labeled as Forest Plots, indicating different sampling locations within the forest, while the y-axis is labelled as Tree Count, representing the number of trees recorded in each plot. The tree count scale ranges from 0 to 500, providing a clear measure of tree density. The bars are shaded in a uniform dark gray color, ensuring clarity and ease of interpretation, while the background grid lines aid in estimating the approximate tree count in each plot.The tree count varies across the ten forest plots, with some plots exhibiting higher densities while others have relatively lower tree counts. Some plots, such as AA and AB, show a relatively low tree count, with values around 250 trees, whereas AG and AH also exhibit lower tree counts, suggesting sparse tree coverage in these areas. Moderate tree count plots include AF, which is slightly above the low-density group but still falls behind the highest-density plots, while AC and AI have intermediate tree densities with moderate variations in tree count. The highest tree counts are observed in AD and AE, with AE reaching the maximum value of approximately 500 trees. AJ also exhibits a high tree count, comparable to AD.Several ecological and environmental factors could contribute to the variation in tree counts across different plots. Soil quality and nutrient availability play a crucial role in tree density, as nutrient-rich soil supports higher tree counts, whereas poor soil conditions may lead to sparse tree coverage. Topographical factors, such as elevation, slope, and aspect, also influence tree growth. For instance, steep slopes may have fewer trees due to soil erosion and water runoff. Additionally, microclimatic conditions like sunlight exposure, temperature, and moisture levels can significantly impact tree density. Forest plots with better moisture retention may support higher tree counts. Human or natural disturbances, including logging activities, land clearance, fire incidents, storms, pest infestations, or disease outbreaks, could also lead to reduced tree densities in certain areas. Furthermore, species composition and growth patterns vary, with some tree species growing in dense clusters while others are sparsely distributed, potentially explaining the differences in tree counts. Understanding tree distribution across forest plots has important ecological implications. Forest plots with higher tree densities may serve as crucial habitats for wildlife, contributing to biodiversity conservation. On the other hand, low-density plots may require conservation efforts such as afforestation and protection from further degradation. High-density plots should be monitored to maintain ecological balance and prevent resource competition among trees. Additionally, forest plots with higher tree densities play a significant role in carbon sequestration, helping mitigate climate change by absorbing atmospheric carbon dioxide. In conclusion, the bar chart effectively illustrates the variation in tree counts across ten forest inventory plots, revealing significant differences in tree density that could be attributed to multiple ecological and environmental factors. Such information is crucial for forest management, conservation planning, and ecological studies. Understanding these patterns helps in making informed decisions regarding sustainable forest resource management and biodiversity conservation.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eAssociation between field data of tree\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe analysis (Fig. 6) of the correlation matrix and pair plot reveals important relationships between key tree characteristics, offering insights into their ecological and practical implications. The correlation between DBH and Volume (0.57) indicates a moderate positive relationship, suggesting that as tree diameter increases, so does its volume, although other factors like tree height, wood density, and shape also play a role. Similarly, DBH shows a moderate correlation with Biomass (0.54), reinforcing its importance in studies of carbon storage and forest inventory. These trends suggest that while DBH is a significant predictor of both Volume and Biomass, it is not the sole determinant. In contrast, the correlations between tree height (TOTH) and Volume (0.30) and Biomass (0.29) are weak, highlighting that height alone does not strongly predict tree volume or biomass. This may be due to variations in diameter, crown shape, branch density, and wood density among trees of similar height. The strongest relationship is observed between Volume and Biomass (0.98), indicating they are almost directly proportional and confirming that Volume is a reliable predictor of Biomass in this dataset. This strong correlation has practical significance for forest inventories, as Biomass can be estimated accurately from Volume, reducing the need for direct measurements. The findings also emphasize the utility of DBH in forest management, though its moderate correlations suggest that combining it with other metrics like height and wood density can improve predictive accuracy. Meanwhile, the limited role of tree height as an independent predictor of Biomass or Volume underscores the need for comprehensive measurements in ecological models. The analysis highlights that correlation does not imply causation and the observed relationships may vary across forest types, species, and environmental conditions. These findings underscore the complexity of tree structure and the importance of using a combination of measurements to accurately assess tree characteristics. Understanding these relationships aids in informed decision-making for forest conservation, resource management, and carbon sequestration, ultimately supporting more sustainable forestry practices.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePlot wise correlation between height and DBH\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scatter plots depict the relationship between tree height and DBH (Diameter at Breast Height) across ten forest plots (Plot1 to Plot10), with each plot featuring a regression line and confidence intervals (Fig. 7). Most plots exhibit a positive trend, indicating that tree height generally increases with DBH, which aligns with expectations since larger diameters are often associated with taller trees. However, the strength of this correlation varies across the plots. Some, such as Plot1, Plot7, and Plot8, display tighter clusters around the regression line, suggesting a stronger relationship, while others, like Plot3 and Plot4, have more scattered data points, reflecting weaker correlations or greater variability. The slopes of the regression lines also differ, with steeper slopes (e.g., in Plot1 and Plot7) indicating a rapid increase in height with DBH, while flatter slopes (e.g., in Plot3 and Plot4) suggest a slower increase. Additionally, the range of DBH and height values varies across plots, with Plot2 covering a wider DBH range and Plot3 focusing on smaller values. Outliers are evident in several plots, with some trees exhibiting unusually tall heights for small DBH or vice versa, potentially due to measurement errors, unique species traits, or atypical growth conditions. Confidence intervals around the regression lines indicate the level of uncertainty, with wider intervals in plots like Plot3 suggesting less confidence in the model fit due to sparse or variable data. These variations highlight the influence of site-specific factors, such as species composition, soil fertility, and climate, on the height-DBH relationship. This underscores the need for localized data when developing allometric equations or estimating tree characteristics and biomass for forest management or research purposes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship Between Tree Height, DBH, and Biomass in Forest Inventory Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scatter plot illustrates the relationship between tree height (in meters) and DBH (Diameter at Breast Height, in meters), with bubble size and color representing biomass (Fig. 8). The majority of the data points are concentrated near the lower range of both DBH and height, indicating that most trees in the dataset are relatively small in size. A few outliers are evident, with some trees exhibiting extremely tall heights (over 90 meters) despite having small DBH values, as well as a few trees with unusually large DBH values (around 2 meters). The size and color of the bubbles provide insight into biomass distribution. Larger and darker bubbles, indicating higher biomass, are generally associated with higher DBH values. This is consistent with the expectation that larger trees tend to have greater biomass. However, there are also instances where trees with moderate DBH and height exhibit significant biomass, suggesting variability in wood density or tree form. Overall, the plot highlights the positive relationship between DBH and biomass while showcasing variability in the relationship between height and biomass. The outliers and variability might be attributed to species differences, site-specific factors, or measurement errors. This visualization underscores the complexity of predicting biomass based on structural parameters and the importance of considering additional variables, such as wood density or crown dimensions, in ecological studies and forest management.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMulti variable of environmental parameter for Biomass estimation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe image displays a correlation matrix that visualizes the relationships between various remote sensing variables and landscape characteristics using a color-coded heat map (Fig. 9). The correlations range from -1 to +1, indicating negative to positive associations. Strong positive correlations are observed among variables such as Sigma0_VH, Sigma0_VH_GLCM Variance, Sigma0_VH_GLCM Mean and Sigma0_VH_GLCM Correlation, with values exceeding 0.8. These relationships suggest that these variables are closely related and likely follow similar patterns, particularly in terms of radar backscatter intensity. Moderate positive correlations are seen between variables like Ratio and Sigma0_VV_Contrast (0.69), as well as between elevation and aspect (0.65), indicating a shared association but with less strength. In contrast, weak or negative correlations are noted with variables such as savi and ndvi which show minimal influence from other landscape metrics, especially radar-based variables. For example, the correlation between savi and other metrics is below 0.2, suggesting that vegetation indices like SAVI and NDVI are less dependent on factors like surface roughness or moisture content (Table 3). Additionally, the variable Diff shows a moderate positive correlation with Sigma0_VH (0.55), while other relationships, such as between slope and elevation, are weak or negative. The heatmap\u0026apos;s color gradient, ranging from blue (positive correlations) to red (negative correlations), provides an intuitive way to assess the strength and direction of relationships between the variables. Hence, strong correlations between radar-based variables indicate they might be used interchangeably in models, while weak correlations with vegetation indices suggest these should be treated separately in analyses of environmental features.\u003c/p\u003e\n\u003cp\u003eTable 3 Predictor Variable for Biomass estimation\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003ePredictor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eDescription\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eFormula\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eRadar backscatter intensity in the vertical-horizontal polarization. Indicates surface texture and structure.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u0026sigma;0,VH=10\u0026times;log10(RVH)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VH_GLCMVariance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMeasures the variance in the Gray Level Co-occurrence Matrix (GLCM), highlighting texture heterogeneity in images.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eGLCM Variance = \u0026Sigma;i,j p(i,j) * (i - \u0026mu;)^2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VH_GLCMMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eThe mean value from the Gray Level Co-occurrence Matrix, indicating average intensity.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eGLCM Mean = \u0026Sigma;i,j p(i,j) * (i + j)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VH_GLCMCorrelation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMeasures the correlation between pixel pairs in the GLCM, reflecting spatial relationship between neighboring pixels.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eGLCM Correlation = \u0026Sigma;i,j p(i,j) * (i - \u0026mu;_x) * (j - \u0026mu;_y) / (\u0026sigma;_x \u0026sigma;_y)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eRadar backscatter intensity in vertical-vertical polarization. Indicates surface roughness and structure.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u0026sigma;0_VV = 10 * log10(R_VV)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eSigma0_VV_Contrast\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eContrast in radar backscatter intensity, revealing the roughness or texture of the surface.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eContrast = \u0026Sigma;i,j p(i,j) * |i - j|\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eRatio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eRatio of the backscatter in different polarization channels, useful for surface characterization.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eRatio = \u0026sigma;0_VH / \u0026sigma;0_VV\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003easpect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMeasures the direction of the slope of the terrain, indicating the orientation of the landscape.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eAspect = atan2(\u0026part;Z/\u0026part;y, \u0026part;Z/\u0026part;x)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eslope\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMeasures the steepness of the terrain, calculated from the elevation data.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eSlope = arctan(\u0026Delta;Z / \u0026Delta;distance)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eelevation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eThe height of the terrain or surface above a reference point, typically sea level.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eElevation = Height above sea level\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003emulti\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMultiplier factor, often used in scaling or transforming certain measurements in data analysis.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eTypically a scaling constant depending on the dataset or model.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eContrast\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eMeasures differences in pixel intensities in the image, related to surface texture or heterogeneity.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eContrast = \u0026Sigma;i,j p(i,j) * |i - j|\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003esavi\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eSoil-Adjusted Vegetation Index (SAVI), a vegetation index that accounts for soil background.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eSAVI = (NIR - RED) * (1 + L) / (NIR + RED + L)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003endvi\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eNormalized Difference Vegetation Index (NDVI), a common vegetation index used to detect vegetation health.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eNDVI = (NIR - RED) / (NIR + RED)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 160px;\"\u003e\n \u003cp\u003eDiff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 253px;\"\u003e\n \u003cp\u003eDifference between two or more variables or features, often used for change detection or comparative analysis.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 192px;\"\u003e\n \u003cp\u003eDiff = X - Y\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eEstimation of AboveGround Biomass\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 4 Model for Aboveground Biomass Prediction\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 178px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMethod \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSE \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMAE \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003erelRMSE \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003er\u003csup\u003e2\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 178px;\"\u003e\n \u003cp\u003eExp Regression \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e2.26 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.97 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.1471322 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 178px;\"\u003e\n \u003cp\u003eLinear Regression \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e2.48 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e1.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.1616262\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e-0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 178px;\"\u003e\n \u003cp\u003eRandom forest\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e1.61 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.84 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.1046609 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 30px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 178px;\"\u003e\n \u003cp\u003eSVM \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e2.00 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e1.06 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.1301456 \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe Table 4 provides a comparative analysis of four different regression methods: Exponential Regression, Linear Regression, Random Forest, and Support Vector Machine (SVM). Each method\u0026apos;s performance is evaluated based on four metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), relative RMSE (relRMSE), and the coefficient of determination (r\u0026sup2;). Exponential Regression shows an RMSE of 2.26, an MAE of 0.97, a relRMSE of 0.1471322, and an r\u0026sup2; value of 0.04. These results indicate that while the model has moderate error rates, its predictive accuracy is low, as reflected by the low r\u0026sup2; value. Linear Regression has an RMSE of 2.48 and an MAE of 1.34, with a relRMSE of 0.1616262 and an r\u0026sup2; of -0.16. This method demonstrates higher error rates compared to Exponential Regression and also shows a negative r\u0026sup2; value, suggesting that it performs worse than a horizontal line model. Random Forest performs the best among the four methods with an RMSE of 1.61 and an MAE of 0.84, alongside a relRMSE of 0.1046609 and an r\u0026sup2; of 0.51. The lower error metrics and a positive r\u0026sup2; value indicate a more accurate and reliable predictive performance. Support Vector Machine (SVM) shows an RMSE of 2.00, an MAE of 1.06, a relRMSE of 0.1301456, and an r\u0026sup2; of 0.25(Fig. 10). This method performs better than both Exponential and Linear Regression but is outperformed by Random Forest in all evaluated metrics. Hence, Random Forest emerges as the most effective regression method among the ones evaluated, demonstrating the lowest error rates and the highest predictive accuracy. While other methods like support vector machine and expotentional result showed that not good for the present dataset.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAboveground Biomass Predicted Vs Actual by Quantiles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe stacked bar chart illustrates the comparison of predicted versus actual Above Ground Biomass (AGB) percentages across four quantiles (0-2, 0-3.6, 0-6.12, and 13-15.4). Each quantile shows the distribution of AGB predictions across four categories (1: 0-0.2, 2: 0-3.6, 3: 6-12, and 4: 13-15.4). In Quantile 1 (0-2), the red category (Predicted Biomass: 1) dominates with 57.69%, followed by orange (2) at 26.32% and green (3) at 13.16%, while blue (4) has the smallest contribution at 2.63%. In Quantile 2 (0-3.6), the red category decreases to 24.32%, and the green category becomes dominant at 43.24%, while orange and blue categories contribute 24.32% and 8.11%, respectively. Quantile 3 (6-12) shows a consistent dominance of the green category at 43.24%, a further decline in the red category to 13.51%, and increased contributions from orange (10.81%) and blue (32.43%). By Quantile 4 (13-15.4), the blue category dominates with 56.76%, indicating improved prediction performance for higher AGB values, while the orange and green categories contribute equally at 18.92%, and the red category accounts for only 5.41%. In conclusion, the chart highlights the model\u0026rsquo;s progression in prediction accuracy across quantiles. Initially, the red category dominates, reflecting an underestimation of biomass in lower quantiles, but its influence decreases in higher quantiles as the blue category gains dominance. This shift indicates an improvement in predicting higher biomass values in later stages. However, the persistence of red and orange categories across quantiles suggests a need for further refinement of the model to enhance consistency and accuracy across all AGB ranges (Fig. 11).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eVariable importance AboveGround Biomass\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe bar chart highlights the importance of various variables in predicting Above Ground Biomass (AGB) using a Random Forest (RF) model. The x-axis represents the percentage increase in Mean Squared Error (% Inc MSE) when a variable is excluded, which indicates the variable\u0026apos;s relative importance to the model. Among the variables, Sigma0_VV_GLCM Variance emerges as the most influential, with the highest % Inc MSE, underscoring its critical role in accurate biomass prediction. Following this, Sigma0_VH and Sigma0_VH_GLCM Correlation are identified as the second and third most significant variables, respectively, highlighting their substantial contributions to the model\u0026apos;s predictive power. Moderately important variables include Diff and multi, which enhance the model\u0026apos;s accuracy but to a lesser extent. Variables such as Ratio, mean, and Sigma0_VV show lower importance but still contribute meaningfully to the predictions. On the other hand, traditional vegetation indices like NDVI and SAVI demonstrate the least impact, with minimal % Inc MSE, suggesting their limited effectiveness compared to radar-based variables. In conclusion, the analysis underscores the dominance of radar-based variables, such as Sigma0_VV_GLCM Variance and Sigma0_VH, in predicting AGB. Their sensitivity to structural and textural properties makes them more effective for biomass estimation compared to traditional vegetation indices. This finding highlights the importance of carefully selecting variables to optimize the performance of biomass prediction models (Fig. 12).\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe study on Forest Biomass Modeling Using In-Situ and Remote Sensing Data reveals the potential of integrating remote sensing data with ground-based measurements for accurate biomass estimation. Various modeling techniques were applied, including Exponential Regression, Linear Regression, Random Forest, and Support Vector Machine (SVM), with each method showing different levels of performance in terms of prediction accuracy. Among the methods tested, Random Forest exhibited the best performance with the lowest RMSE (1.61) and MAE (0.84), along with the highest r\u003csup\u003e2\u003c/sup\u003e value of 0.51, indicating that it is the most reliable model for forest biomass prediction in this study. In contrast, Exponential Regression demonstrated a lower RMSE (2.26) and MAE (0.97) but a relatively low r\u003csup\u003e2\u003c/sup\u003e (0.04), suggesting that while it performed decently in terms of error metrics, its overall explanatory power was limited. Linear Regression showed poorer results, with a higher RMSE (2.48) and r\u003csup\u003e2\u003c/sup\u003e value of -0.16, indicating that it struggled to capture the complexity of the biomass data. SVM also showed moderate performance with a RMSE of 2.00, MAE of 1.06, and an r\u003csup\u003e2\u003c/sup\u003e of 0.25, demonstrating reasonable predictive ability, but not as strong as Random Forest. The integration of remote sensing data, particularly radar-based variables like Sigma0_VV_GLCM Variance and Sigma0_VH, significantly improved the model\u0026rsquo;s predictive capabilities, especially in forests with complex canopies where traditional vegetation indices like NDVI and SAVI were less effective. This study highlights the importance of combining multiple data sources and modeling techniques for accurate forest biomass estimation. The findings underscore the strength of machine learning models, particularly Random Forest, for biomass prediction and their potential application in large-scale forest management and carbon stock estimation. In conclusion, this research demonstrates the effectiveness of using remote sensing and in-situ data for forest biomass modeling. Future work should focus on further refining these models, incorporating additional data sources, and exploring new remote sensing technologies to enhance the precision and scalability of biomass estimation across different forest types and regions\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our sincere gratitude to Forest department of Madhya Pradesh providing for their continuous support over the years. We also extend our thanks to the field staff whose effort and dedication were essential for the data collection throughout the duration of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData will be made available on request\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no conflict of interest to declare that are relevant to the contents of this article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbbas S, Wong MS, Wu J, Shahzad N, Irteza SM (2020) Approaches of Satellite Remote Sensing for the Assessment of Above-Ground Biomass across Tropical Forests: Pan-tropical to National Scales. Remote Sensing 12:3351\u003c/li\u003e\n\u003cli\u003eAhirwar R, Malik MS, Shukla JP (2020) Groundwater vulnerability assessment of Hoshangabad and Budni industrial area, Madhya Pradesh, India, using geospatial techniques. Applied Water Science 10\u003c/li\u003e\n\u003cli\u003eBaccini A, Friedl MA, Woodcock CE, Warbington R (2004) Forest biomass estimation over regional scales using multisource data. Geophysical Research Letters 31\u003c/li\u003e\n\u003cli\u003eBrovkina O, Novotn\u0026yacute; J, Cienciala E, Zemek F, Russ R (2016) Mapping forest aboveground biomass using airborne hyperspectral and LiDAR data in the mountainous conditions of Central Europe. Ecological Engineering 100:219\u003c/li\u003e\n\u003cli\u003eBrown S (1997) Estimating biomass and biomass change of tropical forests: A primer\u003c/li\u003e\n\u003cli\u003eChen Q, McRoberts RE, Wang C, Radtke PJ (2016) Forest aboveground biomass mapping and estimation across multiple spatial scales using model-based inference. Remote Sensing of Environment 184:350\u003c/li\u003e\n\u003cli\u003eCui H (2019) Sub-urban land classification using GF-2 images and support vector machine method. IOP Conference Series Earth and Environmental Science 351:12028\u003c/li\u003e\n\u003cli\u003eDuncanson L, Armston J, Disney M, Avitabile V, Barbier N, Calders K, Carter S, Chave J, Herold M, Crowther TW, Falkowski MJ, Kellner JR, Labri\u0026egrave;re N, Lucas R, MacBean N, McRoberts RE, Meyer V, N\u0026aelig;sset E, Nickeson J, Paul KI, Phillips OL, R\u0026eacute;jou‐M\u0026eacute;chain M, Rom\u0026aacute;n MO, Roxburgh SH, Saatchi S, Schepaschenko D, Scipal K, Siqueira P, Whitehurst A, Williams M (2019) The Importance of Consistent Global Forest Aboveground Biomass Product Validation. Surveys in Geophysics 40:979\u003c/li\u003e\n\u003cli\u003eEisfelder C, Kuenzer C, Dech S (2011) Derivation of biomass information for semi-arid areas using remote-sensing data. International Journal of Remote Sensing 33:2937\u003c/li\u003e\n\u003cli\u003eGao Y, Lu D, Li G, Wang G, Chen Q, Liu L, Li D (2018) Comparative Analysis of Modeling Algorithms for Forest Aboveground Biomass Estimation in a Subtropical Region. Remote Sensing 10:627\u003c/li\u003e\n\u003cli\u003eGeorgopoulos N, Gitas IZ, Korhonen L, Antoniadis K, Stefanidou A (2023) Estimating Crown Biomass in a Multilayered Fir Forest Using Airborne LiDAR Data. Remote Sensing 15:2919\u003c/li\u003e\n\u003cli\u003eGleason CJ, Im J (2011) A Review of Remote Sensing of Forest Biomass and Biofuel: Options for Small-Area Applications. GIScience \u0026amp; Remote Sensing 48:141\u003c/li\u003e\n\u003cli\u003eGriscom BW, Adams J, Ellis PW, Houghton RA, Lomax G, Miteva DA, Schlesinger WH, Shoch D, Siikam\u0026auml;ki J, Smith P, Woodbury PB, Zganjar C, Blackman A, Campari JS, Conant RT, Delgado CL, Elias P, Gopalakrishna T, Hamsik MR, Herrero M, Kiesecker JM, Landis E, Laestadius L, Leavitt SM, Minnemeyer S, Polasky S, Potapov P, Putz FE, Sanderman J, Silvius M, Wollenberg E, Fargione J (2017) Natural climate solutions. Proceedings of the National Academy of Sciences 114:11645\u003c/li\u003e\n\u003cli\u003eHaya B, Evans SL, Brown LB, Bukoski JJ, Butsic V, Cabiyo B, Jacobson R, Kerr A, Potts MD, Sanchez DL (2023) Comprehensive review of carbon quantification by improved forest management offset protocols. Frontiers in Forests and Global Change 6\u003c/li\u003e\n\u003cli\u003eHuang W, Swatantran A, Johnson K, Duncanson L, Tang H, Dunne JO, Hurtt GC, Dubayah R (2015) Local discrepancies in continental scale biomass maps: a case study over forested and non-forested landscapes in Maryland, USA. Carbon Balance and Management 10\u003c/li\u003e\n\u003cli\u003eKaasalainen S, Holopainen M, Karjalainen M, Vastaranta M, Kankare V, Karila K, Osmanoğlu B (2015) Combining Lidar and Synthetic Aperture Radar Data to Estimate Forest Biomass: Status and Prospects. Forests 6:252\u003c/li\u003e\n\u003cli\u003eKellner JR, Armston J, Birrer M, Cushman KC, Duncanson L, Eck C, Falleger C, Imbach B, Kr\u0026aacute;l K, Krůček M, Trochta J, Vr\u0026scaron;ka T, Zgraggen C (2019) New Opportunities for Forest Remote Sensing Through Ultra-High-Density Drone Lidar. Surveys in Geophysics 40:959\u003c/li\u003e\n\u003cli\u003eKilham P, Hartebrodt C, K\u0026auml;ndler G (2018) Generating Tree-Level Harvest Predictions from Forest Inventories with Random Forests. Forests 10:20\u003c/li\u003e\n\u003cli\u003eKnott JA, Liknes GC, Giebink CL, Oh S, Domke GM, McRoberts RE, Quirino VF, Walters BF (2023) Effects of outliers on remote sensing‐assisted forest biomass estimation: A case study from the United States national forest inventory. Methods in Ecology and Evolution 14:1587\u003c/li\u003e\n\u003cli\u003eKong B, Yu H, Du R, Wang Q (2018) Quantitative Estimation of Biomass of Alpine Grasslands Using Hyperspectral Remote Sensing. Rangeland Ecology \u0026amp; Management 72:336\u003c/li\u003e\n\u003cli\u003eKumar L, Sinha P, Taylor S, Alqurashi AF (2015) Review of the use of remote sensing for biomass estimation to support renewable energy generation. Journal of Applied Remote Sensing 9:97696\u003c/li\u003e\n\u003cli\u003eLaurin GV, Pirotti F, Callegari M, Chen Q, Cuozzo G, Lingua E, Notarnicola C, Papale D (2016) Potential of ALOS2 and NDVI to Estimate Forest Above-Ground Biomass, and Comparison with Lidar-Derived Estimates. Remote Sensing 9:18\u003c/li\u003e\n\u003cli\u003eLu D (2006) The potential and challenge of remote sensing‐based biomass estimation. International Journal of Remote Sensing 27:1297\u003c/li\u003e\n\u003cli\u003eLu D, Chen Q, Wang G, Liu L, Li G, Mor\u0026aacute;n EF (2014) A survey of remote sensing-based aboveground biomass estimation methods in forest ecosystems. International Journal of Digital Earth 9:63\u003c/li\u003e\n\u003cli\u003eMinh DHT, Toan TL, Rocca F, Tebaldini S, Villard L, R\u0026eacute;jou‐M\u0026eacute;chain M, Phillips OL, Feldpausch TR, Dubois-Fernandez P, Scipal K, Chave J (2016) SAR tomography for the retrieval of forest biomass and height: Cross-validation at two tropical forest sites in French Guiana. Remote Sensing of Environment 175:138\u003c/li\u003e\n\u003cli\u003eMo L, Zohner CM, Reich PB, Liang J, de‐Miguel S, Nabuurs GJ, Renner SS, Hoogen J van den, Araza A, Herold M, Mirzagholi L, Ma H, Averill C, Phillips OL, Gamarra JGP, Hordijk I, Routh D, Abegg M, Yao YCA, Alberti G, Zambrano AMA, Alvarado BV, \u0026Aacute;lvarez‐D\u0026aacute;vila E, \u0026Aacute;lvarez-Loayza P, Alves LF, Amaral IL do, Ammer C, Ant\u0026oacute;n‐Fern\u0026aacute;ndez C, Araujo‐Murakami A, Arroyo L, Avitabile V, C. GAA, Baker TR, Bałazy R, B\u0026aacute;nki O, Barroso J, Bastian ML, Bastin J, Birigazzi L, Birnbaum P, Bitariho R, Boeckx P, Bongers F, Bouriaud O, Brancalion PHS, Brandl S, Brearley FQ, Brienen R, Broadbent EN, Bruelheide H, Bussotti F, Gatti RC, C\u0026eacute;sar RG, Če\u0026scaron;ljar G, Chazdon RL, Chen HYH, Chisholm C, Cho H, Cienciala E, Clark CJ, Clark DB, Colletta GD, Coomes DA, Valverde FC, Corral‐Rivas JJ, Crim PM, Cumming J, Dayanandan S, Gasper AL de, Decuyper M, Derroire G, DeVries B, Djordjević I, Doležal J, Dourdain A, Obiang NLE, Enquist BJ, Eyre TJ, Fandohan AB, Fayle TM, Feldpausch TR, Ferreira LV, Fin\u0026eacute;r L, Fischer M, Fletcher C, Frizzera L, Gianelle D, Glick HB, Harris DJ, Hector A, Hemp A, Hengeveld G, H\u0026eacute;rault B, Herbohn J, Hillers A, Coronado ENH, Hui C, Ibanez T, Imai N, Jagodziński AM (2023) Integrated global assessment of the natural forest carbon potential. Nature 624:92\u003c/li\u003e\n\u003cli\u003eMohren GMJ, Hasenauer H, K\u0026ouml;hl M, Nabuurs G-J (2012) Forest inventories for carbon change assessments. Current Opinion in Environmental Sustainability 4:686\u003c/li\u003e\n\u003cli\u003eN\u0026aelig;sset E, Gobakken T, Solberg S, Gr\u0026eacute;goire TG, Nelson R, St\u0026aring;hl G, Weydahl DJ (2011) Model-assisted regional forest biomass estimation using LiDAR and InSAR as auxiliary data: A case study from a boreal forest area. Remote Sensing of Environment 115:3599\u003c/li\u003e\n\u003cli\u003eNaik P, Dalponte M, Bruzzone L (2021) Prediction of Forest Aboveground Biomass Using Multitemporal Multispectral Remote Sensing Data. Remote Sensing 13:1282\u003c/li\u003e\n\u003cli\u003ePittman JJ, Arnall DB, Interrante SM, Moffet CA, Butler TJ (2015) Estimation of Biomass and Canopy Height in Bermudagrass, Alfalfa, and Wheat Using Ultrasonic, Laser, and Spectral Sensors. Sensors 15:2920\u003c/li\u003e\n\u003cli\u003ePiza\u0026ntilde;a JMG, N\u0026uacute;\u0026ntilde;ez JM, Romero NC (2016) Remote Sensing-Based Biomass Estimation. In: InTech eBooks\u003c/li\u003e\n\u003cli\u003eRadtke PJ, Walker D, Frank J, Weiskittel AR, DeYoung C, MacFarlane DW, Domke GM, Woodall CW, Coulston JW, Westfall JA (2016) Improved accuracy of aboveground biomass and carbon estimates for live trees in forests of the eastern United States. Forestry An International Journal of Forest Research 90:32\u003c/li\u003e\n\u003cli\u003eRajput A, Yadav SS, Kori PK (2021) Bird diversity of Narmada wetland area and its adjoining habitat of Mathwad Range Alirajpur Division, Madhya Pradesh, India. International Journal of Applied Research 7:184\u003c/li\u003e\n\u003cli\u003eRanson KJ, Sun G, Lang RH, Chauhan NS, Cacciola R, Kıłıc \u0026Ouml; (1997) Mapping of boreal forest biomass from spaceborne synthetic aperture radar. Journal of Geophysical Research Atmospheres 102:29599\u003c/li\u003e\n\u003cli\u003eSaatchi S, Houghton RA, Alval\u0026aacute; RC dos S, Soares JV, Yu Y (2007) Distribution of aboveground live biomass in the Amazon basin. Global Change Biology 13:816\u003c/li\u003e\n\u003cli\u003eSchepaschenko D, Chave J, Phillips OL, Lewis SL, Davies SJ, R\u0026eacute;jou‐M\u0026eacute;chain M, Sist P, Scipal K, Perger C, H\u0026eacute;rault B, Labri\u0026egrave;re N, Hofhansl F, Affum‐Baffoe K, Алейников АА, Alonso A, Amani C, Araujo‐Murakami A, Armston J, Arroyo L, Ascarrunz N, Azevedo CP de, Baker TR, Bałazy R, Bedeau C, Berry N, Bilous A, Bilous S, Bissiengou P, Blanc L, Бобкова КС, Braslavskaya T, Brienen R, Burslem DFRP, Condit R, Cun\u0026iacute;‐Sanchez A, Данилина ДМ, Torres DDC, Derroire G, Descroix L, Sotta ED, d\u0026rsquo;Oliveira MV, Dresel C, Erwin TL, Евдокименко МД, Falck J, Feldpausch TR, Foli EG, Foster RB, Fritz S, Garc\u0026iacute;a‐Abril A, Горнов АВ, Горнова МВ, Gothard-Bass\u0026eacute;b\u0026eacute; E, Gourlet‐Fleury S, Guedes MC, Hamer KC, Susanty FH, Higuchi N, Coronado ENH, Hubau W, Hubbell SP, Ilstedt U, Иванов ВВ, Kanashiro M, Karlsson A, Karminov V, Killeen TJ, Koffi J-CK, Konovalova ME, Kraxner F, Krejza J, Krisnawati H, Krivobokov L, Kuznetsov MA, Lakyda I, Lakyda P, Licona JC, Lucas R, Лукина НВ, Lussetti D, Malhi Y, Manzanera JA, Marimon BS, Marimon BH, V\u0026aacute;squez R, Мартыненко ОВ, Matsala M, Matyashuk RK, Mazzei L, Memiaghe H, Mendoza C, Mendoza AM, Moroziuk OV, Mukhortova L, Musa S, Назимова ДИ, Okuda T, Oliveira LC de, Ontikov P, Осипов АФ (2019) The Forest Observation System, building a global reference dataset for remote sensing of forest biomass. Scientific Data 6\u003c/li\u003e\n\u003cli\u003eShi L, Liu S (2017) Methods of Estimating Forest Biomass: A Review. InTech eBooks\u003c/li\u003e\n\u003cli\u003eSimpson J, Smith TEL, Wooster MJ (2017) Assessment of Errors Caused by Forest Vegetation Structure in Airborne LiDAR-Derived DTMs. Remote Sensing 9:1101\u003c/li\u003e\n\u003cli\u003eSinha S, Jeganathan C, Sharma LK, Nathawat MS (2015) A review of radar remote sensing for biomass estimation. International Journal of Environmental Science and Technology 12:1779\u003c/li\u003e\n\u003cli\u003eSong Q, Albrecht CM, Xiong Z, Zhu XX (2023) Biomass Estimation and Uncertainty Quantification From Tree Height. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 16:4833\u003c/li\u003e\n\u003cli\u003eSun G, Ranson KJ, Guo Z, Zhang Z, Montesano P, Kimes DS (2011) Forest biomass mapping from lidar and radar synergies. Remote Sensing of Environment 115:2906\u003c/li\u003e\n\u003cli\u003eSundquist ET, Ackerman KV, Bliss N, Kellndorfer JM, Reeves M, Rollins M (2016) Rapid Assessment of U.S. Forest and Soil Organic Carbon Storage and Forest Biomass Carbon Sequestration Capacity\u003c/li\u003e\n\u003cli\u003eTian L, Wu X, Yu T, Li M, Qian C, Liao L, Fu W (2023) Review of Remote Sensing-Based Methods for Forest Aboveground Biomass Estimation: Progress, Challenges, and Prospects. Forests 14:1086\u003c/li\u003e\n\u003cli\u003eTsui OW, Coops NC, Wulder MA, Marshall P (2013) Integrating airborne LiDAR and space-borne radar via multivariate kriging to estimate above-ground biomass. Remote Sensing of Environment 139:340\u003c/li\u003e\n\u003cli\u003eWhite JC, Tompalski P, Vastaranta M, Wulder MA, Saarinen N, Stepper C, Coops NC (2017) A model development and application guide for generating an enhanced forest inventory using airborne laser scanning data and an area-based approach\u003c/li\u003e\n\u003cli\u003eYadava RN, Sinha B (2020) Vulnerability Assessment of Forest Fringe Villages of Madhya Pradesh, India for Planning Adaptation Strategies. Sustainability 12:1253\u003c/li\u003e\n\u003cli\u003eZolkos S, Goetz SJ, Dubayah R (2012) A meta-analysis of terrestrial aboveground biomass estimation using lidar remote sensing. Remote Sensing of Environment 128:289\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Forest Biomass, Allometric, conservation, ecological modelling, ecosystem services","lastPublishedDoi":"10.21203/rs.3.rs-6270236/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6270236/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe study estimation of forest Biomass using In-Situ and Remote Sensing data presents a comprehensive investigation into the estimation of forest biomass, a pivotal component of forest ecosystems and a key parameter in understanding carbon dynamics. This research merges in-situ field measurements with cutting-edge remote sensing technologies to develop robust and accurate models for predicting forest biomass. The research leverages data acquired from ground-based measurements, including tree diameter, height, and species composition, in tandem with remote sensing data obtained from satellite platforms. Various modelling techniques, including machine learning algorithms and statistical analyses, are applied to establish the relationship between these datasets and forest biomass. The study evaluates the performance of multiple methods, such as Exponential Regression, Linear Regression, Random Forest, and Support Vector Machines (SVM). The results indicate that Random Forest outperformed other methods with an RMSE of 1.61, MAE of 0.84, relRMSE of 0.1046609, and r\u0026sup2; of 0.51. In comparison, Exponential Regression achieved an RMSE of 2.26, MAE of 0.97, relRMSE of 0.1471322, and r\u0026sup2; of 0.04, Linear Regression produced an RMSE of 2.48, MAE of 1.34, relRMSE of 0.1616262, and r\u0026sup2; of -0.16; while SVM recorded an RMSE of 2.00, MAE of 1.06, relRMSE of 0.1301456, and r\u0026sup2; of 0.25. The outcomes of this study hold significant implications for forest management, climate change mitigation, and conservation efforts. Accurate forest biomass estimates are crucial for assessing carbon storage, understanding ecosystem health, and designing sustainable forestry practices. Moreover, by integrating in-situ and remote sensing data, this research contributes to the ongoing global efforts to monitor and protect the world's forests in an era of environmental challenges. The findings of this study provide valuable insights for policymakers, environmentalists, and researchers engaged in forestry, ecology, and climate change studies, facilitating more informed decisions and sustainable practices in forest management and conservation.\u003c/p\u003e","manuscriptTitle":"Machine learning approach for Forest Biomass Modelling with In-Situ and Remote Sensing Data in Narmadapuram central India","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-24 08:44:20","doi":"10.21203/rs.3.rs-6270236/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-05-21T18:50:08+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-16T07:28:52+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-13T07:07:06+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-23T07:10:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"32297974884687716696396685139418761410","date":"2025-04-01T07:04:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"30123106431222021075303424067837519924","date":"2025-03-30T11:16:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"27910280081671888895289840639661495243","date":"2025-03-29T18:33:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"254012952662628823948202104264397861894","date":"2025-03-29T17:47:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"101045542438902149564412489745223360485","date":"2025-03-29T16:13:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"34829116309442273881464267893786181125","date":"2025-03-29T16:00:40+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-29T15:34:28+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-03-26T14:34:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-26T14:33:55+00:00","index":"","fulltext":""},{"type":"submitted","content":"Modeling Earth Systems and Environment","date":"2025-03-20T13:46:10+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"d363367c-2dd9-444c-a5e1-7a55ca1ee14b","owner":[],"postedDate":"March 24th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-08-07T07:28:57+00:00","versionOfRecord":{"articleIdentity":"rs-6270236","link":"https://doi.org/10.1007/s40808-025-02527-4","journal":{"identity":"modeling-earth-systems-and-environment","isVorOnly":false,"title":"Modeling Earth Systems and Environment"},"publishedOn":"2025-07-19 16:05:45","publishedOnDateReadable":"July 19th, 2025"},"versionCreatedAt":"2025-03-24 08:44:20","video":"","vorDoi":"10.1007/s40808-025-02527-4","vorDoiUrl":"https://doi.org/10.1007/s40808-025-02527-4","workflowStages":[]},"version":"v1","identity":"rs-6270236","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6270236","identity":"rs-6270236","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}