Quantifying Leaf Chlorophyll Concentration of Sorghum from Hyperspectral Data Using Derivative Calculus and Machine Learning

Spatial and Spectral Dependencies of Maize Yield Estimation Using Remote Sensing

Corn (Zea mays L.) is the most abundant food/feed crop, making accurate yield estimation a critical data point for monitoring global food production. Sensors with varying spatial/spectral configurations have been used to develop corn yield models from intra-field (0.1 m ground sample distance (GSD)) to regional scales (>250 m GSD). Understanding the spatial and spectral dependencies of these models is imperative to result interpretation, scaling, and deploying models. We leveraged high spatial resolution hyperspectral data collected with an unmanned aerial system mounted sensor (272 spectral bands from 0.4-1 μm at 0.063 m GSD) to estimate silage yield. We subjected our imagery to three band selection algorithms to quantitatively assess spectral reflectance features applicability to yield estimation. We then derived 11 spectral configurations, which were spatially resampled to multiple GSDs, and applied to a support vector regression (SVR) yield estimation model. Results indicate that accuracy degrades above 4 m GSD across all configurations, and a seven-band multispectral sensor which samples the red edge and multiple near-infrared bands resulted in higher accuracy in 90% of regression trials. These results bode well for our quest toward a definitive sensor definition for global corn yield modeling, with only temporal dependencies requiring additional investigation.

Machine learning in photosynthesis: Prospects on sustainable crop development

Improving photosynthesis is a promising avenue to increase food security. Studying photosynthetic traits with the aim to improve efficiency has been one of many strategies to increase crop yield but analyzing large data sets presents an ongoing challenge. Machine learning (ML) represents a ubiquitous tool that can provide a more elaborate data analysis. Here we review the application of ML in various domains of photosynthetic research, as well as in photosynthetic pigment studies. We highlight how correlating hyperspectral data with photosynthetic parameters to improve crop yield could be achieved through various ML algorithms. We also propose strategies to employ ML in promoting photosynthetic pigment research for furthering crop yield.

UAV Multisensory Data Fusion and Multi-Task Deep Learning for High-Throughput Maize Phenotyping

Recent advances in unmanned aerial vehicles (UAV), mini and mobile sensors, and GeoAI (a blend of geospatial and artificial intelligence (AI) research) are the main highlights among agricultural innovations to improve crop productivity and thus secure vulnerable food systems. This study investigated the versatility of UAV-borne multisensory data fusion within a framework of multi-task deep learning for high-throughput phenotyping in maize. UAVs equipped with a set of miniaturized sensors including hyperspectral, thermal, and LiDAR were collected in an experimental corn field in Urbana, IL, USA during the growing season. A full suite of eight phenotypes was in situ measured at the end of the season for ground truth data, specifically, dry stalk biomass, cob biomass, dry grain yield, harvest index, grain nitrogen utilization efficiency (Grain NutE), grain nitrogen content, total plant nitrogen content, and grain density. After being funneled through a series of radiometric calibrations and geo-corrections, the aerial data were analytically processed in three primary approaches. First, an extended version normalized difference spectral index (NDSI) served as a simple arithmetic combination of different data modalities to explore the correlation degree with maize phenotypes. The extended NDSI analysis revealed the NIR spectra (750-1000 nm) alone in a strong relation with all of eight maize traits. Second, a fusion of vegetation indices, structural indices, and thermal index selectively handcrafted from each data modality was fed to classical machine learning regressors, Support Vector Machine (SVM) and Random Forest (RF). The prediction performance varied from phenotype to phenotype, ranging from R² = 0.34 for grain density up to R² = 0.85 for both grain nitrogen content and total plant nitrogen content. Further, a fusion of hyperspectral and LiDAR data completely exceeded limitations of single data modality, especially addressing the vegetation saturation effect occurring in optical remote sensing. Third, a multi-task deep convolutional neural network (CNN) was customized to take a raw imagery data fusion of hyperspectral, thermal, and LiDAR for multi-predictions of maize traits at a time. The multi-task deep learning performed predictions comparably, if not better in some traits, with the mono-task deep learning and machine learning regressors. Data augmentation used for the deep learning models boosted the prediction accuracy, which helps to alleviate the intrinsic limitation of a small sample size and unbalanced sample classes in remote sensing research. Theoretical and practical implications to plant breeders and crop growers were also made explicit during discussions in the studies.

Combining Fractional Derivatives and Machine Learning: A Review

Fractional calculus has gained a lot of attention in the last couple of years. Researchers have discovered that processes in various fields follow fractional dynamics rather than ordinary integer-ordered dynamics, meaning that the corresponding differential equations feature non-integer valued derivatives. There are several arguments for why this is the case, one of which is that fractional derivatives inherit spatiotemporal memory and/or the ability to express complex naturally occurring phenomena. Another popular topic nowadays is machine learning, i.e., learning behavior and patterns from historical data. In our ever-changing world with ever-increasing amounts of data, machine learning is a powerful tool for data analysis, problem-solving, modeling, and prediction. It has provided many further insights and discoveries in various scientific disciplines. As these two modern-day topics hold a lot of potential for combined approaches in terms of describing complex dynamics, this article review combines approaches from fractional derivatives and machine learning from the past, puts them into context, and thus provides a list of possible combined approaches and the corresponding techniques. Note, however, that this article does not deal with neural networks, as there is already extensive literature on neural networks and fractional calculus. We sorted past combined approaches from the literature into three categories, i.e., preprocessing, machine learning and fractional dynamics, and optimization. The contributions of fractional derivatives to machine learning are manifold as they provide powerful preprocessing and feature augmentation techniques, can improve physically informed machine learning, and are capable of improving hyperparameter optimization. Thus, this article serves to motivate researchers dealing with data-based problems, to be specific machine learning practitioners, to adopt new tools, and enhance their existing approaches.

Combining novel feature selection strategy and hyperspectral vegetation indices to predict crop yield

BACKGROUND

Wheat is an important food crop globally, and timely prediction of wheat yield in breeding efforts can improve selection efficiency. Traditional yield prediction method based on secondary traits is time-consuming, costly, and destructive. It is urgent to develop innovative methods to improve selection efficiency and accelerate genetic gains in the breeding cycle.

RESULTS

Crop yield prediction using remote sensing has gained popularity in recent years. This paper proposed a novel ensemble feature selection (EFS) method to improve yield prediction from hyperspectral data. For this, 207 wheat cultivars and breeding lines were grown under full and limited irrigation treatments respectively, and their canopy hyperspectral reflectance was measured at the flowering, early grain filling (EGF), mid grain filling (MGF), and late grain filling (LGF) stages. Then, 115 vegetation indices were extracted from the hyperspectral reflectance and combined with four feature selection methods, i.e., mean decrease impurity (MDI), Boruta, FeaLect, and RReliefF to train deep neural network (DNN) models for yield prediction. Next, a learning framework was developed by combining the predicted values of the selected and the full features using multiple linear regression (MLR). The results show that the selected features contributed to higher yield prediction accuracy than the full features, and the MDI method performed well across growth stages, with a mean R² ranging from 0.634 to 0.666 (mean RMSE = 0.926-0.967 t ha^-1). Also, the proposed EFS method outperformed all the individual feature selection methods across growth stages, with a mean R² ranging from 0.648 to 0.679 (mean RMSE = 0.911-0.950 t ha^-1).

CONCLUSIONS

The proposed EFS method can improve grain yield prediction from hyperspectral data and can be used to assist wheat breeders in earlier decision-making.

UAV-based multi-sensor data fusion and machine learning algorithm for yield prediction in wheat

Early prediction of grain yield helps scientists to make better breeding decisions for wheat. Use of machine learning (ML) methods for fusion of unmanned aerial vehicle (UAV)-based multi-sensor data can improve the prediction accuracy of crop yield. For this, five ML algorithms including Cubist, support vector machine (SVM), deep neural network (DNN), ridge regression (RR) and random forest (RF) were used for multi-sensor data fusion and ensemble learning for grain yield prediction in wheat. A set of thirty wheat cultivars and breeding lines were grown under three irrigation treatments i.e., light, moderate and high irrigation treatments to evaluate the yield prediction capabilities of a low-cost multi-sensor (RGB, multi-spectral and thermal infrared) UAV platform. Multi-sensor data fusion-based yield prediction showed higher accuracy compared to individual-sensor data in each ML model. The coefficient of determination (R ²) values for Cubist, SVM, DNN and RR models regarding grain yield prediction were observed from 0.527 to 0.670. Moreover, the results of ensemble learning through integrating the above models illustrated further increase in accuracy. The predictions of ensemble learning showed high R ² values up to 0.692, which was higher as compared to individual ML models across the multi-sensor data. Root mean square error (RMSE), residual prediction deviation (RPD) and ratio of prediction performance to inter-quartile range (RPIQ) were calculated to be 0.916 t ha^-1, 1.771 and 2.602, respectively. The results proved that low altitude UAV-based multi-sensor data can be used for early grain yield prediction using data fusion and an ensemble learning framework with high accuracy. This high-throughput phenotyping approach is valuable for improving the efficiency of selection in large breeding activities.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s11119-022-09938-8.

Applying spectral fractal dimension index to predict the SPAD value of rice leaves under bacterial blight disease stress

BACKGROUND

The chlorophyll content is a vital indicator for reflecting the photosynthesis ability of plants and it plays a significant role in monitoring the general health of plants. Since the chlorophyll content and the soil-plant analysis development (SPAD) value are positively correlated, it is feasible to predict the SPAD value by calculating the vegetation indices (VIs) through hyperspectral images, thereby evaluating the severity of plant diseases. However, current indices simply adopt few wavelengths of the hyperspectral information, which may decrease the prediction accuracy. Besides, few researches explored the applicability of VIs over rice under the bacterial blight disease stress.

METHODS

In this study, the SPAD value was predicted by calculating the spectral fractal dimension index (SFDI) from a hyperspectral curve (420 to 950 nm). The correlation between the SPAD value and hyperspectral information was further analyzed for determining the sensitive bands that correspond to different disease levels. In addition, a SPAD prediction model was built upon the combination of selected indices and four machine learning methods.

RESULTS

The results suggested that the SPAD value of rice leaves under different disease levels are sensitive to different wavelengths. Compared with current VIs, a stronger positive correlation was detected between the SPAD value and the SFDI, reaching an average correlation coefficient of 0.8263. For the prediction model, the one built with support vector regression and SFDI achieved the best performance, reaching R², RMSE, and RE at 0.8752, 3.7715, and 7.8614%, respectively.

CONCLUSIONS

This work provides an in-depth insight for accurately and robustly predicting the SPAD value of rice leaves under the bacterial blight disease stress, and the SFDI is of great significance for monitoring the chlorophyll content in large-scale fields non-destructively.

Retrieval of Nitrogen Content in Apple Canopy Based on Unmanned Aerial Vehicle Hyperspectral Images Using a Modified Correlation Coefficient Method

The accurate retrieval of nitrogen content based on Unmanned Aerial Vehicle (UAV) hyperspectral images is limited due to uncertainties in determining the locations of nitrogen-sensitive wavelengths. This study developed a Modified Correlation Coefficient Method (MCCM) to select wavelengths sensitive to nitrogen content. The Normalized Difference Canopy Shadow Index (NDCSI) was applied to remove the shadows from UAV hyperspectral images, thus yielding the canopy spectral information. The MCCM was then used to screen the bands sensitive to nitrogen content and to construct spectral characteristic parameters. Finally, the optimal model for nitrogen content retrieval was established and selected. As a result, the screened sensitive wavelengths for nitrogen content selected were 470, 474, 490, 514, 582, 634, and 682 nm, respectively. Among the nitrogen content retrieval models, the best model was the Support Vector Machine (SVM) model. In the training set, this model outperformed the other models with an R2 of 0.733, RMSE of 6.00%, an nRMSE of 12.76%, and a MAE of 4.49%. Validated by the ground-measured nitrogen content, this model yielded good performance with an R2 of 0.671, an RMSE of 4.73%, an nRMSE of 14.83%, and a MAE of 3.98%. This study can provide a new method for vegetation nutrient content retrieval based on UAV hyperspectral data.

Estimating Carbon, Nitrogen, and Phosphorus Contents of West–East Grassland Transect in Inner Mongolia Based on Sentinel-2 and Meteorological Data

Estimating the carbon (C), nitrogen (N), and phosphorus (P) contents of a large-span grassland transect is essential for evaluating ecosystem functioning and monitoring biogeochemical cycles. However, the field measurements are scattered, such that they cannot indicate the continuous gradient change in the grassland transect. Although remote sensing methods have been applied for the estimation of nutrient elements at the local scale in recent years, few studies have considered the effective estimation of C, N, and P contents over large-span grassland transects with complex environment including a variety of grassland types (i.e., meadow, typical grassland, and desert grassland). In this paper, an information enhancement algorithm (involving spectral enhancement, regional enhancement, and feature enhancement) is used to extract the weak information related to C, N, and P. First, the spectral simulation algorithm is used to enhance the spectral information of Sentinel-2 imagery. Then, the enhanced spectra and meteorological data are fused to express regional characteristics and the fractional differential (FD) algorithm is used to extract sensitive spectral features related to C, N, and P, in order to construct a partial least-squares regression (PLSR) model. Finally, the C, N, and P contents are estimated over a West–East grassland transect in Inner Mongolia, China. The results demonstrate that: (i) the contents of C, N, and P in large-span transects can be effectively estimated through use of the information enhancement method involving spectral enhancement, regional feature enhancement, and information enhancement, for which the estimation accuracies (R2) were 0.88, 0.78, and 0.85, respectively. Compared with the estimation results of raw Sentinel-2 imagery, the RMSE was reduced by 3.42 g/m2, 0.14 g/m2, and 13.73 mg/m2, respectively; and (ii) the continuous change trend and spatial distribution characteristics of C, N, and P contents in the west–east transect of the Inner Mongolia Plateau were obtained, which showed decreasing trends in C, N, and P contents from east to west and the characteristics of meadow > typical grassland > desert grassland. Thus, the information enhancement algorithm can help to improve estimates of C, N, and P contents when considering large-span grassland transects.

Estimation of Salinity Content in Different Saline-Alkali Zones Based on Machine Learning Model Using FOD Pretreatment Method

Soil salinization is a global ecological and environmental problem in arid and semi-arid areas that can be ameliorated via soil management, visible-near infrared-shortwave infrared (VNIR-SWIR) spectroscopy can be adapted to rapidly monitor soil salinity content. This study explored the potential of Grünwald–Letnikov fractional-order derivative (FOD), feature band selection methods, nonlinear partial least squares regression (PLSR), and four machine learning models to estimate the soil salinity content using VNIR-SWIR spectra. Ninety sample points were field scanned with VNIR-SWR and soil samples (0–20 cm) were obtained at the time of scanning. The samples points come from three zones representing different intensities of human interference (I, II, and III Zones) in Fukang, Xinjiang, China. Each zone contained thirty sample points. For modeling, we firstly adopted FOD (with intervals of 0.1 and range of 0–2) as a preprocessing method to analyze soil hyperspectral data. Then, four sets of spectral bands (R-FOD-FULL indicates full band range, R-FOD-CC5 bands that met a 0.05 significance test, R-FOD-CC1 bands that met a 0.01 significance test, and R-FOD-CC1-CARS represents CC1 combined with competitive adaptive reweighted sampling) were selected as spectral input variables to develop the estimation model. Finally, four machine learning models, namely, generalized regression neural network (GRNN), extreme learning machine (ELM), random forest (RF), and PLSR, to estimate soil salinity. Study results showed that (1) the heat map of correlation coefficient matrix between hyperspectral data and salinity indicated that FOD significantly improved the correlation. (2) The characteristic band variables extracted and used by R-FOD-CC1 were fewer in number, and redundancy between bands smaller than R-FOD-FULL and R-FOD-CC5, thus estimation accuracy of R-FOD-CC1 was higher than R-FOD-CC5 or R-FOD-FULL. A high prediction accuracy was achieved with a less complex calculation. (3) The GRNN model yielded the best salinity estimation in all three zones compared to ELM, BPNN, RF, and PLSR on the whole, whereas, the RF model had the worst estimation effect. The R-FOD-CC1-CARS-GRNN model yielded the best salinity estimation in I Zone with R2, RMSE and RPD of 0.7784, 1.8762, and 2.0568, respectively. The fractional order was 1.5 and estimation performance was great. The optimal model for predicting soil salinity in II and III Zone was, also, R-FOD-CC1-CARS-GRNN (R2 = 0.7912, RMSE = 3.4001, and RPD = 1.8985 in II Zone; R2 = 0.8192, RMSE = 6.6260, and RPD = 1.8190 in III Zone), with the fractional order of 1.7- and 1.6-, respectively, and the estimation performance were all fine. (4) The characteristic bands selected by the best model in I, II, and III Zones were 8, 9, and 11, respectively, which account for 0.45%, 0.51%, and 0.63%% of the full bands. This approach reduces the number of modeled band variables and simplifies the model structure.

Evaluation of Total Nitrogen in Water via Airborne Hyperspectral Data: Potential of Fractional Order Discretization Algorithm and Discrete Wavelet Transform Analysis

Controlling and managing surface source pollution depends on the rapid monitoring of total nitrogen in water. However, the complex factors affecting water quality (plant shading and suspended matter in water) make direct estimation extremely challenging. Considering the spectral response mechanisms of emergent plants, we coupled discrete wavelet transform (DWT) and fractional order discretization (FOD) techniques with three machine learning models (random forest (RF), bagging algorithm (bagging), and eXtreme Gradient Boosting (XGBoost)) to mine this potential spectral information. A total of 567 models were developed, and airborne hyperspectral data processed with various DWT scales and FOD techniques were compared. The effective information in the hyperspectral reflectance data were better emphasized after DWT processing. After DWT processing the original spectrum (OR), its sensitivity to TN in water was maximally improved by 0.22, and the correlation between FOD and TN in water was optimally increased by 0.57. The transformed spectral information enhanced the TN model accuracy, especially for FOD after DWT. For RF, 82% of the model R2 values improved by 0.02~0.72 compared to the model using FOD spectra; 78.8% of the bagging values improved by 0.01~0.53 and 65.0% of the XGBoost values improved by 0.01~0.64. The XGBoost model with DWT coupled with grey relation analysis (GRA) yielded the best estimation accuracy, with the highest precision of R2 = 0.91 for L6. In conclusion, appropriately scaled DWT analysis can substantially improve the accuracy of extracting TN from UAV hyperspectral images. These outcomes may facilitate the further development of accurate water quality monitoring in sophisticated global waters from drone or satellite hyperspectral data.

Evaluation of Point Hyperspectral Reflectance and Multivariate Regression Models for Grapevine Water Status Estimation

Monitoring and management of plant water status over the critical period between flowering and veraison, plays a significant role in producing grapes of premium quality. Hyperspectral spectroscopy has been widely studied in precision farming, including for the prediction of grapevine water status. However, these studies were presented based on various combinations of transformed spectral data, feature selection methods, and regression models. To evaluate the performance of different modeling pipelines for estimating grapevine water status, a study spanning the critical period was carried out in two commercial vineyards at Martinborough, New Zealand. The modeling used six hyperspectral data groups (raw reflectance, first derivative reflectance, second derivative reflectance, continuum removal variables, simple ratio indices, and vegetation indices), two variable selection methods (Spearman correlation and recursive feature elimination based on cross-validation), an ensemble of selected variables, and three regression models (partial least squares regression, random forest regression, and support vector regression). Stem water potential (used as a proxy for vine water status) was measured by a pressure bomb. Hyperspectral reflectance was undertaken by a handheld spectroradiometer. The results show that the best predictive performance was achieved by applying partial least squares regression to simple ratio indices (R2 = 0.85; RMSE = 110 kPa). Models trained with an ensemble of selected variables comprising multicombination of transformed data and variable selection approaches outperformed those fitted using single combinations. Although larger data sizes are needed for further testing, this study compares 38 modeling pipelines and presents the best combination of procedures for estimating vine water status. This may lead to the provision of rapid estimation of vine water status in a nondestructive manner and highlights the possibility of applying hyperspectral data to precision irrigation in vineyards.

Comparison of Spaceborne and UAV-Borne Remote Sensing Spectral Data for Estimating Monsoon Crop Vegetation Parameters

Various remote sensing data have been successfully applied to monitor crop vegetation parameters for different crop types. Those successful applications mostly focused on one sensor system or a single crop type. This study compares how two different sensor data (spaceborne multispectral vs unmanned aerial vehicle borne hyperspectral) can estimate crop vegetation parameters from three monsoon crops in tropical regions: finger millet, maize, and lablab. The study was conducted in two experimental field layouts (irrigated and rainfed) in Bengaluru, India, over the primary agricultural season in 2018. Each experiment contained n = 4 replicates of three crops with three different nitrogen fertiliser treatments. Two regression algorithms were employed to estimate three crop vegetation parameters: leaf area index, leaf chlorophyll concentration, and canopy water content. Overall, no clear pattern emerged of whether multispectral or hyperspectral data is superior for crop vegetation parameter estimation: hyperspectral data showed better estimation accuracy for finger millet vegetation parameters, while multispectral data indicated better results for maize and lablab vegetation parameter estimation. This study's outcome revealed the potential of two remote sensing platforms and spectral data for monitoring monsoon crops also provide insight for future studies in selecting the optimal remote sensing spectral data for monsoon crop parameter estimation.

Early Detection of Plant Viral Disease Using Hyperspectral Imaging and Deep Learning

Early detection of grapevine viral diseases is critical for early interventions in order to prevent the disease from spreading to the entire vineyard. Hyperspectral remote sensing can potentially detect and quantify viral diseases in a nondestructive manner. This study utilized hyperspectral imagery at the plant level to identify and classify grapevines inoculated with the newly discovered DNA virus grapevine vein-clearing virus (GVCV) at the early asymptomatic stages. An experiment was set up at a test site at South Farm Research Center, Columbia, MO, USA (38.92 N, −92.28 W), with two grapevine groups, namely healthy and GVCV-infected, while other conditions were controlled. Images of each vine were captured by a SPECIM IQ 400–1000 nm hyperspectral sensor (Oulu, Finland). Hyperspectral images were calibrated and preprocessed to retain only grapevine pixels. A statistical approach was employed to discriminate two reflectance spectra patterns between healthy and GVCV vines. Disease-centric vegetation indices (VIs) were established and explored in terms of their importance to the classification power. Pixel-wise (spectral features) classification was performed in parallel with image-wise (joint spatial–spectral features) classification within a framework involving deep learning architectures and traditional machine learning. The results showed that: (1) the discriminative wavelength regions included the 900–940 nm range in the near-infrared (NIR) region in vines 30 days after sowing (DAS) and the entire visual (VIS) region of 400–700 nm in vines 90 DAS; (2) the normalized pheophytization index (NPQI), fluorescence ratio index 1 (FRI1), plant senescence reflectance index (PSRI), anthocyanin index (AntGitelson), and water stress and canopy temperature (WSCT) measures were the most discriminative indices; (3) the support vector machine (SVM) was effective in VI-wise classification with smaller feature spaces, while the RF classifier performed better in pixel-wise and image-wise classification with larger feature spaces; and (4) the automated 3D convolutional neural network (3D-CNN) feature extractor provided promising results over the 2D convolutional neural network (2D-CNN) in learning features from hyperspectral data cubes with a limited number of samples.

An Evaluation of Eight Machine Learning Regression Algorithms for Forest Aboveground Biomass Estimation from Multiple Satellite Data Products

This study provided a comprehensive evaluation of eight machine learning regression algorithms for forest aboveground biomass (AGB) estimation from satellite data based on leaf area index, canopy height, net primary production, and tree cover data, as well as climatic and topographical data. Some of these algorithms have not been commonly used for forest AGB estimation such as the extremely randomized trees, stochastic gradient boosting, and categorical boosting (CatBoost) regression. For each algorithm, its hyperparameters were optimized using grid search with cross-validation, and the optimal AGB model was developed using the training dataset (80%) and AGB was predicted on the test dataset (20%). Performance metrics, feature importance as well as overestimation and underestimation were considered as indicators for evaluating the performance of an algorithm. To reduce the impacts of the random training-test data split and sampling method on the performance, the above procedures were repeated 50 times for each algorithm under the random sampling, the stratified sampling, and separate modeling scenarios. The results showed that five tree-based ensemble algorithms performed better than the three nonensemble algorithms (multivariate adaptive regression splines, support vector regression, and multilayer perceptron), and the CatBoost algorithm outperformed the other algorithms for AGB estimation. Compared with the random sampling scenario, the stratified sampling scenario and separate modeling did not significantly improve the AGB estimates, but modeling AGB for each forest type separately provided stable results in terms of the contributions of the predictor variables to the AGB estimates. All the algorithms showed forest AGB were underestimated when the AGB values were larger than 210 Mg/ha and overestimated when the AGB values were less than 120 Mg/ha. This study highlighted the capability of ensemble algorithms to improve AGB estimates and the necessity of improving AGB estimates for high and low AGB levels in future studies.