Machine learning approaches for rice crop yield predictions using time-series satellite data in Taiwan

Corn Biomass Estimation by Integrating Remote Sensing and Long-Term Observation Data Based on Machine Learning Techniques

“…Previous studies on biomass estimation involving machine learning models have mainly been conducted in forestlands and grasslands [8,26,30,31]. With regard to crops, the application of machine learning models has largely focused on crop yield estimation [6,17,33,34]. In this research, we explored the performance of four machine learning models in estimating the corn biomass of the whole growing season (RF, SVM, ANN, and XGBoost), all of which exhibited acceptable accuracy (Figures 5 and 6).…”

“…The satellite remote sensing technique is a unique and useful method for crop biomass monitoring in a repeatable manner due to the valuable information it provides on vegetation parameters, high spatial coverage and long time series, which effectively compensates for the deficiency of traditional biomass estimation methods [2,8,[11][12][13]. Previous studies have revealed that remote sensing is a reliable and effective technique to obtain biophysical and biochemical crop information [6,[14][15][16][17]. For example, the moderate resolution imaging spectroradiometer (MODIS) of the Earth Observation System (EOS) instrument provides long time series observations at a spatial resolution ranging from 250 to 1000 m in multiple spectral bands at the visible to shortwave infrared (SWIR) wavelengths, with a global coverage of one to two days, which has been widely applied in studies on the variation in vegetation parameters at the regional and global scales [18][19][20].…”

“…Machine learning techniques have become a basic and effective way to model and extract patterns from remote sensing data due to their high computational efficiency, few required variables, and reliable results [14,[21][22][23]. Machine learning techniques, including artificial neural networks (ANNs) [24], random forests (RFs) [25][26][27], support vector machines (SVMs) [28,29], and extreme gradient boosting (XGBoost) [30][31][32] have been widely implemented to evaluate vegetation biomass based on spectral signature parameters [1,17,33,34]. Although great progress has been attained in improving the spatially explicit estimation of crop biomass with machine learning methods based on multisource data, regional crop biomass estimates still exhibit large uncertainties at local scales [6,21,35].…”

“…The second limitation is that there is a need for the further comparison of available techniques, particularly machine learning techniques, for crop biomass estimation. Comparisons of these machine learning techniques have shown that each has its own advantages and drawbacks [11,17,22,30,33]. The third uncertainty originates from the input variable selection process according to the remote sensing data during the simulation process.…”

“…The third uncertainty originates from the input variable selection process according to the remote sensing data during the simulation process. Crop biomass was highly correlated with spectral signature parameters derived from satellite imagery, such as vegetation indices, which were calculated from different integrations of visible and NIR reflectance, leaf area, and reflectance for certain bands [1,6,10,11,17]. The input variable selection was based on its sensitivity to vegetation biomass.…”