Improving the accuracy of models to map alpine grassland above-ground biomass using Google earth engine

Accurate modelling and mapping of alpine grassland aboveground biomass (AGB) are crucial for pastoral agriculture planning and management on the Qinghai Tibet Plateau (QTP). This study assessed the effectiveness of four popular models (traditional multiple linear regression (MLR), support vector machine (SVM), artificial neural network (ANN), and deep neural network (DNN)) with various input combinations (geospatial variables [GV], vegetation types [VT], field measurements [FM], meteorological variables [MV] and observation time [OT]) for AGB estimation based on a new framework for AGB modelling and mapping using Google Earth Engine. The results showed that the input feature of GV had a poor performance in AGB estimation (0.121 < R-2 < 0.591). FM improved the accuracy the most when incorporated with GV (0.815 < R-2 < 0.833). Although MV, VT and OT improved the accuracy (R-2) only by 0.112-0.216 with an importance rank order of MV > VT > OT for machine learning models, their outputs could be used to map AGB. Grass AGB was less accurately predicted than shrub AGB, but the pooling of both VTs improved estimation accuracy (R-2) by 0.171-0.269. The performance of the models followed the ranked order of DNN > ANN > SVM > MLR. DNN had the highest accuracy (R-2 = 0.818) using all non-field measured variables (excluding FM) as the inputs, and it was successfully applied to a new dataset (not associated with the data used in the training and testing) with a R-2 of 0.676. This study presents an effective and operational framework for modelling and mapping grassland AGB. Accordingly, it provides the scientific foundations to determine of sustainable grazing carrying capacity in alpine grasslands.

Automated summary

This study assessed the performance of four models (MLR, SVM, ANN, and DNN) with different input combinations for alpine grassland AGB estimation. The results showed that FM had the most significant improvement on accuracy when combined with GV. MV, VT, and OT also improved the accuracy, with the highest accuracy achieved by DNN. The study presents an effective framework for modelling and mapping AGB in grassland, which contributes to determining sustainable grazing carrying capacity.