A risk-based model for grassland management using MODIS data: The case of Gannan region, China

The grassland classification management model is considered to be a sustainable approach for grassland resources use. However, it is difficult to accurately calculate the index of classification management of grasslands (ICG) due to high variability of grassland productivity and livestock prices. Inaccurate ICG indices may lead to adoption of unreliable grassland management strategies. To overcome the problem, this article introduces a risk-based ICG model accounting for sheep prices, grassland productivity, and their variability. In this model, grassland forage dry biomass and coverage from 2001 to 2016 were estimated with enhanced vegetation index (EVI) of each pixel in MODIS using the power model and logarithmic model, respectively. A simple linear regression model was employed to estimate sheep prices. Finally, the ICG_m was estimated by variability of grassland productivity and sheep prices derived from the models. A test of the performance of the model showed that the area of conservative and moderately productive grassland predicted with the risk-based model fits very well with the observed area. This risk-based ICG model thus provides a reliable tool for classification management of grassland.