Ecological risk assessment of grassland vegetation change based on Bayesian model in Xilin Gol League, China

As an essential ecological barrier in the north of China, the grassland degradation situation in the Xilin Gol League of Inner Mongolia has remained severe in recent years. Based on data spanning from 2010 to 2022, this study utilizes a set of ten indicators, encompassing altitude, temperature, precipitation, wind speed, FVC, LAI, FRAR, NPP, ET, and above-ground biomass, to construct a Bayesian model for assessing ecological risks associated with vegetation change in the Xilin Gol League. In the case of the most crucial parameter, above-ground biomass, we integrate spectral bands, vegetation indices, and ground-truth data, employing the Random Forest algorithm to estimate the above-ground biomass in the Xilin Gol League. The study further applies the Mann-Kendall trend test and Sen's slope estimation to analyze changes in this parameter and employs it as an indicator for assessing grassland degradation. The results showed that (1) Overall, the risk level decreases gradually from the west to the east and aligns with the grassland types of the Xilin Gol League. (2) In terms of spatial and temporal distribution, medium and high-risk areas are primarily located in the western region of the Xilin Gol League, showing a distinct westward shift from 2010 to 2022. (3) Areas with higher grassland degradation risk are characterized by lower altitudes, higher temperatures, less precipitation, lower vegetation coverage, and reduced potential evapotranspiration. (4) Under the pressures of both climate warming and vegetation change, there has been a significant shift in grassland degradation risk levels in the Xilin Gol League. The application of the Bayesian model effectively reveals intricate causal relationships among various variables, such as climate, terrain, and vegetation dynamics. This study could provide guidance for the prevention and control of grassland degradation in the Xilin Gol League.