Dynamic monitoring of grassland vegetation degradation based on multi-source data

In recent years, due to the interference of human activities and the joint action of natural disasters, the grassland in Hongyuan County has been degraded, accompanied by the emergence of a large number of toxic and harmful weeds, the productivity of grassland has been seriously decreased, and the speed of grassland ecological restoration has been slow. Grassland is the core of grassland ecosystem and the material basis of sustainable development of grassland economy. However, in recent years, grassland degradation has become increasingly serious, which has seriously affected the healthy development of grassland and the balance of ecological environment. Therefore, it is of great significance to monitor and diagnose the degree of degradation in real time and accurately. In recent years, $3 S$ technology has gradually become the mainstream of grassland degradation research based on its advantages of rapid, large-scale, continuous and economic, and its research results have more diagnostic and simulation capabilities. This paper takes Hongyuan County of Sichuan Province as the research area, with a total area of 8,400 square kilometers. The natural grassland area in Hongyuan County reaches $11,580,300 \mathrm{mu}$, accounting for $91.8 \%$ of the total area of the county. Therefore, the whole Hongyuan County is taken as the research area in this study. Previous studies mainly focused on vegetation coverage and productivity, but each index is independent of each other, it is difficult to achieve a comprehensive assessment of grassland ecosystem degradation. Grassland vegetation cover, average height of grass layer and biomass were selected. Principal component analysis was used to determine the weight of Grassland Degradation Index (GDI). Min-Max standardization method was introduced to process the index. Finally, through Landsat image inversion and reasonable classification of grassland degradation index change rate, the grassland degradation degree from 2000 to 2023 was evaluated. The results show that: (1) The results of vegetation coverage between different sensors show that the correlation between Gaofen-6 data and the measured value is the largest, reaching 0.7259, which verifies the rationality of the grassland degradation model construction based on Gaofen-6 data in this paper. (2) The long-term analysis results of vegetation coverage from 2003 to 2023 show that, The grassland has been slightly degraded in the last 20 years. (3) Through comparison and analysis with the calculation results of land use type map and GDI model, the results show that the grassland degradation results of Hongyuan County in 2023 are of no obvious degradation level. The research results can provide scientific basis for assessing the degree of grassland degradation and the protection of grassland ecosystem. Keywords-Hongyuan County, grassland degradation monitoring, grassland degradation index, vegetation index, inversion model