关键词: FY-3, MODIS, 变异系数, 一致性, 稳定性

Abstract: In this paper, the reliability and accuracy of the FY-3 and MODIS NDVI data in Xinjiang from 2014 to 2020 was analyzed by using numerical comparison, correlation analysis and trend consistency analysis, and the reliability and accuracy of FY-3 and MODIS were studied. Based on FY-3 NDVI data, the vegetation change in Xinjiang from 2014 to 2020 was studied by using coefficient of variation analysis method. The results show that the MODIS NDVI monthly mean from 2014 to 2020 were generally lower than those of FY-3, however, the change rules of the two sets of data were consistent, but there were certain differences in NDVI between the two data in sandland, grassland, woodland, cultivated land, spring 2017, summer 2016 and 2018, autumn 2015 and 2019, and winter 2019, the difference areas were mainly located in Altai mountain district, Ili river valley and the eastern section of Tianshan mountains; in 2020, the average NDVI data for each land use type and season in FY-3 was greater than the average MODIS NDVI data, and the FY-3 NDVI data were closer to the true value. At the same time, the NDVI data R2, RMSE, MAD, RE, θ for each land use type and season in FY-3 were lower than the MODIS NDVI data, indicating that the FY-3 data tends to be concentrated, with higher stability than the MODIS data, the NDVI data of FY-3 superior to MODIS NDVI. In Xinjiang, the internal stability of vegetation varies greatly seasonally and spatially, with small fluctuations in summer and large fluctuations in spring and autumn on a seasonal scale. Spatially, regions with better vegetation coverage, such as grasslands, forests, and oasis farming areas, have stronger internal stability of vegetation, and vice versa.

Key words: FY-3, MODIS, Coefficient of variation, Uniformity, Stability