Mapping the fractional cover of non-photosynthetic vegetation and its spatiotemporal variations in the Xilingol grassland using MODIS imagery (2000−2019)

Abstract

Obtaining accurate information on the fractional cover of non-photosynthetic vegetation (NPV) (f NPV) on grasslands is essential for monitoring soil erosion risk, assessing grassland productivity and managing grassland ecosystems. However, few studies have monitored f NPV in the Xilingol grassland region (XGR) and its spatiotemporal variations based on Moderate Resolution Imaging Spectroradiometer (MODIS) images. In this study, we determined the upper dead fuel index (DFI) threshold for NPV in the XGR. Then, a remote-sensing inversion model for f NPV was established based on the ground-measured f NPV and the DFI derived from MODIS images. Based on the inversion model, we obtained the spatial distribution and spatiotemporal variations in f NPV from 2000 to 2019. The results indicated that the DFI can reflect the NPV in the XGR with an upper threshold of 27.2. The DFI-f NPV linear regression model showed a good performance, with a coefficient of determination (R2) of 0.60 and a root mean square error of leave-one-out cross-validation (RMSECV) of 0.1574. Furthermore, the spatial distribution of f NPV exhibited significant heterogeneity, and f NPV decreased from the northeastern XGR to the southwestern XGR. The overall trend of the interannual f NPV in the XGR increased in a fluctuating manner during 2000–2019. The f NPV increased in 66.92% of the XGR and decreased in a relatively small proportion (18.21%) of the XGR.