Spectral analysis of seasonal rock and vegetation changes for detecting karst rocky desertification in southwest China

Karst rocky desertification (KRD) has become a major long-term problem for the ecological environment in karst areas of southwest China. The aim of this study was to monitor and evaluate the level of rocky desertification in Pingguo County, Guangxi Province, southwest China, by tracking the seasonal spectral changes in rock and vegetation cover. Landsat-8 Operational Land Imager (OLI) images and field measurements of karst areas were collected and a novel karst rocky desertification index (KRDI) was constructed as an indicator of rocky desertification based on greenness, humidity, and brightness. Then, the time-spectrum feature space (TSFS) model was used to determine the level of rocky desertification. The main findings were as follows. (1) KRDI is a good indicator of karst rocky desertification; the highest KRDI value was obtained for severe rocky desertification, followed by moderate rocky desertification, with slight rocky desertification exhibiting the lowest KRDI value. (2) KRDI values exhibited some differences in different seasons, with the lowest KRDI occurring in summer and the highest in winter. (3) The TSFS model can accurately monitor and identify different levels of rocky desertification. The selected key indicators performed well, with a Kappa coefficient of 0.90 and overall accuracy of 0.92. Therefore, the proposed index, combining spectral information with spatiotemporal information, is a promising method for extracting information on karst rocky desertification, with significant application potential for areas with similar geology.