Spatiotemporal changes of desertification areas in the Alxa Desert obtained from satellite imagery

Abstract Desertification is defined as land degradation in arid, semi‐arid and dry sub‐humid areas resulting from various factors. High‐spatial‐resolution desertification monitoring with long time series and accurate area quantification in the Alxa Desert has yet to be fully elucidated. Here, we exploited Landsat satellite images to develop a method for the monitoring of high‐resolution, large‐scale desertification dynamics using a Desertification Difference Index (DDI) model based on albedo and Topsoil Grain Size Index (TGSI). On this basis, we examined the spatial–temporal changes in the extent of desertified land and ascertained the impact of various factors (temperature, precipitation, total livestock) on the desertification process. We made a detailed classification of desertification (five types) and found that non‐desertification accounted for the smallest proportion of the entire study region (annual mean 2.00 × 10 4 km 2 , 7.8%), while severe desertification contributed the largest proportion (annual mean 7.88 × 10 4 km 2 , 30.9%). Over the past 20 years, there has been a substantial reduction in extremely severe (−251 km 2 /yr) and moderate (−230 km 2 /yr) desertification areas, demonstrating the effectiveness of desert management. Regionally, considerable attention should be paid to the eastern Tengger Desert in terms of desert control; temporally, special attention should be paid to summer. High temperatures can exacerbate extremely severe, and severe desertification, contrary to the effect of increasing precipitation. Dynamic changes in desertification will become more complex under predicted climate change patterns, indicating that desertification prevention should be prioritized over control.