The impact of climate change on aeolian desertification: A case of the agro-pastoral ecotone in northern China

Land desertification, one of the gravest eco-environmental problems in the world, has been proven to be critically in-fluenced by climate change. However, the information on the future spatial-temporal patterns of land desertification under climate change has been rarely explored, which restricts the proposal of reasonable desertification control coun-termeasures to adapt to climate change. The agro-pastoral ecotone in northern China (APENC) is the most critical eco-environmental barrier in China and is also a climate change-sensitive area prone to aeolian desertification. We quan-titatively assessed the risk of aeolian desertification in the APENC to climate change and social-economic development in the near-term (2010-2039), mid-term (2040-2069) and long-term (2070-2099) by integrating the representative concentration pathway (RCP) scenarios and the shared socioeconomic pathway (SSP) scenarios using a data-mining approach. The C5.0 decision tree algorithm demonstrated acceptable reliability in aeolian desertification classifica-tion. Aeolian desertification in the APENC shows a significant persistent decreasing trend in 2010-2099 under RCP2.6-SSP1 and RCP8.5-SSP3 scenarios, whereas first increased in mid-term then decreased under RCP6.0-SSP2 sce-narios. Aeolian desertification risk is lowest under the RCP2.6-SSP1 scenarios, while it is highest under the RCP6.0-SSP2 scenarios. With climate change and socioeconomic development, the risk of aeolian desertification in APENC was generally dominated by a slight grade, i.e., >70 %. While the moderate and severe grades still occupy vast areas, approximately 20 %, and 10 %, respectively, which mainly distributed in and around the Hulunbuir Sandy Land and the Horqin Sandy Land, showing the hot spots of desertification in the APENC. The reversal trend of aeolian desertification risk in the APENC might be initiated by the significant decrease of wind speed. This work highlights the great potential of data-mining approaches on climate change and social-economic development-related land desertifi-cation assessment.