During the 21st century,artificial intelligence methods have been broadly applied in geosciences to simulate complex dynamic ecosystems,but the use of artificial intelligence(AI)methods to reproduce land-use/cover change(LUCC)in arid ecosystems remains rare.This paper presents a hybrid modeling approach to understand the complexity in LUCC.Fuzzy logic,equation-based systems,and expert systems are combined to predict LUCC as de-termined by water resources and other factors.The driving factors of LUCC in this study in-clude climate change,ecological flooding,groundwater conditions,and human activities.The increase of natural flooding was found to be effective in preventing vegetation degradation.LUCCs are sensitive under different climate projections of RCP2.6,RCP4.5,and RCP8.5.Simulation results indicate that the increase of precipitation is not able to compensate for the additional evaporation losses resulting from temperature increases.The results indicate that grassland,shrub,and riparian forest regions will shrink in this study area.The change in grasslands has a strong negative correlation with the change in groundwater salinity,whereas forest change had a strong positive correlation with ecological flooding.The application of artificial intelligence to study LUCC can guide land management policies and make predic-tions regarding land degradation.

YU Yang;CAO Yiguo;HOU Dongde;DISSE Markus;BRIEDEN Andreas;ZHANG Haiyan;YU Ruide

State Key Laboratory of Desert and Oasis Ecology,Xinjiang Institute of Ecology and Geography,CAS,Urumqi 830011,China;Advanced Research Institute,Southwest University of Political Science&Law,Chongqing 401120,China;Chair of Hydrology and River Basin Management,Technical University of Munich,Munich 80333,Germany;Chair of Statistics and Risk Management,Universitaet der Bundeswehr Muenchen,Neubiberg D-85577,Germany

地理学报（英文版）

[1]YU Yang;CAO Yiguo;HOU Dongde;DISSE Markus;BRIEDEN Andreas;ZHANG Haiyan;YU Ruide-.The study of artificial intelligence for predicting land use changes in an arid ecosystem)[J].地理学报（英文版）,2022(04):717-734

LUCCs

study,artificial,intelligence,predicting,use,changes,arid,During,21st,century,methods,have,been,broadly,applied,geosciences,simulate,dynamic,ecosystems,but,reproduce,cover,remains,rare,This,paper,presents,hybrid,modeling,approach,understand,complexity,Fuzzy,equation,expert,combined,termined,by,resources,other,factors,driving,this,clude,climate,ecological,flooding,groundwater,conditions,human,activities,natural,was,found,effective,preventing,vegetation,degradation,sensitive,different,projections,RCP2,RCP4,RCP8,Simulation,results,indicate,that,precipitation,not,able,compensate,additional,evaporation,losses,resulting,from,temperature,increases,shrub,riparian,forest,regions,will,shrink,area,grasslands,has,strong,negative,correlation,salinity,whereas,had,positive,application,can,guide,management,policies,make,regarding

0.552158