基于GeoDetector-MCCA的土地利用变化驱动因素分析与多情景模拟——以黄河流域甘肃段为例  

Driving factors analysis and multi-scenario simulation of land use change based on GeoDetector-Mixed-cell CellularAutomata:Acase of the Gansu section in the Yellow River Basin

作　　者：张艳[1,2] 杨维新 吕韬 ZHANG Yan;YANG Weixin;LYU Tao(School of Land Engineering,Chang’an University,Xi’an 710054,Shaanxi,China;Shaanxi Provincial Key Laboratory of Land Improvement,Xi’an 710054,Shaanxi,China)

机构地区：[1]长安大学土地工程学院,陕西西安710054 [2]陕西省土地整治重点实验室,陕西西安710054

出　　处：《干旱区研究》2025年第4期668-681,共14页Arid Zone Research

基　　金：陕西省重大科技项目(2022ZDLSF07-05);国家自然科学基金项目(42371210)资助。

摘　　要：土地资源是人类生存发展最基本的生产要素,研究土地利用变化驱动因素及未来土地利用情景模拟对区域可持续发展具有重要意义。本文以黄河流域甘肃段为研究区,基于多源数据,采用土地利用转移矩阵、地理探测器(GeoDetector)和混合元胞自动机(MCCA)模型等方法揭示黄河流域甘肃段土地利用演变特征,并开展2035年土地利用多情景模拟。结果表明:(1)2000-2020年,研究区土地利用覆被以耕地、林地、草地为主,林地草地覆盖度较高,耕地面积显著下降,且耕地和草地之间的转换最为明显。(2)影响黄河流域甘肃段土地利用变化的主导因素包括高程、气温、降水、距农村居民点距离和人口密度,各驱动因素交互作用后的q值均有所增大。(3)MCCA模型在研究区土地利用模拟过程中有良好的模拟精度,总体精度达到0.903;2035年不同情景模拟结果各异,自然演变情景耕地、未利用地收缩,其余地类均扩张;耕地保护情景保持了耕地现状存量,草地面积下降明显;生态优先情景林地、草地面积上升显著;经济发展情景表现为更积极的开发模式,建设用地显著扩张。研究结果可为黄河流域甘肃段土地管理及高质量发展提供参考。Land resources are the most fundamental production factors for human survival and development.In-vestigating the driving factors of land use change and simulating future land use scenarios are of great signifi-cance for regional sustainable development.Taking the Gansu section in the Yellow River Basin as the research area,this paper,based on multi-source data,employs methods such as the land use transfer matrix,GeoDetector,and the Mixed-cell Cellular Automata(MCCA)model to reveal the evolution characteristics of land use and con-duct multi-scenario simulations for 2035.The results are as follows:(1)From 2000-2020,the land use/cover in the research area mainly comprised cultivated land,forest land,and grassland.The extent of forest and grassland cover was relatively high,and the area of cultivated land decreased significantly.Moreover,the conversion be-tween cultivated land and grassland was the most obvious.(2)The dominant factors influencing land use change of the Gansu section in the Yellow River Basin include elevation,temperature,precipitation,distance from rural settlements,and population density.The q-values of the interaction effects of all driving factors have increased.(3)The MCCA model exhibits high simulation accuracy,with an overall accuracy of 0.903.In 2035,the simula-tion results vary among scenarios.Under the natural evolution scenario,cultivated land and unused land contract,while other land types expand.Under the cultivated land protection scenario,the current stock of cultivated land is maintained,but the area of grassland decreases significantly.In the ecological priority scenario,the areas of for-est land and grassland increase significantly.The economic development scenario is manifested in a more aggres-sive development paradigm,under which construction land experiences a remarkable expansion.The research re-sults provide references for land management and high-quality development of the Gansu section in the Yellow River Basin.

关 键 词：土地利用 多情景模拟 地理探测器 MCCA模型 黄河流域甘肃段 

分 类 号：F301.2[经济管理—产业经济]