中国西南地区未来不同情景下土地利用变化模拟
Simulation of future land use changes under different scenarios in Southwest China
摘要
Abstract
[Objective]Southwest China(Yunnan,Guizhou,Sichuan,and Chongqing)is highly sensitive to soil erosion,necessitating urgent simulation and prediction of future land use changes to support regional ecological protection and land management decisions.[Methods]Based on MCD12Q1 historical land use data(2001-2020)and LUH2 future data,the patch-generating land use simulation(PLUS)model was used to simulate land use distribution from 2020 to 2100 under three Shared Socioeconomic Pathway-Representative Concentration Pathway(SSP-RCP)scenarios:SSP1-RCP1.9(sustainable development,SSP119),SSP2-RCP4.5(medium development,SSP245),and SSP5-RCP8.5(high-speed development,SSP585).Spatiotemporal variation characteristics of land use change in this region during the 21st century were analyzed using land use dynamic degree,transition matrices,and spatial expansion analysis.[Results]1)Historically(2001-2020),the comprehensive land use dynamic degree was 0.47%,indicating a relatively high change intensity.Forestland increased significantly by 42 191 km2(reaching 47.73%coverage),primarily converted from grassland(-52 614 km2).Construction land surpassed unused land to become the fourth largest category,with its expansion concentrated in urban agglomerations such as the Chengdu-Chongqing economic zone.2)The PLUS model demonstrated high accuracy with Kappa=0.85,overall accuracy(OA)=0.91,figure of merit(FoM)=0.18.From 2020 to 2100,grassland was the primary land use type converted to other uses under all scenarios.Construction land exhibited an inverted U-shaped trend,peaking in the 2050s before decreasing.3)Under the SSP119,land use change intensity was the highest,with the peak comprehensive dynamic degree reaching 1.71%.Forestland and cropland increased by 57.33%and 66.96%,respectively.Forestland covered most areas except the Sichuan Basin,and cropland expanded in northwestern and eastern Sichuan and eastern Yunnan.Grassland area experienced a net decrease of 96.23%.Under the SSP245,the change intensity was relatively low,with comprehensive dynamic degree ranging from 0.08%to 0.34%.Forestland and cropland increased by 10.18%and 18.40%,while grassland decreased by 19.21%.Under the SSP585,the change intensity was high before 2040 and then slowed down.By 2100,the area of each land use type converged with that under the SSP245,and construction land in urban agglomerations exhibited high-intensity concentric expansion.[Conclusions]The PLUS model can reliably simulate future land use change in the complex terrain of Southwest China.The simulation results reveal the spatiotemporal characteristics of land use change in this region under different development scenarios,providing robust scientific support for territorial spatial planning(e.g.,delineation of ecological protection red lines)and land management strategy formulation(e.g.,soil erosion prevention and control).关键词
土地利用变化/PLUS模型/SSP-RCP情景/草地转化/中国西南地区Key words
land use change/PLUS model/SSP-RCP scenarios/grassland conversion/Southwest China分类
管理科学引用本文复制引用
倪语轩,姜超,李阳春..中国西南地区未来不同情景下土地利用变化模拟[J].中国水土保持科学,2026,24(3):267-279,13.基金项目
国家自然科学基金面上基金"持续性降水对滑坡影响的动力过程和预报模式改进研究"(42175170) （42175170）
自然资源部重庆典型矿区生态修复野外科学观测研究站开放课题项目"重庆市矿山生态修复植被碳汇时空格局变化及其水热条件响应机制研究"(CQORS-2023-3) （CQORS-2023-3）
中国气象局气候资源经济转化重点开放实验室开放研究课题"重庆气候生态产品价值核算研究"(2024007K) National Natural Science Foundation of China"Research of the Dynamical Processes and Improving Forecasting Models for the Impact of Persistent Precipitation on Landslides in China"(42175170).The Open Research Fund of the Observation and Research Station of Ecological Restoration for Chongqing Typical Mining Areas,Ministry of Natural Resources"Study on the Spatiotemporal Pattern Changes of Vegetation Carbon Sink in Mine Ecological Restoration in Chongqing and Its Response Mechanism to Hydrothermal Conditions"(CQORS-2023-3).The CMA Key Open Laboratory of Transforming Climate Resources to Economy"Research on the Calculation of Climate Ecological Product Value in Chongqing"(2024007K) （2024007K）
