基于改进Markov-CA模型的黄土高原土地利用多情景模拟

土地利用/覆被的时空变化研究能为区域生态环境恢复和生态系统集成管理提供科学支持。集成Logistic回归模型、改进的Markov与FLUS模型模拟黄土高原2020—2050年3种典型情景土地利用变化。发现各情景土地利用面积变化及空间置换转移主要集中在农用地、草地和城镇用地;历史趋势延续情景下农用地减少15 205 km²,草地、城镇用地分别增加2742 km²和16 007 km²;生态保育管护情景中草地增加7076 km²,林草用地增长存在权衡关系（r_{主要林地-草地}=-0.66）,在典型区域的生态恢复管理中应加以重视;城镇发展建设情景中农用地减少20 256 km²,城镇用地增加22 032 km²,变化均达到极值,其中,南部城镇扩张与农用地减少存在强权衡关系（r_{农用地-城镇用地}=-1）。改进的Markov-FLUS模型适用于黄土高原地区的土地利用模拟,情景分析可有效揭示区域生态保护与城镇扩张的阈值变化,为区域土地利用政策权衡管理和水土保育提供科学依据。

Multi-scenario Simulation of Land Use Dynamic in the Loess Plateau using an Improved Markov-CA Model

Simulating the spatiotemporal land use and land cover dynamic can provide a scientific support for regional ecological environment restoration and integrated ecosystem management. In this study, an improved Logistic-Markov-FLUS model was used to simulate land-use changes (2020—2050) under multi-scenarios in the Loess Plateau. Results show that the changes of land use area and spatial displacement were mainly concentrated in agricultural land, grassland, and urban land. In the historical trend scenario, a decreasing trend was observed in cropland with a decreased area of 15 205 km², while an increasing trend was found in the grassland and urban land with an increased area of 2742 km² and 16 007 km², respectively. In the ecological conservation management scenario, the policy for vegetation protection had a significant effect on the increase of grassland (7076 km²), showing a significant trade-off ($r_{main forestland-grassland}=-0.66$) between grassland and forest growth due to the coexistence of land transformation. Ecological protection and construction should fully consider this ecological trade-off in specific regions. The reduction of cropland (20,256 km²) and urban land (22 032 km²) in the urban development and construction scenario was the largest. There was a strong trade-off ($r_{agricultural-urban}=-1$) between the expansion of urban land and the reduction of cropland in the southern regions. The improved Markov-FLUS model had better performance on multi-scale land use dynamic simulation in the Loess Plateau. The scenario-based method can also be used to explore the abrupt changes in ecological protection and urban expansion for better LUCC decision-making and trade-off management of ecosystems.