基于CLUE-S和GMOP模型的青龙满族自治县土地利用情景模拟

目的]对青龙满族自治县未来土地利用格局进行多情景预测模拟,并探索土地经济和生态综合效益双提升目标下的土地利用发展格局,以期为区域土地资源优化配置和国土空间规划的编制提供参考。方法]基于土地利用现状数据和社会经济统计资料,分别利用CLUE-S单模型、灰色多目标优化(GMOP)+CLUE-S模型对研究区进行了土地利用情景模拟。结果](1)所选驱动因子对土地利用类型的解释能力较好,ROC值均大于0.7,满足Auto-Logistic回归要求,模拟检验的Kappa系数为91.03%,模拟效果比较理想;(2)到2020年,基于CLUE-S单模型的自然发展情景,研究区建设用地扩张占用耕地和林地现象突出,土地经济效益提升较大,但生态效益却呈现负增长;基于GMOP+CLUE-S模型的多目标优化情景,研究区建设用地扩张趋势得到控制,耕地、林地面积有所增加,未利用地得到较大程度开发,实现了土地经济和生态效益双提升,分别比2015年提升了3.12%和1.96%。结论]GMOP+CLUE-S模型能够发挥两个模型各自在结构优化和空间分配方面的优点,有利于实现区域土地利用数量结构、空间布局和综合效益的协同优化。

LAND USE SCENARIO SIMULATION IN QINGLONG MANCHU AUTONOMOUS COUNTY BASED ON CLUE S AND GMOP MODEL

This paper aims to forecast and simulate the future land use pattern of Qinglong Manchu Autonomous county, and explores the land use development pattern under the goal of improving land economy and ecological comprehensive benefits, which are expected to provide references for the optimal allocation of regional land resource and the spatial planning of national land. Based on the land use status data and socio economic statistics, the land use scenarios were simulated by the CLUE S model, GMOP+CLUE S model. The results showed as follow. (1) The selected driving factors had good interpretation ability for land use types, and the ROC values were all greater than 0.7, which met the requirements of Auto Logistic regression. The Kappa coefficient of the simulation test was 91.03%, and the simulation effect was ideal. (2) By 2020, based on the natural development scenario of CLUE S model, the phenomenon of cultivated land and woodland occupied by construction land expansion in the study area was prominent, and the economic benefits of land use would greatly improve, but the ecological benefits would decrease. Based on the multi objective optimization scenario of GMOP+CLUE S model, the expansion trend of construction land in the study area would be controlled, the area of cultivated land and forest land would increase, the unused land would be exploited to a greater extent, and the economic and ecological benefits of land would be improved by 3.12% and 1.96%, respectively, compared with 2015. It concludes that the GMOP+CLUE S model can give full play to the advantages of the two models respectively in structural optimization and in spatial allocation, and the use of the model is conducive to the coordinated optimization of regional land use quantity structure, spatial layout and comprehensive benefits.