基于CLUE-S模型的黑河中游土地利用情景模拟研究——以张掖市甘州区为例

Land Use Change Scenarios Simulation in the Middle Reaches of the Heihe River Basin Based on CLUE-S Model—A Case of Ganzhou District of Zhangye City

由于传统的土地利用空间统计分析问题中存在固有的空间自相关效应,进而影响到不同土地利用类型空间分布概率模拟的精度。研究在CLUE-S模型中传统的二值Logistics回归的基础之上引入了空间自相关因子形成Autologistic回归模型,并将其用于区域土地利用情景模拟。结果表明:(1)考虑了土地利用类型空间自相关性的Autologistic回归模型在模拟土地利用空间格局时能更好地反映真实土地利用分布格局。耕地、林地、草地、水域及未利用地的空间格局拟合优度ROC值分别从0.914、0.820、0.697、0.635和0.798提高到0.924、0.892、0.766、0.716和0.835;(2)基于Autologistic回归分析的建模对CLUE-S模型的模拟精度有一定的提高。Autolo-gistic回归分析下模拟结果的Kappa指数0.935 4大于Logistic回归模拟结果 0.888 0;(3)通过模拟2020年研究区5种情景方案下土地利用格局,表明在不同情景方案下的土地利用/覆被格局存在显著的空间差异:①自然增长情景方案下,耕地的增加对于保障粮食安全有重要意义,而建设用地的扩张则会促进研究区经济建设的快速发展,但林地和草地转化为未利用地会加剧土地的退化和生态环境的恶化;②在3种水资源约束情景方案下,水资源对干旱区土地利用/覆被变化的制约非常明显,提高水资源利用率是改善干旱区生态环境的一个重要措施;③土地结构优化情景方案下土地利用比较全面地考虑到了流域粮食安全、生态保护以及经济发展等需要,具有较强的合理性;④经济发展情景方案下建设用地快速扩张,大量侵占周边耕地和草地,粮食安全会受到较大的威胁;⑤生态保护情景方案下林地、草地和水域等生态保护用地面积呈显著扩张,区域生态环境质量明显改善。

Spatial data, like land-use data, have a tendency to be dependent （ spatial autocorrela- tion）, which means that when using spatial models, a part of the variance may be explained by neighboring values. Through incorporating components describing the spatial autocorrelation into a classic logistic model, this study sets up a regression model （Autologistic model）, and uses the model to simulate and analyze the spatial land use patterns in Ganzhou District of Zhangye City. Then the scenario simulation of the land use/cover change in 2020 in the study area was modeled by CLUE-S （The Conversion of Land Use and its Effects at Small Region Extent） model. The re- suits show that ： （ 1 ） Through comparison with the classic logistic model without considering the spatial autocorrelation, the Autologistic model showed better goodness of fitting and higher accura- cy of fitting. The distribution of land use types of cultivated land, forestland, water area and un- used land yielded areas under the ROC curves （AUC） were improved to 0. 924, 0. 892, 0. 766, 0. 716 and 0. 835 respectively when using autologistic model. （2） The land use type spatial pattern of 2009 was simulated based on the LUCC data of 2005. The Kappa index based on Autolo- gistic model and logistic model is 0.9354 and 0. 8880, respectively ,which means the accuracy of CLUE-S model was increased by Autologistic model. （3） This paper simulates the land use pattern of 2020 based on the five scenarios indicated that there existed an obvious spatial difference during different scenarios of LUCC model. 1） Under natural growth scenario, the increase of ara- ble land is important for food security, and the expansion of construction land area will promote the economic development, but the woodland and grassland conversion to unused land will exacerbate land degradation and ecological environmental degradation. 2） Under three water resource restriction scenarios, the water resources serve as very important restriction factors for land use/ cover change, and improvement of the utilization of water resources is an important measure to improve the ecological environment in arid areas. 3） Under the land use structure optimization scenario, the land use was more rational because of taking into account the needs of food security, ecological protection and economic development in the study area. 4） Under the scenario of eco- nomic development, the construction land was rapidly expanded, and food security would be threatened due to occupation of lots of high quality cultivated land and grassland. 5 ） Under the ecological protection scenario, the regional eco-environmental quality could be improved to a large extent in woodland, grassland and water area