基于CA-Markov模型的藉河流域土地利用变化动态模拟

为了探讨黄土丘陵沟壑区藉河流域的土地利用变化规律及驱动力,基于1995和2008年两期遥感影像解译得到的土地利用数据,通过土地利用动态度模型和空间测算模型对藉河流域土地利用类型的动态变化程度进行分析,并运用CA-Markov模型对2022年土地利用空间格局进行预测.结果表明,空间分析测算模型由于同时考虑了各土地利用类型的空间转移及新增过程,能更为准确地测算出各种土地利用类型的动态变化程度与速率；CA-Markov模型预测结果表明,整个流域总体的Kappa系数为0.9515,预测结果可信；除草地和未利用地外,2008-2022年期间其他土地利用类型继续原有的变化趋势,到了2022年表现出坡耕地、草地、水域和未利用地面积减少趋势,梯田、林地和居民地面积呈增加趋势.黄土丘陵沟壑区水土流失严重,此研究有助于加强土地资源的保护和实施退耕还林还草等措施,为当地土地利用规划管理提供参考.

Dynamic simulation of land use change in Jihe watershed based on CA-Markov model

To explore the law of land use change and driving force in Loess hilly-gully area in Jihe watershed, based on the land use data interpretation from remote sensing images in 1995 and 2008, the degree of dynamic change of land use type in Jihe watershed was analyzed by dynamic degree model and spatial analysis model, and distribution of land use spatial patterns in 2022 was forecast by using the CA-Markov model. The results showed that spatial analysis model not only took the conversion process into account but also the spatial expansion process of land use change, so it could more precisely measure the dynamic change rate of land use. The simulation result by the CA-Markov method indicated that, the value of Kappa coefficients of agreement in the whole watershed was 0.9515, and forecasting results were credible. From 2008 to 2022, except for grassland and unused land, trends of land use evolution as well as its rate will keep constant. That is, the area of slope farmland, grassland, water and unused land will continue to reduce in some degree, while terrace, forest land and settlements will present a increasing trend. Because the Loess hilly-gully is a serious soil erosion area, this research is helpful to reinforce the protection of land resource and to enforce the agro-forestry policies of reusing farmland for forestland and grassland, and it is meaningful to serve as a scientific basis for land planning and management.