[目的]对快速城市化的河流城市——江西省德兴市的土地利用结构变化及其影响因素进行研究,为河流城市土地利用结构的调整和我国正在经历快速城市化的河流城市发展提供科学依据。[方法]基于德兴市2000--2014年遥感影像的土地利用数据,应用信息熵模型和灰色关联方法分析其土地利用结构信息熵变化与各相关因素之间的关联度,并以此为基础进一步运用向量自回归模型,以脉冲响应函数和方差分解解释了德兴市土地利用结构信息熵变化与主要因素的动态相关性。[结果](1)2000--2014年德兴市耕地、林地、草地、水域、其他用地类型面积下降,建设用地面积上升;土地利用结构朝无序方向发展,土地利用类型的均衡性增强,但单一类型的优势度降低。(2)基于VAR模型的脉冲响应函数和方差分解分析还表明,长期人口总数、城市化和粮食总产量对德兴市土地利用结构信息熵变化基本呈现正向效应,且随着时间滞后期数的递增,冲击效应逐步减弱,最终趋向稳定状态;其中人口总数和城市化水平对德兴市土地利用结构信息熵变化有促进作用,但粮食总产量对其方差贡献度却随时间推移作用减弱。[结论]2000--2014年德兴市土地利用程度不高,土地利用结构朝无序方向发展;人口总数、粮食总产量和城市化水平是影响其土地利用结构信息熵的主要因素。
[Objective] Land use change and its influence factor analysis of river urban--Dexing City of Jiangxi Province in the process of rapid urbanization were explored, in order to provide scientific basis for the land use structure adjustment of river city and the development of similar river city experiencing rapid urbanization in China. [Methods] Based on the land use change data of remote sensing image of Dexing City from 2000 to 2014, the association degree between information entropy change of land use structure and related factors in Dexing City were analyzed using entropy model and grey correlation analysis. And on this basis, the dynamic correlations between information entropy change of land use structure and the main factors in Dexing City were further explained using impulse response function and variance decomposition of vector autoregressive model. [Results] (1) The areas of construction land, forest land, grassland, water, and other land-use types was decreasing, on the contrary, construction land area was rising. Moreover, the land use structure was in a disorderly development state, and the equilibrium of land use type was enhanced but the dominance of a single type was reduced. (2) According to the results of the impulse response function and variance decomposition analysis of VAR model, long-term total population, gross output of grain and urbanization level had positive effects on information entropy change of land use structure in Dexing City, and the impact effects gradually diminished and finally tended to be stable with time lag increased. Thereinto, total population and urbanization level had promoting effects on the information entropy change of land use structure in Dexing City, but the role of gross output of grain contributing to the degree of variance was declined with time going by. [Conclusoin] The degree of land use in Dexing City was not high. The land use structure was in a disorderly development state from 2000 to 2014. Total population, gross output of grain and urbanization