中文摘要：

利用黄土高原南部地区TM和ETM＋遥感影像,在获取区域农业土地利用变化及其驱动因子数据的基础上,采用Binary Logistic回归模型,定量分析区域内农业用地的动态演化特征与空间分布格局,并运用ArcGIS软件进行不同类型农业用地变化模拟与输出。结果表明：1）黄土高原南部地区耕地的分布主要受距县城中心的距离、高程、坡度、坡向、人口密度、地均GDP的影响,其中受坡度影响最大,发生比expβ是0.83130204,高于其他因子;其次是高程和人口密度,发生比expβ分别为0.99932988和1.00011472。2）林地的分布主要受坡度和坡向的影响,坡度影响最大,坡向次之,发生比expβ分别为1.09136459和0.99209087。3）草地的分布受坡度和人口密度的影响较大,发生比expβ分别为1.02784089和0.99843133。影响黄土高原南部地区农业土地利用的自然和人文因素相对集中,主要包括坡地形度和人口密度,这说明该区域农业土地利用的人为干扰较强烈,人类活动是研究区农业土地资源和土地生态系统变化速率不断加快的重要因素。

英文摘要：

Accessing to the regional agricultural land-use change and its driving factors data from TM and ETM ＋ image in the southern Loess Plateau and using of Binary logistic regression model,we quantitatively analyzed the dynamic evolution characteristics and spatial distribution pattern of agricultural land.And simulated the different types of agricultural land-use change using ArcGIS.The results showed that： 1）The distribution of arable land in the southern Loess Plateau is mainly affected by the distance from the town center,elevation,slope,aspect,population density,GDP unit area.The greatest effect is slope,and the second effects are elevation and population.2）The distribution of woodland is mainly affected by slope and aspect.the biggest factor is slope and the bigger factor is aspect.3）The distribution of the grass is mainly affected by slope and population density.The influencing factors for agricultural land-use are natural and human factors,including the terrain of slope and population density in the southern Loess Plateau.It shows that the most important influencing factors is human activity.Human activity is the important factor for study the agricultural land resources and the rate of change in land ecosystems.