The Spatial Coupling Relationship between Settlements and Land-Water Resources in Zhangye Irrigation Districts
利用空间自相关模型，分析了张掖灌区渠系、灌溉土地与聚落的空间分布格局，发现不同灌区中水土资源与聚落的空间耦合程度存在差异。
Heihe River Basin is the second largest inland river basin in the arid northwest China.Arid inland basin lies in the water,the core area is an oasis.You can also say that it is a complex interdependence of Irrigation District,through spatial integration after the formation of the modern oasis.Under the strong control and guidance of the artificial canal system,water supply quantity determines Irrigation District,and irrigated land determines population carrying capacity,so water-soil-human systems exist a high degree of interdependence and symbiosis.One of the factors beyond the reasonable range,will be a profound impact on the system-wide coordination and stability as well as the direction of succession.Using spatial autocorrelation model,the paper interpreted the pattern of spatial organization between the water-soil resources and settlements,The results showed that:The centers of standard deviation oval of the canal system,settlements and irrigated land are the centers of gravity of thore elements,show spatial distribution characteristic of the elements of Irrigation Districts integrated with each other.In the hillside independent Irrigation District( Xinba Irrigation Districts) :spatial patterns of irrigated land and canal systems substantially coincide;compared with canal systems-irrigated land,the spatial patterns of settlements slightly shrink in the northwest direction,instructions canal systems and irrigated land have more spatial coupling relationship.Pingchuan Irrigation District in edge of the Oasis,spatial distribution patterns of settlements,canal systems and irrigated land are coincidence.Yingke Irrigation District( in a oasis core),spatial distribution patterns of canal systems and irrigated land are coincidence,but there is a big difference with the settlements,described that the spatial coupling settlements with soil and water resources are poor.