福建省龙海市土地利用空间分布影响因子的尺度效应分析

Scale Effect Analysis of Driving Forces of Land-use Patterns in Longhai City of Fujian Province

土地利用在不同的规模尺度上具有不同的特征,某个研究尺度上的影响因子可能在其它尺度上并不发生作用。论文运用统计方法与GIS技术从自然环境、社会经济以及基础设施条件各方面分析福建省龙海市土地利用空间分布影响因素及其尺度效应。综合考虑9种主要的土地利用方式和23个候选影响因子,其中土地利用数据来自2000年11∶0 000土地利用详查数据,地形数据来自15∶0 000DEM,而社会经济统计数据如人口分布来自以乡镇为统计单元的统计年鉴,因此必须根据其空间分布与影响因子的关系进行空间化处理。采用的基本研究单元为100m×100m,在此基础上生成200m×200m,300m×300m～2 000m×2 000m多个空间尺度序列数据图层。通过分别构建不同聚合规模上的土地利用空间分布驱动模型,探讨了福建省龙海市土地利用空间分布影响因子的尺度规模效应。研究表明,不仅模型的解释能力会随聚合规模发生变化,影响因子本身及其影响系数也随研究尺度发生不同程度的变化,呈现出显著的尺度依赖性规律。龙海市主要地类受坡度、海拔高程等地形条件的严格限制,而地形因素是不容易随时间发生改变的,因此自然条件优越的地区适合多种用地类型,因而将是用地矛盾的焦点。

Land use patterns are governed by a broad variety of potential driving forces and constraint which act over a large range of scales.It has been recognized that the types and effects of land use drivers may vary with spatial scale and multi-scale investigation of land use patterns which is essential for fully understanding of its complexity.The main purpose of this paper was to perform a multi-scale analysis of land use patterns of Longhai City in Fujian province by means of statistical analysis on the basis of bio-geophysical,socio-economic and infrastructural conditions.Twenty-three variables were selected as the candidate land use drivers and 9 main land use types were considered.Land use data was derived from the 1：10000 survey map,terrain data from the 1：50000 DEM,accessibility data, i.e.distance to the nearest rural road,is derived from 1：10000 distribution map of rural road,river,residential area,etc.But socio-economic data such as population census data was collected based on the administration areas.As a result,the spatial distribution of population data on cells was conducted based on the analysis of the relationship between population density and its influencing factors.The basic spatial organization in the analysis was a 100×100m geographical grid.Through aggregations of these cells,a total of 20 artificial aggregation levels were obtained.The independent of 9 main land use types,namely paddy field,dry land,garden plot,woodland,town land,agricultural residential area, industrial land, water body and unused land,were constructed at multiple scales respectively. The results showed that： （1）Land use models varied with aggregation level,indicating spatial scale effects.Independent variables explained more of the variance for the explanation of land use type at higher aggregation levels.Relationships obtained at a certain scale of analysis may not be directly applied at other scales.The variables included in the models and their relative importance also varied between land use types,different studying year of 1996 and 2000, respectively. （2）The distribution of paddy field was mainly restricted by slope,distance to the nearest rural road or city,aspect,agricultural population density whose influences increase with scale,elevation and distance to the nearest cover fiver whose influences occurs only in the medium or small aggregation levels.For garden plot,elevation,distance to the nearest coast or fresh water sea-route is the highest ranking variables and their contributions increase with aggregation levels.Slope,distance to the nearest town or line-river or city are the second ranking variables.For woodland,slope and distance to city or town are the most leading variables at almost all aggregation levels.Important variable also includes of distance to the nearest highroad or low road or elevation at lower aggregation levels.Variables residential areas contribute to the models to a certain extent and their contributions increase with aggregation levels.Industrial land is mainly related with distance to the nearest fresh water sea-route,total population density,distance to the nearest coast or road,slope and distance to the nearest city whose influence all increases with aggregation levels. Most land use types in Longhai City were restricted by topographic factors little along with time.h is argued that these types of analyses can support the quantitative multi-scale understanding of land use,needed for the spatially explicit land use change models.