Analyzing spatial variations in land use/cover distributions: A case study of Nanchang area, China. (May 2017)

• Record Type: Journal Article
• Authors: Zhang, Yang; Wang, Chuansheng; Xie, Hongbin; Zhang, Jie; Wu, Jiaping
• Is Part Of: Ecological indicators. Volume 76(2017)
• Page Start: 52
• Page End: 63
• Publication Date: 2017-05
• Subjects: Spatial pattern -- Landscape -- Gradient -- Density indicator -- Scale effect; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2017.01.007
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: A density framework was integrated to measure spatial variations within land cover classes. Intra-class variations in area, distance and scale were comparatively analyzed. Patch and distribution characteristics in the classes have left imprints on variations. Assessing spatial variations provides information applicable to management practices. Abstract: Quantification of spatial variation is important for analyzing and predicting the environmental and social impacts of land development. This paper presents a density-based framework to analyze spatial variations within land use/cover classes through a case study of the Nanchang area, China. By means of grid sampling, the categorical patches were represented by grid densities, and spatial indicators of class abundance, scale-area curve and neighborhood density were constructed to measure the spatial variables of area, distance and scale. The scale variations within each class were demonstrated by abundance indicators and were divided into three types with different similarity degrees, which were measured by coefficients of congruence. These variations roughly corresponded to the distribution patterns revealed by the scale-area indicators. The scaling behaviors of these patterns exhibited discontinuity and coherence, which were possibly affected by the change rates of some patch characteristics in the classes. The neighborhood density indicators showed that every class was more aggregated at short distances, while multimodal patterns fluctuating in nearly random distributions occurred at considerable distances. The degree of clumping correlated positively with the abundance of each class. The characteristics of distribution sizes, ranges and patch isolation in these classes left some imprints on the variations in aggregation intensity. These findings have implications for data integration, mechanism exploration and methodological framework, which are also needed for management practices.