Analysis of land use spatial autocorrelation patterns based on DEM data

The paper aimed to prospectively study the impact of topographic change on land use spatial autocorrelation patterns. Firstly, methods known as Moran's I, Moran scatter-plot and Anselin local Moran's I were adopted to analyze the spatial dependence in observations of land use among administrative units of Gucheng county. The results reveal that the spatial cluster and anomaly characteristics of each land use type in local areas of the whole region. Secondly, based on surface analysis, terrain factors were extracted from DEM data, including evaluation, slope, aspect, relief and roughness which were used for terrain analysis by means of zonal statistics in land use areas with spatial cluster or anomaly characteristics test at 0.05 significant level. The results show that typical features of mountains, hills and plains can greatly influence the spatial structure of land use spatial autocorrelation in regions, as it easily leads to the formation of local cluster characteristics of land use. While the generation of local anomaly characteristics usually derived from the special terrain or human activities. This study can provide a foundation for the adjustment of land use structures and optimization of land use patterns.