Effects of spatial grain size on different types of cultivated land landscape pattern in fragmented agricultural landscapes based on GF-2 imagery

Abstract

The landscape pattern of the cultivated land determines the type, number, distribution and the allocation of cultivated land resource, which could provide theoretical basis for ecological security research, provide reference for policy formulation. The high resolution satellite image provides macroscopic, real-time, fine and dynamic data source for the study of cultivated land landscape pattern. The study of spatial grain effects on the different types of cultivated land is an important part of the study of landscape pattern, and also provides a basis for the selection of suitable spatial resolution of remote sensing image. In this study, a concentrated area of cultivated land in the suburban area of Nanjing City in Jiangsu Province was taken as the study area. Based on the analysis of China high spatial resolution GF-2 image, different types of cultivated land in the study area were extracted, including rice, soybean, vegetables, fallow land and greenhouse cultivation. On this basis, a series of landscape pattern indices were used to analyse the effects of spatial grain size on different types of cultivated land landscape pattern characteristics. The Percentage of landscape index (PLAND) of five types of cultivated land in the study area remained stable with the increase of grain size and the Area weighted mean patch fractal index (FRAC_AM) of the five types of cultivated land landscape in the study area changed slightly especially the rice landscape. The Mean patch area (AREA_MN) and Mean nearest neighbour index (ENN_MN) increased in varying degrees with the increase of grain size, while Edge density (ED), Area weighted mean contiguity (CONTIG_AM) and Patch cohesion index (COHESION) decreased in varying degrees with the increase of grain size. Among the five types of cultivated land landscape, the ED and COHESION of soybean and fallow land decreased more obviously with the increase of grain size, and the AREA_MN a and ENN_MN of rice landscape increased most obviously with the increase of grain size.