Researchon the Landscape Pattern of Tidal Flat Reclamation Area in Jiangsu Coastal Zone Based on GF-1 Satellite Imagery

The tidal flat in the transitional belt between land and ocean is deeply disturbed by human reclamation activity. The reclamation and development of tidal flat is the specific land use pattern of coastal zone. Jiangsu province possesses abundant tideland resources and a large-scale tidal flat reclamation has taken place in Jiangsu during the past several decades. Analysis of the landscape pattern of reclamation is one of the important components to realize the sustainable development in the coastal zone. This paper takes Yancheng City of Jiangsu as a case to analysis the landscape pattern of reclamation area by using GF-1 satellite imagery. The panchromatic and multispectral data were resampled for geometric correction and fused to obtain the multispectral data of 2m resolution. Object-based classification method was taken to classify different landscape categories in study area. Wheat paddies, rape paddies, forest and grass, rive and lake, aquaculture ponds, facility aquaculture, facility planting, industrial land and traffic land were extracted separately from the GF-1 imagery through multi-resolution segmentation, training samples selection, classification features establishment. Landscape indices of area, shape and distribution characteristics were calculated by using Fragstats software for analyses. The results show that within the study area, high resolution GF-1 satellite imagery could identify landscape types in tidal flat reclamation area and offer reliable information for the landscape pattern analysis. The cultivated land (including wheat paddies, rape paddies) and aquaculture area (including facility aquaculture and aquaculture ponds) are the main landscape types in the study area and the shape of most of the landscape type is very regular, which reflected the characteristics of intensification. As the corridors of the study area, traffic land, river and lake played a better connecting function and the overall connectivity of the landscape is relatively high. The distribution of landscape types is uneven, but the overall concentration trend is obvious.