| Due to the catch limits for wild fish,the aquaculture industry has developed rapidly in the past decades.As a result,the environmental impacts of the aquaculture industry have also received increasing attention.As the world’s largest producer,consumer and exporter of aquatic products,China produced 644.45 million tons of aquatic products in 2017.Since 1993,aquaculture production has exceeded that of wild-caught fish and now accounts for 76%of domestic production.Aquaculture products play a crucial role in nutrition and global food security,however,the expansion and construction of aquaculture ponds also resulted in the reduction of tidal flats and mangrove areas,the destruction of wetland habitats,the loss of ecological functions,and the water pollution.Therefore,accurate and efficient acquisition of the spatial and temporal distribution information of aquaculture ponds is of great significance for scientific management of aquaculture industry and environmental sustainability.Based on the Google Earth Engine（GEE）cloud computing platform,this paper designed an automated algorithm for extracting aquaculture ponds using remote sensing images,and explored the spatial and temporal change and its influencing factors of the aquaculture ponds.The main works include the following aspects:（1）the automated extraction and spatio-temporal evolution of Chinese coastal aquaculture ponds based on Landsat dataBased on Landsat images,a 30 km buffer zone on both sides of China coastline as the study area,and a decision tree model was built to realize automated extraction of coastal aquaculture ponds.The decision tree model mainly relies on the MNDWI and AWEI water index to highlight spectral features,the Laplacian convolution operator in the 8-neighborhood region to enhance spatial structure features,and the SRTM DEM data mask for inland hinterland with complex landform to realize the extraction of aquaculture ponds.The overall accuracy of the extraction results is 0.96 and the kappa coefficient is 0.82.The analysis of temporal and spatial changes of coastal aquaculture ponds from 1988 to 2017 showed that the area of aquaculture ponds increased from4095.59 km~2to 10865.99 km~2,with a net growth of 6,770 km~2and an average annual increase of 233 km~2in the past 30 years.The Yellow River delta,the central and northern Jiangsu plain and the Pearl River Delta showed the most obvious area increase trend.Until 2017,2,944.66 km~2of coastal tidal flats was converted to the aquaculture ponds,accounting for 34%of the total coastal reclamation area.（2）Spatial distribution pattern analysis of aquaculture ponds in China in 2019In this paper,the Sentinel-1/2 data were used to improve the automated extraction algorithm of aquaculture ponds.In detail,sentinel-1 SAR data was used to enhance the spatial texture of aquaculture ponds,with OTSU method for adaptively thresholding water index images.By using these techniques,the efficiency and accuracy of the extraction algorithm greatly improved,and we derived the spatial distribution information of large-scale aquaculture ponds of China in 2019.The results showed that,the area of large-scale aquaculture ponds of China in 2019 was about 19500 km~2,of which the coastal area（30 km buffer of coastline）was about 11232 km~2,accounting for57.6%.（3）Temporal and spatial evolution and driving forces analysis of aquaculture ponds in the typical inland areas of China in the past 30 years--a case study of Jiangsu ProvinceBased on the method of automated extraction of aquaculture ponds with Landsat images,the temporal and spatial evolution of aquaculture ponds in recent 30 years in Jiangsu Province was analyzed.The results showed that the area of aquaculture ponds in Jiangsu Province increased from 660.29 km~2in 1988 to 4097.95 km~2in 2018,with an increase of 6.2 times and a growth rate of 114.59 km~2/a.In Jiangsu Province.At present,there were three concentrated distribution areas of aquaculture ponds in the lake strand zone of Southern Jiangsu,central Jiangsu and coastal areas.The center of gravity of aquaculture ponds in Jiangsu Province moved toward the south before 1990s and thereafter to the north,and the scale of inland aquaculture ponds was decreasing after2013.The social and economic development,population growth,policy factors and coastal industrial transformation were the the fundamental causes for the aquaculture ponds changes in Jiangsu Province.The research results of this paper can provide important technical support and data for comprehensive understanding of the development of China’s aquaculture industry and environmental protection planning.The automated extraction algorithms proposed have potential to be applied to the research on aquaculture in the other countries and regions,even in the global scale. |
PDF Full Text Request