Spatiotemporal changes of wetlands in China during 2000–2015 using Landsat imagery

Abstract

China has abundant wetland resources; however, wetland ecosystems are experiencing critical challenges owing to continuous anthropogenic activities and climate change. Therefore, rapid and accurate mapping of wetlands and analysis of the drivers of spatiotemporal changes in wetlands are of great importance for wetland resource conservation. In this study, we proposed a comprehensive method for wetland mapping in China based on Landsat images, in which inland wetlands were identified by the object-oriented and hierarchical classification method, and coastal wetlands were extracted by visual interpretation based on sea boundaries and sea level nodes. Additionally, partial least squares structural equation modelling was utilized to exploit the mechanisms of spatiotemporal variability in wetlands during 2000–2015. The application of the comprehensive method to Landsat TM/OLI resulted in wetland maps of China with overall classification accuracies above 85%. The national wetland maps demonstrated that the wetland areas of China in 2000 and 2015 were 476,305.68 and 456,262.01 km2, respectively, and swamps accounted for the largest proportion of the total wetland area in China (over 38%). Statistics show that wetlands are mainly located in the northern inland areas, such as Tibet, Inner Mongolia, and Xinjiang Autonomous Region, as well as the Qinghai and Heilongjiang Provinces. Wetland changes were concentrated in the Tibetan Plateau, Loess Plateau, Song-Liao Plain, and middle and lower reaches of the Yangtze River. Human activities and climate change were the primary contributors to wetland changes in China. These two factors can not only directly influence wetland changes but also can be indirectly influenced by other factors, such as soil and topography. The results of this study are intended to provide basic data for monitoring and conserving wetland resources in China.