China wetland extraction and classification using MODIS data

Wetland, as a unique ecological system with highest productive forces on earth, is one of the most important environmental resources, which are named “kidney of the earth”. Wetland mapping plays an important role on the scientific exploitation and conservation of wetland resources. It is convenient for a big range of wetlands extraction based on MODIS and one of the most significant advantages of MODIS data is able to provide time-series data sets. At present, there are many mapping applications using a variety of MODIS data products. This study completed China wetland mapping in 2012. Wetlands were classified into three classes, including Open water, Vegetated wetlands and Flooded wetlands. The study used MODIS Reflectance product MCD43A4 at 8 days temporal resolution and 500 meters spatial resolution to estimate modified normalized difference water Index (MNDWI) and normalized difference vegetation index (NDVI). The other factors included the 7 bands spectral reflectance characteristics of typical ground objects and the SRTM digital elevation information. Firstly, China was divided into four zones based on eco-geographic regional system. Then the study used the decision tree classification method to complete classification in each zone. The total accuracy was as much as 70%. The wetlands map will provide decision makers with valuable information about wetlands distribution and change in response to global warming and human activities.