A suitable method for alpine wetland delineation: An example for the headwater area of the yellow river, Tibetan Plateau

Abstract

Alpine wetlands are one of the most important ecosystems in the Three Rivers Source Area, China, which plays an important role in regulating the regional hydrological cycle and carbon cycle. Accordingly, Wetland area and its distribution are of great significance for wetland management and scientific research. In our study, a new wetland classification model which based on geomorphological types and combine object-oriented and decision tree classification model (ODTC), and used a new wetland classification system to accurately extract the wetland distributed in the Headwater Area of the Yellow River (HAYR) of the Qinghai-Tibet Plateau (QTP), China. The object-oriented method was first used to segment the image into several areas according to similarity in Pixels and Textures, and then the wetland was extracted through a decision tree constructed based on geomorphological types. The wetland extracted by the model was compared with that by other seven commonly methods, such as support vector machine (SVM) and random forest (RF), and it proved the accuracy was improved by 10%–20%. The overall classification accuracy rate was 98.9%. According to our results, the HAYR’s wetland area is 3142.3 km2, accounting for 16.1% of the study area. Marsh wetlands and flood wetlands accounted for 37.7% and 16.7% respectively. A three-dimensional map of the area showed that alpine wetlands in the research region are distributed around lakes, piedmont groundwater overflow belts, and inter-mountain catchment basin. This phenomenon demonstrates that hydrogeological circumstances influence alpine wetlands’ genesis and evolution. This work provides a new approach to investigating alpine wetlands.