Mapping and assessment of wetland conditions by using remote sensing images and POI data

Abstract

Wetlands are one of the most valuable natural resources on earth and play an important role in preserving biodiversity. However, due to economic development and human disturbances, many wetlands across the world have deteriorated and disappeared over the past several decades. By using remote sensing images and point of interest (POI) data, we proposed a knowledge-based raster mapping (KBRM)-based framework and implemented it in the assessment of wetland ecological conditions in Suzhou, China. Density maps of waterbodies, vegetation covers, imperviousness, roads, and POI values were derived and used as five ecological indicators that can represent the ecological conditions of wetlands. The KBRM approach was used to integrate these indicators into an overall rating and map wetland ecological conditions efficiently. Thus, spatial variations in wetland ecological conditions can be distinguished and represented in detail. Cross validation was conducted with water quality data at 15 field sampling sites. The validation results demonstrated that the overall wetland condition scores generated by our approach and the water quality index (WQI) values calculated from water quality data were strongly correlated. These findings confirm that our framework could be used to effectively map and evaluate spatial variations in wetland ecological conditions and provide more support for policy-making in wetland protection and management

Keywords

• Wetland
• Remote sensing
• Knowledge-based raster mapping (KBRM)
• Point of interest (POI)
• Mapping and assessment