Estimating aboveground biomass of urban trees by high resolution remote sensing image: a case study in Hengqin, Zhuhai, China

Abstract

Aboveground biomass of trees is the weight of organic matter per unit area of trees. The monitoring and evaluation of aboveground biomass of urban trees is helpful to the construction of garden city. With the development of urbanization and the Continuous urban ecological research, the aboveground biomass of urban trees is getting more important for the study of urban carbon cycle and the construction of ecological city. High resolution remote sensing image, which overcomes the time-consuming and labor-consuming drawbacks of traditional methods, is a data source for estimating the aboveground biomass. In this paper, taking Hengqin New Area, Zhuhai City, Guangdong Province, China as an example, we use the method of object-oriented classification of Worldwire-03 data to extract Hengqin tree area, and analyze 15 feature parameters in remote sensing images including vegetation indexes and texture features. Combined with field measurement and aboveground biomass computation, we set up Geographical Weighted Regression (GWR) model and Random Forest (RF) model for regression. Finally we estimate and discuss the aboveground biomass of trees in Hengqin.