Assessing of Urban Vegetation Biomass in Combination with LiDAR and High-resolution Remote Sensing Images. Issue 3 (1st February 2021)

• Record Type: Journal Article
• Authors: Zhang, Ya; Shao, Zhenfeng
• Is Part Of: International journal of remote sensing. Volume 42:Issue 3(2021)
• Page Start: 964
• Page End: 985
• Publication Date: 2021-02-01
• Subjects: Remote sensing -- Periodicals; Télédétection -- Périodiques; 621.3678
• DOI: 10.1080/01431161.2020.1820618
• Languages: English
• ISSNs: 0143-1161

Abstract

ABSTRACT: The urban vegetation ecosystem is a vegetation ecosystem that is deeply influenced by human beings. The rapid urbanization process brings a great influence on the growth environment of urban vegetation. Urban vegetation has a certain mitigating effect on the urbanization process. It is irreplaceable by other urban ecosystems in functions such as maintaining atmospheric carbon–oxygen balance, reducing heat island effects, purifying, and beautifying the urban environment. At present, the estimation of aboveground biomass (AGB) mainly focuses on the original forest, grassland, desertification vegetation and crops, and the data sources are mostly medium- and low-resolution data such as Land Remote-Sensing Satellite Thematic Map (Landsat TM), Enhanced Thematic Mapper Plus (ETM+), and Moderate Resolution Imaging Spectroradiometer (MODIS). Compared with the research objects such as forest and grassland, urban vegetation has higher heterogeneity of underlying surface within the city scope, and there are more mixed pixels of medium- and low-resolution data. Therefore, high-resolution data are needed for classification and estimation. This study uses Light Detection and Ranging (LiDAR) data to expand the sample size, combines high-resolution image data to classify urban vegetation areas, and quantitatively estimates and inverts biomass. The spatial and temporal variation of urban vegetation biomass was analysed by comparing the inversion accuracy of five different models and the advantages and disadvantages of the research models. The research results show that: (1) In the absence of urban forest sample points, the biomass background data are expanded with the help of LiDAR data and more data is provided for further inversion; (2) Through five model method comparison experiments, the optimal method for estimation is based on the Random Forest (RF) model; (3) Analysed the changes of urban vegetation in the study area in the past 10 years, the development of different types of urban vegetation has experienced a trend of increasing first and then decreasing. Due to the high heterogeneity of features in urban areas, this study improved the inversion accuracy of estimating urban vegetation biomass by classification, and provided reference value and the basis for urban ecological management and regional planning.