Quantifying the Relationship between Urban Form and the Lost Vegetation Biomass under Urban Expansion Based on GWR Model

ZHANG Tong,CHEN Shuang,LI Guangyu,YAN Yue

DOI: https://doi.org/10.16258/j.cnki.1674-5906.2016.12.001

2016-01-01

Abstract:Under the background of global urbanization, optimization of urban morphological pattern is one of important issues for the sustainable urban form. The vegetation biomass in the city is a parameter standing for multiple environmental targets, including carbon sequestration, climate regulation, pollution reduction etc. For the process of urban expanding to encroach on vegetation, the lost vegetation biomass quantitatively connects with the urban morphological pattern. This paper established the quantitative relationship between six selected urban form indicators and biomass loss caused by urban expansion (TBLI), with the aim of providing a specific optimization to the urban space configuration. The results indicate that six urban form indicators in fifty cities had their own characteristics, in which the Richardson compactness (0.452~0.402), density gradient (0.034~0.326) and Gini coefficient (0.527~0.995) show that the compactness of city center in large cities were generally lower than those of small or medium-sized cities. TBLI in Yangtze River Delta had reached 2.97 Tg during the study period, in which biomass loss in small or medium-sized cities were less than those of large cities, for example, Shanghai (0.529 Tg). TBLI was significantly related to urban form indicators, OLS model (TBLI=0.00051 [Urban land increment]-0.09127 [Richardson compactness]-0.11352 [density gradient]-0.05313 [Gini coefficient]+0.118) represented the average estimation of biomass loss in the whole region. However, topographic conditions and land cover types were the main causes of spatial non-stationarity. Through urban form optimization, such as enhancing circularity of external contour of Nanjing-Zhenjiang-Yangzhou urban agglomeration, improving the compactness of elements in Taizhou urban agglomeration, and breaking the balance distribution pattern in southeast part of Zhejiang Province, biomass loss could be reduced significantly. GWR model was superior to the OLS model on fitting accuracy and visualization of modeling parameters.