Estimation of changes in urban ecosystem service values from remote sensing data in Guangzhou, China

Based on three TM/ETM+ images of Guangzhou acquired in the winters of 1995, 2000 and 2009, land use/land cover (LULC) changes and the corresponding changes of ecosystem service value (ESV) in Guangzhou were estimated by using the artificial neural network (ANN) classifier and ESV coefficients of Xie et al. Results show that the annual total ESVs of Guangzhou were $1248.53, 1255.20, 1287.58 million in 1995, 2000 and 2009, with increases of $6.67 million from 1995 to 2000, $32.38 million from 2000 to 2009, and $39.25 million from 1995 to 2009, respectively. The annual average increasing rate was on average 0.24% during the study period. The combined ESV of woods land and water body together was more than 82.5% of the annual total values of Guangzhou in the three dates. The ESVs of gas regulation, climate regulation, water conservation, waste treatment, biodiversity protection, raw material and recreational culture functions increased while the ESVs of soil formation and conservation and food production functions decreased from 1995 to 2009. It is suggested that the LULC changes were mainly responsible for the ESV changes in Guangzhou and a sound land use planning should be taken effectively for sustainable urban development in Guangzhou.