Forecasting loss of ecosystem service value using a BP network: a case study of the impact of the South-to-north Water Transfer Project on the ecological environmental in Xiangfan, Hubei Province, China.
- Author:
Yun-Feng CHEN
1
;
Jing-Xuan ZHOU
;
Jie XIAO
;
Yan-Ping LI
Author Information
- Publication Type:Journal Article
- MeSH: China; Conservation of Natural Resources; economics; Costs and Cost Analysis; Ecosystem; Forecasting; Models, Theoretical; Neural Networks (Computer); Social Conditions; Water Supply
- From: Biomedical and Environmental Sciences 2003;16(4):379-391
- CountryChina
- Language:English
-
Abstract:
OBJECTIVETo recognize and assess the impact of the South-to-north Water Transfer Project (SNWTP) on the ecological environment of Xiangfan, Hubei Province, situated in the water-out area, and develop sound scientific countermeasures.
METHODSA three-layer BP network was built to simulate topology and process of the eco-economy system of Xiangfan. Historical data of ecological environmental factors and socio-economic factors as inputs, and corresponding historical data of ecosystem service value (ESV) and GDP as target outputs, were presented to train and test the network. When predicted input data after 2001 were presented to trained network as generalization sets, ESVs and GDPs of 2002, 2003, 2004... till 2050 were simulated as output in succession.
RESULTSUp to 2050, the area would have suffered an accumulative total ESV loss of RMB104.9 billion, which accounted for 37.36% of the present ESV. The coinstantaneous GDP would change asynchronously with ESV, it would go through an up-to-down process and finally lose RMB89.3 billion, which accounted for 18.71% of 2001.
CONCLUSIONSThe simulation indicates that ESV loss means damage to the capability of socio-economic sustainable development, and suggests that artificial neural networks (ANNs) provide a feasible and effective method and have an important potential in ESV modeling.
