Spatiotemporal Simulation And Risk Assessment Of Farmland Environmental Quality In The Pearl River Delta

The extensive economic growth model has led to severe challenges and increasingly serious risk in environmental quality of farmland in the Pearl River Delta. Researches on the evolution law of farmland environmental quality and the environmental risk assessment provide scientific basis for improvement of farmland environmental quality and risk management decision for environmental quality in the Pearl River Delta and similar regions. In this study, the spatiotemporal evolution law and influencing factors of farmland environmental quality in Pearl River Delta were systematically analyzed. On this basis, we established a spatiotemporal simulation model of farmland environmental quality evolution based on logistic and cellular automata (CA). An empirical analysis is carried out in Dongguan City, which is a typical representative of the Pearl River Delta region. Then based on the framework of risk assessment, namely problem formulation, risk analysis and risk characterization, an improved fuzzy comprehensive evaluation model based on group-decision analytical hierarchy process and entropy weight model was proposed for the risk of farmland environmental quality. Risk of farmland environmental quality in the Pearl River Delta was evaluated and the corresponding risk management strategies were put forward. The main results were concluded as follows:(1) The spatiotemporal evolution law of farmland environmental quality were analyzed using multivariate statistical and geostatistical methods, and the driving factors of farmland environmental quality evolution were identified by cluster analysis and correlation analysis. The contribution rates of pollutant sources were calculated by principal component analysis/absolute principal component scores (PCA/APCS) model. The case analysis results showed that farmland environmental quality in Dongguan as a whole was in the warning domain (Grade 4) in both 2002 and 2012, and there was a slight downward trend in the study period. To a large extent, the change of farmland environmental quality was characterized by a degradation of the more lightly polluted regions and an improvement in the seriously polluted regions. The evolution of farmland environmental quality was mainly influenced by natural factors, industrialization and urbanization, and agricultural production activities. The results of PCA/APCS model showed that heavy metals were mainly from 3 types of sources in 2002:water area and parent material, electroplating, chemical industry and railway, and electronic industry, electroplating, leather making and highway traffic; there were two main types of sources in 2012:water area and parent material, and electronic industry, electroplating and domestic waste.(2) The customized logistic-based cellular automata (CA) model for the simulation of farmland environmental quality evolution has high simulation accuracy. The model can effectively predict the evolution of farmland environmental quality under different scenarios in the future. The results showed that the spatial variables, which were related to farmland environmental quality, can be divided into 5 categories, namely, distance-based variables, current status of farmland environmental quality, industrialization, natural factors, and land use. The scenario simulation showed that environmental regulation can mitigate the pollution of farmland environmental quality but cannot completely prevent the development of the pollution because of neighborhood effects. Therefore, the government should not only implement stricter environmental regulation but also strengthen the remediation of the currently polluted area.(3) The improved fuzzy comprehensive evaluation model based on group-decision analytical hierarchy process and entropy weight model has achieved good results in the case study, and the risk assessment result was consistent with the theoretical expectation. The evaluation index weights were determined by group decision making analytic hierarchy process (the subjective weight) and entropy weight method (the objective weight). The weights reflected both subjectivity and objectivity of the evaluation indexes, which improves accuracy of the risk assessment. And the validity of maximum membership degree law is tested in the process of determining risk level. The case analysis results showed that the risks of Guangzhou, Shenzhen, Foshan and Dongguan were higher, reaching the level of "moderate risk". These cities have received more intense human disturbance than other cities in the study area. The risks of Zhuhai, Huizhou, Zhongshan and Jiangmen were all in the level of "slight risk". According to the risk assessment and its spatial distribution, risk management strategies were put forward, which were mainly included implementation of risk zoning management, optimization of the management sequence of risk factors, establishing the whole process chain of risk management, and constituting a sound system of risk management.