A grey systems analysis of water quantity allocation and quality protection in Xiamen, China

This study was designed to introduce the concepts of grey systems theory into water resources management as a means for accounting for uncertainty, and to conduct a grey systems analysis of the tradeoffs between meeting water quantity/quality objectives and maximizing economic income in the specific case of Xiamen, China. The literature on water resource systems analysis was reviewed to arrive at an understanding of how water quantity and quality problems were analyzed and incorporated, how uncertainty was accounted for, and what cases have been studied in water quantity and quality management. The literature revealed that ( 1) previous studies of water quantity and quality management were related to river or lake basins, and none was about a canal basin with strict water quality requirements; (2) none of the studies in China combined both quantity and quality problems in an optimization framework; and (3) no previous study attempted to communicate uncertain messages directly into optimization processes and solutions. This study has developed a grey linear programming (GLP) model for water quantity allocation and quality planning, and advanced a new solving approach which can effectively incorporate uncertain messages into the optimization framework. This method has been applied to water quantity and quality management in a water delivery canal in Xiamen, China. Results of the case study indicate that the derived decision schemes are feasible for the study area. When the canal water quality has precedence, the scheme for lower limit of objective function has to be adopted. Under this alternative, less cropping area, manure application and livestock numbers, and no fertilizer application are programmed. When agricultural income has precedence, the scheme for upper limit of objective function can be adopted. Under this alternative, more cropping areas, manure application and livestock numbers, and some fertilizer application are programmed. Therefore, decision makers can adjust the grey decision variables (including cropping area, manure and ferti1izer applications and livestock numbers) within their grey intervals according to the detailed situations. Reliability of the method has been proved through sensitivity tests of the impacts of pollutant loss constraints on agricultural income, the costs of reducing pollutant losses, the impacts of water quantity constraints on agricultural income, and the effects of grey inputs on grey outputs