Uncertainty Analysis For Decision Making In Urban Water System Planning

The status and performance of urban water system directly affect the sustainability of water use and city developing. Urban water system planning is important for water resource conservation as well as social/economic development. During the decision procedures of urban water system planning, abundant information/data/models/methods/tools are used by decision makers with less consideration of their uncertainty, inducing uncertainty problems in the results of decision. Decision makers and the system which is analyzed can also bring uncertainty risks to decision making. These uncertainty problems may have big influence on the reliability of decision results. Thus the uncertainty analysis of urban water system planning is necessary and meaningful. For uncertainty analysis, the uncertainty problems in decision procedures are described and evaluated with mathematical methods. Then the uncertainty of decision results can be calculated and the reliability of planning can be discussed.Firstly, the concept and constitution of urban water system are confined. Following the summary of decision procedures and decision making methods, the uncertainty problems in decision procedures are brought forward. In order to describe and calculate these uncertainties, the theories and methods about uncertainty analysis are demonstrated. Then the uncertainties in urban water system planning are classified and discussed according to decision elements and decision procedures. Some kinds of methods for uncertainty analysis are tested with examples.In the case study-"Water System Planning for Beijing New Residential Area", WISYS software is used and developed for decision support. The uncertain data in WISYS database are described as interval numbers with certain probability distribution. In this way, three scenarios are set up with uncertain parameter values in representation of different water reuse levels. For the evaluation of 3 scenarios, 4 comprehensive indicators are selected to indicate the surface/ground water use efficiency, cost and financial burden. Other 4 sets of indicators are defined as baseline standards for the scenarios. Monte Carlo simulation realized in Visual Basic program is used to simulate interval numbers with certain probability distribution. Samples generated by Monte Carlo simulation are put in indicator calculation module. Indicator results by Monte Carlo simulation are analyzed by SPSS and compared with the ones by certain parameter values input. It is indicated that latter is not inside the 95% confidence interval of the former. Thus the uncertain data input has obvious effect on the decision results in this case study. This case provides a useful practice on uncertainty analysis for urban water system planning.In the finality, the problems requiring further studies are discussed.