Open Environmental Sciences, 2008, 2, 101-106 101 Open Access Development of New Water Quality Model Using Fuzzy Logic System for Malaysia Md. Pauzi Abdullah*,2, Sadia Waseem2, Raman Bai V1 and Ijaz-ul-Mohsin3 1Departmentof Civil & Environmental Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lum- pur, Malaysia 2School of Chemical Sciences and Food Technology, University Kebangsaan Malaysia, 43600 Bangi, Selangor Darul Ehsan. 3University of Engineering and Technology, Pakistan Abstract: Proper assessment of water quality status in a river system based on limited observations is an essential task for meeting the goals of environmental management. Various classification methods have been used for estimating the chang- ing status and usability of surface water in River basin. However, a discrepancy frequently arises from the lack of a clear distinction between each water utilization mode, the uncertainty in the quality criteria of water employed and vagueness or fuzziness embedded in the decision making output values. Owing to inherent imprecision, difficulties always exist in some conventional methologies like water quality index (WQI) when describing integrated water quality conditions with respect to various chemical constituents, biological aspects, nutrients, and aesthetic qualities. In recent years, the fuzzy logic based methods have demonstrated to be appropriated to address uncertainty and subjectivity in environmental issues. In the present study, a methodology based on fuzzy inference systems (FIS) to assess water quality is proposed. This pa- per presents a comparative study using fuzzy logic technique to assess status of river water quality by comparing the out- put generated by fuzzy with that of conventional methods. The model is based on observations made from Semenyih River in West Malaysia. The findings clearly indicate that the fuzzy inference system (FIS) may successfully harmonize inherent discrepancies and interpret complex conditions. This river water quality model can be extended to determine the non-regulated contaminants in water. Keywords: Water quality determinant, non-regulated contaminants, water management, water quality index, environmental monitoring, fuzzy inference system. INTRODUCTION river water quality in particular to policy makers and the public at large is with indices [3]. Environmental protection and water quality management has become an important issue in public policies throughout Traditional reports on water quality tend to be too techni- the world. More than governments are concerned about the cal and detailed, presenting monitoring data on individual quality of their environmental resources because of the com- substances, without providing a whole and interpreted pic- plexity of water quality data sets [1]. Many countries have ture of water quality. To resolve this gap, various water qual- introduced a scheme for river water quality monitoring and ity indices have been developed to integrate water quality assessment, examining separate stretches of fresh water in variables worldwide. Most of these indices are based on the terms of their chemical, biological and nutrient constituents water quality index (WQI) developed by the U.S. national and overall aesthetic condition. General indices are used as Sanitation Foundation [4]. Malaysian Department of Envi- comprehensive evaluation instruments to help assess condi- ronment (DOE) has followed their own interim national wa- tions at the earliest stage to clarify monitoring priorities for ter quality standards (INWQS) for evaluating water quality regulatory agencies dealing with pollution abatement prob- of the river for intended purposes. Environmental protection lems [2]. agencies define comprehensive sets of determinants to moni- tor water quality. In order to protect the ecological status, as Water quality indices are computed for classification of declared in the Water Framework Directive [5] not only en- water wherein the integration of parametric information on vironmental concentrations of chemicals in rivers are being water quality data and the expert’s knowledge base on their used to assess water quality, but also their effects on trophic importance & weights are considered. Considerable uncer- chains. The Catalan Water Agency (Catalonia, Spain) uses tainties are involved in the process of defining water quality 150 chemical indicators to survey the condition of water [6]. for designated uses. One of the most effective ways to com- municate information on environmental trends in general and One of the difficult tasks facing environmental managers is how to transfer their interpretation of complex environ- mental data into information that is understandable and use- *Address correspondence to this author at the School of Chemical Sciences ful to technical and policy individuals as well as the general and Food Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Se- langor Darul Ehsan, Malaysia; Tel: +603-89215447; Fax: +603-89215410; public. This is particularly important in reporting the state of E-mail: [email protected] the environment. Internationally, there have been a number 1876-3251/08 2008 Bentham Open 102 Open Environmental Sciences, 2008, Volume 2 Abdullah et al. of attempts to produce a method that meaningfully integrates program results. Fuzzy logic can be considered as a language the data sets and converts them into information [7]. In mod- that allows one to translate sophisticated statements from eling complex environmental problems, researchers often natural language into a mathematical formalism [12]. Fuzzy fail to make precise statements about inputs and outcomes, logic can deal with highly variable, linguistic, vague and but fuzzy logic could be applied to the development of envi- uncertain data or knowledge and, therefore has the ability to ronment indices in a manner that solves several common allow a logical, reliable and transparent information stream problems, including the incompatibility of observations and from data collection to data usage in environmental applica- the need for implicit value judgements [8]. Conventional tions. The Fuzzy logic has been used to assess water quality water quality regulations contain quality classes which use by developing a water quality index based on fuzzy reason- crisp sets (crisp sets means classical sets express two level in ing [13]. The basic architecture of a fuzzy inference system classical mathematic, instead of multiple level among two was shown in Fig. (1). level [0,1]) and the limits between different classes have Rule-based models often deal with the linguistic descrip- inherent imprecision [9]. tions. The linguistic aspect could be based on two different In Malaysia, the classification of rivers by the Depart- approaches in river water quality management: (1) expert ment of Environment (DOE) is based on a water quality in- knowledge and or (2) actual water quality data are available dex (WQI). A water quality index relates a group of water in a linguistic format. In this study, linguistic variables were quality determinants to common scale and combines them based on actual water quality data. Fuzzy logic provides a into a single number in accordance with a chosen method or framework to model uncertainty, the human way of thinking, model of computation. The WQI was established based on reasoning, and the perception process. Fuzzy systems were the results of an opinion poll of a panel of experts who de- first introduced by Zadeh (1965). The fuzzy logic can be used termined the choice of determinants and weights assigned to for mapping inputs to appropriate outputs. Fig. (2) shows an each chosen water quality determinants. DOE-WQI consists input-output map for the water quality classification prob- of six determinants: dissolved oxygen (0.22), biological lem. oxygen demand (0.19), chemical oxygen demand (0.16), ammonical nitrogen (0.15), suspended solid (0.16) and pH In this study, the fuzzy logic formalism has been used to (0.12). In parentheses are given the weight factors according access River water quality by developing a water quality to the importance of parameters. WQI system is to assess index based on fuzzy reasoning. Comparison has been done water quality trends for management purposes even though it over conventional method (DOE-WQI). River water quality is not meant specially as an absolute measure of the degree fuzzy model based on six input and one output was used to of pollution or the actual water quality [10]. DOE-WQI was evaluate the Semenyih River quality in Malaysia. originally designed to include six determinants designed for In this methodology, membership functions of the quality making an integrated assessment of water quality conditions determinants and fuzzy rule bases were defined and then the in order to meet utilization goals. The methods which con- fuzzy logic toolbox of MATLAB7 package was used. The tain upper and lower limits have two ambiguities. Firstly, the method was applied to the 1997-2004 observations of quality traditional water quality evaluation methods use discrete determinants of the Sememyih river basin based on Mam- form. This classification technique may cause a rough and dani fuzzy inference system. The Fuzzy model was devel- imprecise approach for data, as in this approach, a parameter oped with physicochemical determinants of their weight as- being close or far from the limit has equal importance for signed and significance ratings curve to evaluate the Se- evaluation