Water quality management is inevitably complicated since it involves a number of environmental, socio-economic, technical, and political factors with dynamic and interactive features. Nonpoint source (NPS) pollution from agricultural activities is the most significant source of water quality deterioration in agricultural lands (Cho et al., 2008, Huang and Xia, 2001, Leóna et al., 2004 and USEPA, 1992). Due to its diffuse characteristics, NPS contamination is difficult to be controlled since it is hard to isolate and quantify the contribution from individually dispersed sources (Berka, Schreier, & Hall, 2001). One effective means for NPS pollution control is to plan the related agricultural activities for causing the water quality deterioration. In planning of such water quality management systems, uncertainties exist in many system components and may affect the system behaviours. It is thus desired that such complexities and uncertainties be effectively addressed for providing decision support for practical water quality management (Chen et al., 2005, Huang et al., 2008 and Nie et al., 2008).
Mathematical models can facilitate in identifying effective decision schemes for water quality management. Previously, a variety of inexact water quality management models have been developed for dealing with various types of uncertainties (Chang et al., 1997, Chen et al., 2005, Huang, 1996, Huang, 1998, Karmakar and Mujumdar, 2006, Lee and Wen, 1997, Luo et al., 2006, Nie et al., 2008, Sasikumar and Mujumdar, 1998, Sasikumar and Mujumdar, 2000, Zhang et al., 2009 and Zhang et al., 2010). They could be categorized into fuzzy programming models, stochastic programming models, and interval programming models. Due to the inherent complexities and uncertainties, these mathematical models were highly complicated and involved in a number of mathematical knowledge on uncertainty analysis, modeling formulation and solution algorithms. The decision makers often encounter difficulties in understanding the inexact modeling results and formulation of desired policies and strategies for water quality management.
A decision support system (DSS) can be helpful for handling the above situations and facilitating the decision making processes. Previously, a number of decision support systems have been developed and applied in the field of environmental management (Loucks and da Costa, 1991, Recknagel et al., 1991, Ito et al., 2001, Matthies et al., 2006, Obropta et al., 2008, Quinn, 2009, Simonovic, 1996a, Simonovic, 1996b, Srinivasan and Engel, 1994 and Zhang et al., 2009). Nasiri, Maqsood, Huang, and Fuller (2007) proposed a fuzzy multiple-attribute decision support expert system to compute the water quality index for assessment and evaluation of water quality policies. Assaf and Saadeh (2008) developed an integrated decision support system to assess and evaluate alternative management plans for sewage induced degradation of surface water quality. Argent, Perraud, Rahman, Grayson, and Podger (2009) described a catchment modeling software system named E2 to improve the flexibility of the specific DSS. Kao, Pan, and Lin (2009) developed a web-based budget allocation system for regional water quality management to improve environmental sustainability. However, there was a lack of research efforts in incorporating the inexact water quality management models into the decision support systems, where the hybrid uncertainties expressed by fuzzy membership functions and interval numbers associated with the coefficients of the objective function and the constraints of the models could be effectively reflected.
Therefore, the objective of this study is to develop a model-based decision support system for supporting water quality management under hybrid uncertainties, named FICMDSS, which is based on a hybrid uncertain programming (HFICP) model with fuzzy and interval coefficients. The different functions are effectively organized and integrated within an integrated decision support framework. The coefficients in the objective function can be modeled as interval numbers, and those in the constraints can be expressed by fuzzy membership functions. The HFICP model can improve upon the existing inexact programming methods through incorporation of hybrid fuzzy and interval uncertainties into the optimization management processes and resulting solutions. An agricultural water management case is proposed for demonstrating the applicability of the developed FICMDSS. Feasible decision alternatives for cropping area, amounts of manure and fertilizer application, and sizes of livestock husbandry can be generated for achieving the maximum agricultural system benefit subject to the given water-related constraints under hybrid uncertainties.
This papers proceeds as follows. Section 2 describes the model development. Section 3 presents the development of FICMDSS. Section 4 describes the implementation of FICMDSS through a water quality management case study, and Section 5 concludes the paper.
