Regionalization by fuzzy expert system based approach optimized by genetic algorithm

Summary

In recent years soft computing methods are being increasingly used to model complex hydrologic processes. These methods can simulate the real life processes without prior knowledge of the exact relationship between their components. The principal aim of this paper is perform hydrological regionalization based on soft computing concepts in the southern strip of the Caspian Sea basin, north of Iran. The basin with an area of 42,400 sq. km has been affected by severe floods in recent years that caused damages to human life and properties. Although some 61 hydrometric stations and 31 weather stations with 44 years of observed data (1961–2005) are operated in the study area, previous flood studies in this region have been hampered by insufficient and/or reliable observed rainfall-runoff records. In order to investigate the homogeneity (h) of catchments and overcome incompatibility that may occur on boundaries of cluster groups, a fuzzy expert system (FES) approach is used which incorporates physical and climatic characteristics, as well as flood seasonality and geographic location. Genetic algorithm (GA) was employed to adjust parameters of FES and optimize the system. In order to achieve the objective, a MATLAB programming code was developed which considers the heterogeneity criteria of less than 1 (H < 1) as the satisfying criteria. The adopted approach was found superior to the conventional hydrologic regionalization methods in the region because it employs greater number of homogeneity parameters and produces lower values of heterogeneity criteria.

Introduction

Modeling of hydrological processes is immensely complicated since majority of descriptive variables such as soil, topography, land-use, and rainfall vary in space under different scales. To partially overcome the difficulties involved, soft computing is being increasingly used in recent years in modeling complex hydrologic processes. One of the most important advantages of soft computing (SC) is that it allows multi-source data management (Raman and Tewari, 2011). It provides the possibility of collecting information from a large source of data, which is significantly important for comprehending complex hydrological processes. Consequently, if significant factors are discovered, without knowledge on the exact relationships, SC is still able to perform function fitting and predict possible future relationships (Raclot and Puech, 2003). Due to these advantages, soft computing techniques have been applied by numerous researchers in recent years (e.g. Maier and Dandy, 2000, Dolling and Varas, 2001, Wright and Dastorani, 2001, Patrick et al., 2002, Coppola et al., 2003, Singh and Datta, 2004, Daliakopoulos et al., 2005, Tayfur and Singh, 2006, Garbrecht, 2006, Antar et al., 2006, Srinivasulu and Jain, 2006, Tayfur and Moramorco, 2007, Manisha et al., 2008).

Lack of hydrologic and weather station records in a given study area makes flood studies problematic such that any prediction based on short and/or sparse flood records would not be statistically reliable. Moreover, flooding is an inherently uncertain natural process which has a complex interaction with drainage basin components such as soil, topography and rainfall. Thus flood phenomena cannot be described as a linear process and thus conventional methods, such as regression or empirical equations, often fail to simulate its behavior. In this paper, soft computing techniques were used to partially overcome the deficiencies in flood records and trying to make use of all available basin data. The main objective of this study is to delineate hydrological homogeneous regions for regional flood prediction in northern basins of Iran.