Monitoring of Water Quality Using Remote Sensing Data Mining

Abstract

Remote sensing techniques play increasingly important role over recent decades in both problems of global climate change and frequent deterioration of the status of aquatic ecology, driven by the ever-increasing needs of growing populations for drinking water, polluted by overland runoff from point and non-point sources, as well as fish and other seafood (Pozdnyakov et al., 2005). With the advent of new sensor technologies, it is possible to monitor land cover / land change in large area simultaneously and quickly (Wen and Yang, 2009a, Wen and Yang, 2009b). Remote sensing techniques have been widely used in water quality assessment (Alparslan et al., 2007, Brando and Dekker, 2003, Chen et al., 2007, Giardino et al., 2007, Hadjimitsis and Clayton, Kondratye et al., 1998, Koponen et al., 2002, Pozdnyakov et al., 2005, Ritchie et al., 2003, Seyhan and Dekker, 1986, Wang and Ma, 2001). Many documents describe water quality monitoring using different satellite sensor (Maillard and Pinheiro Santos, 2008, Giardino et al., 2007, He et al., 2008, Alparslan et al., 2007, Verma et al., 2008, Wang and Ma, 2001, Zhang et al., 2003, Martinez et al., 2007, Boken, 2007, Wang et al., 2004). The spectrum characteristics of water and pollutants are essential to water quality monitoring and assessment. The spectral characteristics of the signal received from water are a function of hydrological, biological and chemical characteristics of water, and other interference factor (Seyhan and Dekker, 1986). Suspended sediments increase the radiance emergent from surface waters in the visible and near infrared proportion of the electromagnetic spectrum(Ritchie et al., 1976), so it is promising and feasible to detect water pollutants using spectral signatures in the visible and near infrared band. Wang assessed the water quality of Taihu lake using Landsat TM imagery (Wang and Ma, 2001), and the result indicated that three visible bands of TM1, TM2 and TM3 were correlated with some water quality parameters from the lake. Alparslan (Alparslan et al., 2007) assessed water quality at Omerli Dam using the first four bands of Landsat 7 ETM +satellite data. Hadjimitsis (Hadjimitsis and Clayton, 2009) assessed temporal variations of water quality in inland water bodies using atmospheric corrected satellite remotely sensed image data. It found that atmospheric correction was essential to water quality assessment using satellite remotely sensed imagery because it improved significantly the water reflectance. In this paper, atmospherically corrected Landsat ETM+ imagery is used to monitors water quality using stepwise multiple linear regression analysis, southwest China. Five over 30 square