No abstract

No abstract

No abstract

The concern about water quality in inland water bodies such as lakes and reservoirs has been increasing. Owing to the complexity associated with field collection of water quality samples and subsequent laboratory analyses, scientists and researchers have employed remote sensing techniques for water quality information retrieval. Due to the limitations of linear regression methods, many researchers have employed the artificial neural network (ANN) technique to decorrelate satellite data in order to assess water quality. In this paper, we propose a method that establishes the output sensitivity toward changes in the individual input reflectance channels while modeling water quality from remote sensing data collected by Landsat thematic mapper (TM). From the sensitivity, a hypothesis about the importance of each band can be made and used as a guideline to select appropriate input variables (band combination) for ANN models based on the principle of parsimony for water quality retrieval. The approach is illustrated through a case study of Beaver Reservoir in Arkansas, USA. The results of the case study are highly promising and validate the input selection procedure outlined in this paper. The results indicate that this approach could significantly reduce the effort and computational time required to develop an ANN water quality model.

ABSTRACT. Lake water quality monitoring using traditional water sampling and laboratory analyses is very expensive and time consuming. Application of neural networks to predict water quality using satellite imagery data has a potential to make the water quality determination process cost-effective, quick, and feasible. This paper includes an indirect method of determining the concentrations of chlorophyll-a (chl-a) and suspended matter (SM), two optically active parameters of lake water quality. Radial basis function neural (RBFN) network models are developed to predict the chl-a and SM concentrations in the lake. The low cost commercially available Landsat-TM imagery spectral information was used as the input with chl-a or SM concentrations as output. The model is trained and validated with data from the years 2001, 2002, 2003, and 2004. The model testing resulted in a coefficient of determination (R 2 ) of 0.55, and 0.90, respectively, for actual and predicted chl-a and SM concentrations. The root mean square error (RMSE), standard error of prediction (SEP), and average testing accuracy indicated the merit of the developed models.