20th Annual Symposium of the European-Association-of-Remote-Sensing-Laboratories (EARSeL), Dresden, Germany, 14 - 16 June 2000, pp.311-316
This study reports on an investigation using a genetic algorithm model for polluting substances and satellite data in a study area in the strait of Istanbul (Bosphorus). This is being affected negatively by the waste discharged from industrial plants and residential areas as the result of population growth and industrial development in Istanbul,. This situation is investigated using satellite digital data and surface water quality measurements, such as total suspended solids (TSS), humic materials (HM), chemical oxygen demand (COD), polyaromatic hydrocarbons (PH) and hydrodynamic conditions of water. In order to find possible significant relationships between remote sensing variables such as the reflectance values and the surface measurement a genetic algorithm regression approach is used for refined determination of regression coefficients. Adaptive parameter estimation in the genetic algorithm provides efficient computation in an economic manner. These observed reflectance shows a strong relationship with the water quality observation. The necessary values are provided in single pixel values for each band at the station point in the Bosphorus. Satellite data provide a useful index of TSS, HM and PAH. As the reflectance tin the turbidity area) in the longer red and near IR increases faster than the reflectance in shorter blue and green wave lengths, it can be seen that turbidity levels are positively related to reflectance.