Abstract:
Two decadal trends of chlorophyll-a concentration (Chl-a) in Negombo estuary, Sri Lanka, were analyzed with satellite optical sensor data observed from 1987 to 2011 and in-situ measurements. Landsat band ratios were first regressively analyzed with insitu Chl-a data, and the regression equation obtained was used for generation of 30 m-resolution Chl-a distribution maps using time-series Landsat images which had been atmospherically corrected. This Landsat-based regression equation was also used to correct MODIS Ocean Color 3 (OC3) Chl-a products with an overestimation error for a tropical region, and the corrected
MODIS OC3 values were then regressively analyzed with ASTER band ratios on a day-by-day basis for each ASTER observation date due to unavailability of in-situ measurements at ASTER observation dates. These regression equations based on ASTER/ MODIS-OC3 were then used for generation of 15 m-resolution Chl-a distribution maps using time series ASTER and MODISOC3 images. The above equations obtained from the Negombo estuary were also applied to Chilaw lagoon near Negombo estuary for evaluation of this approach. The results indicated that channel segments and some other parts of the Negombo estuary increased eutrophication during 1987-2011, particularly during 2000-2011, with consistency between Landsat-based
and ASTER-based estimates. As a conclusion, it is mentioned that our methodology can be effectively used for prediction and mitigation of localized environmental problems including sudden fish kills frequently occurred in Negombo estuary, as a costeffective tool complementing regular monitoring programs.
