ASSESSMENT OF LAKE TUMBA WATER QUALITY EMPLOYING REMOTE SENSING, LANDSAT 8 IMAGES

Abstract

Lake Tumba has been chosen in this particular study because of the far-reaching role it plays in the Congo’s basin. It is a water body harboring many kinds of fish and other aquatic organisms, and constitutes an important water supply source for the surrounding villages. The quality of lake Tumba is till now not too much impaired apparently due to less human activities taking place in the catchment. This study attempts to map and estimate lake Tumba water quality in terms of variables like TSS, TP and Chl-a concentration, employing remote sensing techniques (Landsat 8 OLI data). Concentration of various water quality variables (viz. Temperature, pH, DO, Turbidity, TSS, TP and Chl-a) was obtained at 18 points spatially distributed through the lake monitored physically in the months of September and October 2018. This study employies remote sensing techniques (Landsat 8 OLI data) combined with a limited number of monitored field water quality parameter samples. Regression models were implemented using band ratios mean reflectance values to predict measured water quality variables. Turbidity did not give good result for both September and October conditions. Only three parameters (TSS, TP and Chl-a) were taken into consideration in this particular study. The September image was also too much impacted by clouds. Therefore, only the clear portion of the lake was extracted for regression analysis purpose. The clear part of the lake contained only 9 sampling points out of 18, which were considered in the analysis for September. TSS concentrations were ranging from 1.2-2.8mgl/l for September monitoring. High levels of TSS were recorded near inflow village Ikoko. TP and Chl-a concentrations were ranging from 1.6-3.5 mug/l; and 0.018-0.038 mug/l respectively and were found near lake surface5, inflow lokongoli, village Ikoko, and lake surface1, lake surface2, village lokongoli and village nkoso respectively. Concentration of TSS and Chl-a could successfully be estimated through remote sensing for September. However, similar estimation could not show a good result for October. This could be due to the impact of clouds held up in the ninth band of October image. Only TP estimation was successful for October using entire spatial information of the lake Tumba water quality. The temporal variation in the September and October water quality data did not emerge to be significant. The Land Use Land Cover classified maps obtained from ESA allowed a study of the change detection in different components in the catchment area. The findings of this research laid vi a foundation for successfully employing remote sensing techniques for estimation and mapping of lake Tumba water quality parameters in future field monitoring studies for facilitating the overall management of this lake.