| Inland water quality monitoring using remote sensing and GIS techniques: A case study on Tigris River, Iraq |
| Paper ID : 1096-ICRSSSA-FULL |
| Authors |
|
Wael Ahmed *1, Suhaib Mohammed2, Salem Morsy3, Adel El-Shazly1 1Public Works Department, Faculty of Engineering, Cairo University, Giza 12613, Egypt 2ENV. Engineer, Ministry of Environment, Karada, Baghdad 10062, Iraq 3Public Works Department, Faculty of Engineering, Cairo University, Giza 12613, Egypt. |
| Abstract |
| Remote sensing and GIS are effective technologies for surface water quality assessment and monitoring. These technologies help in making correct decisions that contribute to pollution reduction, its causes, and the time needed for treatment. This study aims to analyze the water quality along the Tigris River in Baghdad city, Iraq by developing mathematical and statistical models to predict water parameters from satellite imagery. In 2018, fourteen different locations along the Tigris River were surveyed. At each location, continuous measurements for eight variables, including temperature (Temp), electrical conductivity (Cond), total dissolved solids (TDS), pH, turbidity (Turb), chlorophyll A, blue-green algae (BGA), dissolved oxygen (DO) were provided. The spectral bands from Landsat-8 were modified by geometric and radiometric corrections. Then, spectral indices for soil, vegetation, and water were calculated from the corrected bands. Both spectral bands and indices were implemented in the least absolute shrinkage and selection operator (LASSO) for the prediction of those eight water variables. Evaluation of the prediction model showed that the temperature has a maximum root mean square error (RMSE) of 0.093% with 0.8 coefficient of determination (R2), while DO has a minimum RMSE of 0.012% with 0.76 R2. The predictive model for each water variable provides cost-effective alternatives to frequent monitoring of Tigris water quality using field data. |
| Keywords |
| surface water, LASSO, RMSE, GIS, WQI |
| Status: Accepted (Oral Presentation) |
