Abstract:
Water contamination in Pakistan is a serious public health concern made worse by the country's growing population. Few studies have been conducted to relate remote sensing data to model the water quality parameters. This study used the multispectral and hyperspectral data to relate and model the water quality parameters. Twenty-five randomly selected surface water quality samples were collected from the Khanpur Dam. The spectral data was collected using Landsat 8 and ASD spectroradiometer. Water quality samples were analyzed in the laboratory, using standard analytical techniques for physico-chemical properties (EC, pH, turbidity, nitrates and phosphates) and the regression analysis were applied for model development. The water qaulity parameters turbidity and phosphate indictaes above persmissible limits however paramters pH, EC and nitrates were found to be within permissible limits. The nitrate had a high correlation with turbidity(r=0.71) and phosphate (r=0.68), similarly turbidity with phosphate indicates significant correlation(r=0.81). The results of regression models for water quality prediction reveal compelling insights. ASD ratio 560/891 nm for predicting turbidity, yields coefficient of determination (R2=0.75) and root mean square (RMSE) of 0.68. In comparison, the application of the OLI ratio Band3/Band5 within the green spectral range produces an RMSE of 1.04 and R2=0.41. For phosphates prediction, the ASD ratio 600/820 nm resulted in RMSE of 0.09 and an R2 of 0.69, while the OLI ratio Band4/Band5 within the corresponding spectral range (i.e., red), shows similar performance with an RMSE of 0.09 and R2 of 0.49. Furthermore, for nitrates prediction, the ASD ratio 465/729 nm yielded R2 = 0.74 and an RMSE of 0.27, whereas, the OLI ratio Band2/Band5 in blue spectral range produced R2 = 0.45 and an RMSE of 0.38. In summary, the findings show that for all three water quality criteria, ASD ratios consistently performed better than OLI ratios. These generated models may help in informed decision making in water resource management, emphasizing the relevance of established spectral ratios in shaping effective strategies.
