Water quality monitoring using modeling of suspended sediment estimation (a case study: sefidroud river in Northern Iran)

Abstract

The Sefidroud River, the second largest river in Iran, is located in the north. Since the operation of the Sefidroud (Manjil) dam on the said river, about half of the storage volume of the dam has decreased as a result of the accumulation of sediments. The present research, using 516 suspended sediment data from four regional sediment monitoring stations, was conducted between 2013 and 2020 to check the accuracy of single-linear, bi-linear and middle-class sediment rating curves (SRCs) of four sediment monitoring stations as well as Landsat 8 images to estimate suspended sediment concentration (SSC). After drawing the SRCs based on 46 satellite images and SSC data, 70% of samples were used to prepare the regression models of spectral data versus suspended sediment discharge (Qs) and 30% of samples to evaluate the accuracy of SRC and Landsat 8 data. According to results, the middle-class SRCs had the highest coefficient of determination (R2, exponential). Four band ratios B4/B3, B4/B2, B6/B5 and B7/B5 had exponential and power correlation with Qs, with the highest value for the band ratio B4/B3 (R2 = 0.74, exponential). To conclude, the results of the current research showed that the B4/B3 band ratio was more efficient for Qs estimation.