A new method for point pollution source identification based on remote sensing inversion and water streamlines

The accurate identification of the location and released mass of pollution sources using remote sensing technology is essential for emergency responses and environmental conservation. However, pollution source identification is still difficult to perform effectively when based on remote sensing technology because of the uncertainty of the inverse problem. Based on a physical mechanism, a point pollution source identification method combined with remote sensing inversion and water streamlines is proposed. This approach is designed to attempt to provide a deterministic solution for the uncertain inverse problem through clustering estimations along streamlines. The method is carried out based on the reverse flow field, which is simulated by a twodimensional hydrodynamic model. The spatial distribution of the water quality parameters obtained from remote sensing inversion is transformed into estimates of released pollutant mass. Hypothetical cases and realworld cases are conducted to validate the proposed method. In hypothetical cases, the relative errors of the identified released mass for continuous release cases are less than 5%, and the relative error for the instantaneous release case is 1.54%. In the real-world case, the relative errors of identified pollution source locations are less than 3.5%. The relative errors of the identified released mass of the two pollution sources are 5.50% and 4.42%, respectively. The results show that this method can provide high-accuracy identification results for the location and released mass of pollution sources without requiring observed water quality data. It will help improve the water environment in areas where data of pollution sources are lacking.