Integration of remote sensing and hydrological modeling to estimate a basin-scale sediment transport in the data-scarce Himalaya region

The Tibetan Plateau region is experiencing increasing runoff and sediment load in its headwater regions driven by the impacts of climate change, such as glacial retreat, permafrost degradation, and alterations in precipitation patterns. However, study on the changes in sediment load in the high-altitude Himalayas region remains challenging due to sparse observation data under the harsh climatic and topographic conditions. Recently, remote sensing has emerged as a promising tool in sediment studies, with several applications in the Himalayas; however, high cloud coverage during the high-flow season often leads to underestimation of sediment load. To address this issue, we introduce a remote-sensing approach to supplement the hydrological model calibration process using a less suspended sediment concentration (SSC) to quantify the long-term sediment transport in the Koshi River Basin (KRB). Landsat 8-9 OLI and the Landsat 4-5 TM images were selected to estimate Landsat-SSC with observed SSC data taken from the Chatara gauging station. Then the SWAT model was calibrated using the Landsat-SSC and validated by applying the monthly observed data from both the Chatara and Mulghat gauging stations. After model calibration and validation, the sediment load was simulated for 42 years (1981–2022). Additionally, the partial least squares-structural equation model (PLS-SEM) was used to quantify the complex relationships of sediment regimes with the potential influencing factors including hydro-climatic conditions, topographic variables, vegetation cover, and soil types. Results show that the surface reflectance of visible band combinations (R+G-B) exhibited the highest Pearson correlation with observed SSC data, allowing a power regression equation to estimate SSC from 1987 to 2022. The statistical analysis demonstrates a strong agreement between SWAT-SSC, Landsat-SSC, and Observed-SSC during calibration and validation. The annual sediment load of KRB at Chatara station is estimated at 75 Million tons (Mt) with a significant contribution during the monsoon season. The basin scale sediment load shows a significant increasing trend (p<0.01), with an average rate of 6.97 Mt/10a, which became more pronounced after 2001. PLS-SEM analysis shows that the above-considered potential influencing factors can explain 72% of the total variations, with a significant impact of hydro-climatic conditions (β=0.86, p<0.01) and vegetation cover (β=-0.56, p<0.05). The increasing sediment load in the KRB is primarily due to the strong influence of hydro-climatic changes. The negative influence of land cover changes highlights the buffering effect of increased vegetation cover on sediment export. Above all, by integrating remote sensing with hydrological modeling, this study applied new methods to estimate sediment loads with limited data and subsequently obtained critical insights into the impact of climatic and environmental changes on sediment transport, offering valuable information for soil conservation planning in the data-scarce Himalayan region.