InSAR-based investigation on spatiotemporal characteristics of river sediment behavior

Abstract

For the purpose of investigating spatiotemporal characteristics of river sediment behavior, we proposed a strategy integrating the PS-InSAR technique together with gridding spatial calculation to estimate time series change of sediment volume over the Baibufan section in Daan River in Taiwan. Firstly, the PS-InSAR performance based on ascending/descending and all-season/dry-season-only images were examined to identify the most optimal Sentinel-1 image data set. Subsequently, the estimated results were validated through internal and external validations. Once the PS-InSAR results were proved reliable, we utilized the advantages of the long-term wide-extent observation to investigate the sediment behavior over sequential wet/dry seasons and relatively long hydrological seasons. Based on the estimation of sediment volume change, the most severe erosion occurred in the wet season starting from 2019/04. The erosion volume reached 1513.652 m3 and 1194.829 m3, respectively, in high and median water coverage areas. Moreover, based on the sediment maps, the sediment distribution along the river was visualized, in which the regions prone to erosion or deposition were also identified. As the strategy for investigating the detailed spatial- and temporal-wise river sediment was proved feasible, we are able to perform a regular observation to determine the deposition and erosion zones. The corresponding change of sediment volume can subsequently be estimated for an early warning of damage of hydraulic facilities or settlements along the river.