The present study proposes a suitable method to monitor sandbars using satellite-based information. The target area is Jamuna River, Bangladesh where sandbars often form and deform both in space and time. The authors employ threshold and gradient based methods to identify sandbar behavior. The Threshold-Based Method (TBM) extracts water and land by computing the Modified Land Soil Water Index (MLSWI) using MODIS 8-day composite dataset. The Gradient-Based Method (GBM) focuses on the edges between the surface types and produces comparatively better results using MODIS daily surface reflectance and without any threshold. The results suggest that the sandbars' sizes and shapes identified by the proposed method closely coincide with the field investigation.
展开▼
