Summary Results from:

A high-resolution bathymetry dataset for global reservoirs using multi-source satellite imagery and altimetry
As they relate to the validation of MOD28

Authors: Li, Y., Gao, H., Zhao, G., Tseng, K.H.

Source: Remote Sensing of Environment, 2020, 244, 111831

Link to: Access Publication


Although accurate 3-D reservoir bathymetry is a key input for multiple applications (such as global hydrological models, local water resources related studies, and others), such information is only available for very few locations. To fill this knowledge gap, we generated a 30 m resolution bathymetry dataset for 347 global reservoirs, representing a total volume of 3123 km3 (50% of the global reservoir capacity). First, Area-Elevation (A-E) relationships for the identified reservoirs were derived by combining altimetry data from multiple satellites with Landsat imagery data. Next, the resulting A-E relationships were applied to the Surface Water Occurrence (SWO) data from the JRC Global Surface Water (GSW) dataset to obtain bathymetry values for the dynamic areas of the reservoirs. Lastly, an extrapolation method was adopted to help achieve the full bathymetry dataset. The remotely sensed bathymetry results were primarily validated against the following: (1) They were validated against Area-Elevation (A-E) and Elevation-Volume (E-V) relationships derived from the in situ elevations and volumes for 16 reservoirs, with root-mean-square error (RMSE) values of elevation from 0.06 m to 1.99 m, and normalized RMSE values of storage from 0.56% to 4.40%. (2) They were also validated against survey bathymetric maps for four reservoirs, with R2 values from 0.82 to 0.99 and RMSE values from 0.13 m to 2.31 m. The projected portions have relatively large errors and uncertainties (compared to the remotely sensed portions) because the extrapolated elevations cannot fully capture the underwater topography. Overall, this approach performs better for reservoirs with a large dynamic area fractions. It can also be applied to small reservoirs (e.g., reservoirs with surface areas of a few square kilometers or less), where ICESat observations are available, and to large natural lakes. With the contribution of ICESat-2, this dataset has the potential to be expanded to thousands of reservoirs and lakes in the future.