Estimating water volume variations in lakes and reservoirs from four operational satellite altimetry databases and satellite imagery data

• Four satellite altimetry products were evaluated for three lakes.
• Lake volume variations were estimated from only satellite data.
• Satellite water level and volume variations were validated with in-situ data.
• The relative RMSE were within 4.6–13.1% for the estimated volumes.

Water levels in lakes and reservoirs can currently be obtained from four different satellite altimetry databases: (i) Global Reservoir and Lake Monitoring (GRLM), (ii) River Lake Hydrology (RLH), (iii) Hydroweb and (iv) ICESat-GLAS level 2 Global Land Surface Altimetry data (ICESat-GLAS). This paper proposes a new method for estimating water volume changes in lakes and reservoirs from these four databases in combination with satellite imagery data, without any in-situ measurements and bathymetry maps. Three lakes/reservoirs with different characteristics were studied, i.e. Lake Mead (U.S.A.), Lake Tana (Ethiopia) and Lake IJssel (The Netherlands). Compared to in-situ water levels, satellite altimetry products provided accurate water level variations for Lake Mead and Lake Tana but not for Lake IJssel. The long-term lowest water level in each satellite altimetry database was used as the reference level for water volume estimation. All water levels were converted to the Water Level Above the Lowest Level (WLALL), and the series of Landsat TM/ETM + imagery data were selected to extract corresponding surface areas for establishing area–WLALL relationships. Subsequently, the relationships of the Water Volume Above the Lowest water Level (WVALL) and WLALL were obtained through the analytical integration of area–WLALL relationships. The WVALL–WLALL relationships are site-specific and database-specific and can be used to convert water levels from the four databases directly into water volumes above the identified minimum levels for the same lake. Validation showed that estimated water volumes agreed well with in-situ measurements (R2 from 0.95 to 0.99) and the root mean square error (RMSE) was within 4.6 to 13.1% of the mean volumes of in-situ measurements.