Informing a hydrological model of the Ogooué with multi-mission remote sensing data

Cecile M. M. Kittel, Karina Nielsen, Christian Tøttrup, Peter Bauer-Gottwein
2017 Hydrology and Earth System Sciences Discussions  
Remote sensing provides a unique opportunity to inform and constrain a hydrological model and to increase its value as a decision-support tool. In this study, we applied a multi-mission approach to force, calibrate and validate a hydrological model of the ungauged Ogooué river basin in Africa with publicly available and free remote sensing observations. We used a rainfall–runoff model based on the Budyko framework coupled with a Muskingum routing approach. We parametrized the model
more » ... ng the SRTM DEM, and forced it using precipitation from two satellite-based rainfall estimates, FEWS-RFE and TRMM 3B42 v.7, and temperature from ECMWF ERA-Interim. We combined three different datasets to calibrate the model using an aggregated objective function with contributions from: (1) historical in-situ discharge observations from the period 1953-1984 at 6 locations in the basin, (2) radar altimetry measurements of river stages by Envisat and Jason-2 at 12 locations in the basin and (3) GRACE total water storage change. Additionally, we extracted CryoSat-2 observations throughout the basin using a Sentinel-1 SAR imagery water mask and used the observations for validation of the model. The use of new satellite missions, including Sentinel-1 and CryoSat-2, increased the spatial characterization of river stage. Throughout the basin, we achieved good agreement between observed and simulated discharge and river stage, with a RMSD between simulated and observed water amplitudes at virtual stations of 0.74 m for the TRMM forced model and 0.87 m for the FEWS-RFE forced model. The hydrological model also generally captures total water storage change patterns, although the amplitude of storage change is generally underestimated. By combining hydrological modelling with multi-mission remote sensing from ten different satellite missions, we obtain new information on an otherwise unstudied basin. The proposed model is the best current baseline characterisation of hydrological conditions in the Ogooué in light of the available observations.
doi:10.5194/hess-2017-549 fatcat:p7zdbwfhsree3p3v32vg37qmva