Assessing water storage variability in large exorheic river basins from global hydrological models over 1948–2016

• Main Author: Cáceres, D.
• Format: Book
• Language: English
• Published: IIASA 2020
• Subjects: Greenland; Antarc*; Antarctica
• Online Access: http://pure.iiasa.ac.at/id/eprint/16916/; http://pure.iiasa.ac.at/id/eprint/16916/1/Report_Caceres.pdf

Total land water storage anomalies (TWSAs) give valuable insights as to the hydrological behaviour of a basin and, as such, constitute one of the basin signatures. However, to this day, an accurate assessment of this hydrological variable remains a challenge. The Gravity Recovery and Climate Experiment (GRACE) mission enabled its monitoring at global scale. However, these observations are limited in time and do not give information as to the individual components and drivers of TWSAs. This variable can also be estimated by the means of global hydrological models (GHMs). Here, we did a long-term assessment (1948–2016) of TWSAs in 17 large exorheic basins worldwide using two state-of-the-art GHMs, namely the Community Water Model (CWatM) and WaterGAP. TWSAs were decomposed into individual mass components by the means of the models, and into individual temporal signals (seasonal, linear trend and interannual) by the means of harmonic analysis. In a first instance, the models were validated against GRACE-based TWSAs aggregated over the global continental area (except Greenland and Antarctica) for the time period 2002–2016. Model evaluation was also performed against in situ streamflow observations and showed a highly variable model performance depending on the model and gauging station considered. The long-term assessment of TWSAs and individual mass and temporal components revealed the importance of seasonal and annual water storage fluctuations in the soil, groundwater and river compartments. The comparison between CWatM and WGHM showed significant model discrepancy with respect to anomalies in these water storage compartments. These discrepancies are presumably related to different model parameterization as to maximum soil water storage capacity, groundwater recharge and groundwater abstractions for human use, and to the fact that WaterGAP was calibrated against observed mean streamflow. Furthermore, we found an El Niño-Southern Oscillation (ENSO) signature in the interannual signal of TWSAs by the means of a correlation analysis against the Multivariate ENSO Index (MEI).