Evaluation of terrestrial water storage variations in regional climate simulations over Europe using basin-scale combined water-balance data

Terrestrial water storage plays an important role in the hydrological cycle, but is an unknown in most parts of the world. There are very few large-scale measurements of its main components (soil moisture, groundwater, snow and surface water), and therefore model results cannot be validated in most regions. In this study, we analyze an ensemble of regional climate models over Europe using a derived data set of monthly variations in terrestrial water storage.

By combining the terrestrial and atmospheric water balances, the monthly changes in terrestrial water storage can be expressed as the sum of three terms: the water vapor flux convergence, the change in atmospheric water vapor content (both from ERA-40 reanalysis data), and the measured river stream flow.

This approach has been developed over the Mississippi region (Seneviratne et al. 2004) and then applied to other major mid-latitude river basins in Europe, Asia, North America and Australia to form a diagnostic data set of monthly terrestrial water storage variations (Hirschi et al. 2005). The analysis covers the full 44-year time period (1958-2001) of the ERA-40 reanalysis from ECMWF (European Centre for Medium-Range Weather Forecasts) but is temporally limited in some river basins depending on the availability of runoff data from the Global Runoff Data Center, the U.S. Geological Survey or local sources. Validation has shown good agreement between diagnosed estimates and observations of terrestrial water storage in Illinois (soil moisture, groundwater, snow observations) and Asia (soil moisture and snow observations), both in terms of the mean seasonal cycle and its interannual variations.

The derived data set has illustrated its potential for the analysis and validation of regional climate models in the Rhine basin (van den Hurk et al. 2005). Here, this work is extended to other sub-continental scale river basins in Europe (i.e. Danube, as well as compounds of French, Central and NE-European and Baltic Sea river basins), where the water budget of various regional climate models involved in the EU-project PRUDENCE is evaluated with the derived data set and direct observations of precipitation and runoff.

Results demonstrate that there are substantial differences between the models regarding the different components of the hydrological cycle.

Figure 1 shows the comparison of PRUDENCE regional climate models against the derived water-balance estimates and observed precipitation from CRU in the Danube region. Several models underestimate the summer drying in TWS, connected with an underestimation in precipitation and runoff (not shown). There are also considerable differences in winter (likely relating to the representation of snow in the models).

Hirschi, M., S. I. Seneviratne, and C. Schär, 2005: Seasonal variations in terrestrial water storage for major mid-latitude river basins. Journal of Hydrometeorology, conditionally accepted.

Seneviratne, S. I., P. Viterbo, D. Lüthi, and C. Schär, 2004: Inferring changes in terrestrial water storage using ERA-40 reanalysis data: The Mississippi river basin. Journal of Climate, 17, 2039–2057.

van den Hurk, B., M. Hirschi, C. Schär, G. Lenderink, E. van Meijgaard, A. van Ulden, B. Rockel, S. Hagemann, P. Graham, E. Kjellström, and R. Jones, 2005: Soil control on runoff response to climate change in regional climate model simulations. Journal of Climate, in press.

Fig.1: Comparison of PRUDENCE climate model runs against (a) derived water-balance estimates, and (b) CRU precipitation for the Danube basin (772 220 km²) and the period 1961-1990. The labels refer to the model names.