Articles | Volume 15, issue 4
Hydrol. Earth Syst. Sci., 15, 1257–1271, 2011
Hydrol. Earth Syst. Sci., 15, 1257–1271, 2011

  20 Apr 2011

20 Apr 2011

Seasonal evaluation of the land surface scheme HTESSEL against remote sensing derived energy fluxes of the Transdanubian region in Hungary

E. L. Wipfler1, K. Metselaar1, J. C. van Dam1, R. A. Feddes1, E. van Meijgaard2, L. H. van Ulft2, B. van den Hurk2,3, S. J. Zwart4,5, and W. G. M. Bastiaanssen4,5 E. L. Wipfler et al.
  • 1Wageningen University, Soil Physics, Ecohydrology and Groundwater Management Group, Wageningen, The Netherlands
  • 2Royal Netherlands Meteorological Institute, the Bilt, The Netherlands
  • 3Institute for Marine and Atmospheric Research, Utrecht, The Netherlands
  • 4WaterWatch, Wageningen, The Netherlands
  • 5Delft University of Technology, Department of Water Management, Delft, The Netherlands

Abstract. The skill of the land surface model HTESSEL is assessed to reproduce evaporation in response to land surface characteristics and atmospheric forcing, both being spatially variable. Evaporation estimates for the 2005 growing season are inferred from satellite observations of the Western part of Hungary and compared to model outcomes. Atmospheric forcings are obtained from a hindcast run with the Regional Climate Model RACMO2. Although HTESSEL slightly underpredicts the seasonal evaporative fraction as compared to satellite estimates, the mean, 10th and 90th percentile of this variable are of the same magnitude as the satellite observations. The initial water as stored in the soil and snow layer does not have a significant effect on the statistical properties of the evaporative fraction. However, the spatial distribution of the initial soil and snow water significantly affects the spatial distribution of the calculated evaporative fraction and the models ability to reproduce evaporation correctly in low precipitation areas in the considered region. HTESSEL performs weaker in dryer areas. In Western Hungary these areas are situated in the Danube valley, which is partly covered by irrigated cropland and which also may be affected by shallow groundwater. Incorporating (lateral) groundwater flow and irrigation, processes that are not included now, may improve HTESSELs ability to predict evaporation correctly. Evaluation of the model skills using other test areas and larger evaluation periods is needed to confirm the results.

Based on earlier sensitivity analysis, the effect of a number of modifications to HTESSEL has been assessed. A more physically based reduction function for dry soils has been introduced, the soil depth is made variable and the effect of swallow groundwater included. However, the combined modification does not lead to a significantly improved performance of HTESSEL.