Exploring the transferability of a land-surface hydrology model

The utility of hydrological models as a research tool is strongly linked to their ability to capture changes in regional hydrologic regimes in response to changes in climate forcing, or to simulate the hydrologic regimes of distinct climatic regions: that is, model transferability. To assess the transferability of an existing land-surface hydrology model (LSHM), three case-studies were conducted without changing the model's physical parameterizations and without the calibration of model parameters. A 1D implementation of the LSHM was used with data sets from Cabauw in the Netherlands (field plot scale, one year), and from Valdai in Russia (small catchment scale, 18 years). Simulations of runoff, latent and sensible heat fluxes, soil moisture and soil temperature, and snow accumulation and melt were compared against observations at hourly, daily, monthly, annual and inter-annual time scales. The model can reproduce well the monthly, seasonal and interannual variability of the hydrological regime in response to applied forcing, especially regarding snow accumulation patterns, and the timing and duration of melting. The results also show that, where shallow watertable fluctuations are important in determining the mechanisms of runoff generation (i.e. Valdai), the dynamic interaction between the saturated and the unsaturated zones is an essential hydrologic process that cannot be ignored. Continental-scale simulations were performed using a 1° × 1° global data set to assess the model's ability to capture seasonal cycles and interannual variability of hydrological variables across diverse climatic regions in the continental USA before and during the 1988 drought. The results obtained for the three case-studies suggest that the model can be used to study the range of variability caused by environmental change on mid-latitude hydrological regimes. Further work must be conducted for arid and semiarid regions.