Sensitivity of surface hydrologic fluxes predicted by a macroscale hydrologic model to energy balance closure assumptions

The Variable Infiltration Capacity (VIC) macroscale hydrology model can be run in two modes: energy or water balance. In energy balance mode, the model iterates on a surface temperature that closes the energy balance, whereas in water balance mode, the surface temperature is assumed to equal air temperature. For parameter estimation purposes, it is desirable to do parameter searches in water balance mode, but understanding the implications of model performance for a given set of parameters in energy balance mode is clearly important. Furthermore, the energy balance mode often requires shorter time steps to avoid numerical and conceptual problems associated with diurnal variations of radiative forcings, for instance. The objective of this study was to establish relationships between model performance in the different modes and temporal resolutions at which the VIC model is run. Three model configurations were analyzed; 1) Daily water balance, 2) 3 hourly water balance, and 3) 3 hourly energy balance simulations. The basic meteorological input data were daily total precipitation (uniformly apportioned within the day for subdaily time steps), and temperature maxima and minima, from which the diurnal cycle of temperature and radiative forcings was estimated. Two areas were studied; 1) A transect of 201 one-eighth degree grid cells in the central U.S., and 2) 145 grid cells within the Columbia River basin. In general, evapotranspiration decreases and total runoff increases when the model's temporal resolution was decreased, and also when going from water balance to energy balance mode. The mean runoff ratio for Area 1 increased from 0.22 in daily water balance mode to 0.28 for 3 hourly energy balance mode, and from 0.35 to 0.38 for Area 2. The difference in time step accounted for 47 and 32 percent of the difference in total runoff for Areas 1 and 2, respectively. Diagnosis of the results showed that daily average temperature and vapor pressure deficits favor transpiration, as compared with temporally disaggregated values. In energy balance mode, the iterated surface temperatures leads to lower net radiation as compared with water balance mode, and in turn reduced evapotranspiration.