| |
The use of gridded climate data is often necessary to capture the spatial and temporal variability in weather patterns when forcing hydrologic models. Several gridded climate datasets exist and differ based on the interpolation method used to grid the data, grid resolution and climate parameters available to the end user. The variability in the interpolation methods can create different values across datasets even when the same weather station data is used, which can vary in how accurately weather is reproduced on a local or regional scale. Thus, the selection of a gridded climate dataset for hydrologic modeling would depend upon the area of study, forcing an assessment of different gridded climate data’s ability to produce meaningful hydrologic output. We assess the efficacy of 3 gridded climate datasets, Maurer, Livneh and NLDAS-2, to yield reliable weather data and streamflow in the Santa Fe River basin via the Soil and Water Assessment Tool (SWAT). The Maurer and Livneh datasets have been shown to more often match observed precipitation and temperature with greater accuracy than NLDAS-2. However, the Maurer and Livneh datasets do not provide all the parameters necessary to calculate potential evapotranspiration (PET) using Penman-Monteith or Priestley-Taylor; which are the recommended PET estimation methods in Florida. Accurately estimating PET is essential for reliable hydrologic simulations. We therefore also compare the efficacy of different PET methods within SWAT, and assess the utility of the SWAT weather generator to produce reliable estimates of variables used to calculate PET when the model is forced with climate data where variables necessary to calculate PET are missing. We compare generated streamflow and weather data to observed values and assess which dataset would be more reliable for use in the Santa Fe River basin. Our results can then be used to guide gridded climate data selections for other watersheds in the region. |
