LOCA-VIC runs

The LOCA meteorological data (daily Tmin, Tmax, and precipitation) were used to force the Variable Infiltration Capacity land surface model, which produces such quantities as runoff, snow cover, surface long- and short-wave radiation, soil moisture, etc. on the 16th degree LOCA grid. Note in particular that VIC estimates near-surface humidity from the time history of supplied temperature and precipitation, which is a different product than the directly LOCA-downscaled GCM humidity fields described here. Some discussion of whether the VIC-estimated humidity can be used for climate change simulations can be found in Pierce et al. 2013. The brief answer is that the VIC-estimated humidities appear to have problems capturing long-term trends. However, not all GCMs in the CMIP5 archive saved daily near-surface humidity, and so were not able to be downscaled directly.

Currently the VIC output is only available over the California-Nevada domain at the Cal-Adapt site, even though the VIC runs covered the entire LOCA domain from central Mexico through southern Canada. Releasing more of the data awaits vetting by the U.S. Army Corps of Engineers, sponsors of the extended domain effort.

The LOCA-VIC runs were started about the same time that VIC version 5 was first released. Because that was the first release after a major change to VIC’s software architecture, there were concerns it might have bugs. It was therefore decided to use the last pre-version 5 release of VIC for the LOCA-VIC runs, which at the time was version 4.2.c.

Small modifications to the VIC 4.2.c code were made to treat all grid cells as existing in the same time zone. Without this modification, the daily LOCA input data was split across different days in the VIC simulation, resulting in the odd effect that the daily averaged VIC output precipitation field did not match the daily averaged precipitation input field. With the imposition of a constant time zone, the daily averaged VIC output matches the daily averaged VIC input for the forcing meteorological variables.

VIC requires at a minimum daily Tmax, Tmin, precipitation, and wind speed. For wind speed, we used the same daily climatology for all GCMs, based on the CFS reanalysis wind speeds over the period 1980-1984. This relatively short period was used to try to capture some day-to-day variability in wind speed, although obviously, being a daily climatology, it is not synchronized to the actual weather patterns in the GCMs. Our sensitivity tests have indicated that VIC’s sensitivity to wind speed is modest.

The VIC parameter files were kindly supplied by Ben Livneh at the University of Colorado, and can be downloaded directly from his site. A local copy of the parameter files as used in the LOCA-VIC runs, including the daily wind speed climatology, can be found here.