Climate of the Reynolds Creek Experimental Watershed ranges from arid to temperate. Average annual precipitation ranges from less than 25 cm in the northeast part of the watershed to more than 115 cm at the highest elevations in the southwest. Annual snowfall varies from about 20% of annual precipitation in the lower elevations to more than 70% at the higher elevations. At the higher elevations, the wind interacts with local terrain features to redistribute about two-thirds of this snow to deep drifts on northeast-facing slopes. Annual precipitation in the irrigated areas ranges from 25 to 38 cm with 40% falling from April through September. Irrigation water, diverted from Reynolds Creek, depends mostly on spring runoff from melting snow. The irrigation season, depending on individual water rights, rarely lasts beyond mid-July.
During July, August, and September, most of the basin is subject to convection storms (cloudbursts) of small areal extent with intensities recorded in excess of 25 cm per hour. These storms have resulted in severe erosion in areas of sparse vegetation. However, major floods and erosion of record were caused by winter storms with rain, soil frost and snowmelt.
The average annual air temperature is more than 7.2 °C in the valley below about 1525 meters elevation and about 3.9 °C at 1830 meters; less than 3.9 °C above this elevation. The average freeze-free season in the valley bottom ranges from 90 to 110 days becoming shorter as the elevation increases. Because of the relatively short growing season crops are limited to hay, grain, and pasture.
Two kinds of meteorology data are available. Provisional, or “L1” data, have undergone preliminary quality control to remove obvious errors, which are replaced with a missing data indicators so that the data are serially complete. Processing is done on a routine basis such that the previous water year data are available each year in March. Model Reference data, designated as “L2” data, have received additional scrutiny to correct for more subtle errors data errors and are temporally continuous so that missing data have been estimated and filled in. These data are designed to provide forcing data for model applications. This processing is done on a more ad hoc basis, the latest suite of data is processed through water year 2008.