Precipitation Variability at High Spatial Resolution in the Desert Southwest

Andrew C. Comrie
University of Arizona
and
Bill Broyles
USFWS Volunteer, Cabeza Prieta National Wildlife Refuge

The deserts of the southwestern United States experience the most severe combination of hot and dry conditions in North America. In the region, as with most arid regions, tremendous temporal and spatial variability in precipitation is the norm. Often, appropriate management decisions for wildlife habitat, rangelands, and water resources in these areas are dependent on adequate information about this precipitation variability. Nevertheless, budgetary concerns of government agencies managing many of these areas strongly limit widespread meteorological observations using typical approaches and instrumentation (e.g., digital sensors, data loggers, computers, power sources, trained technicians, etc.). For applications requiring relatively simple kinds of data (e.g., monthly rainfall), there are alternative instrumentation choices that require only minimal financial and personnel resources.

We report results from a low-tech, low-budget network of precipitation storage gauges in and around the Cabeza Prieta National Wildlife Refuge. A storage gauge closely resembles a regular non-mechanical precipitation gauge, except that the reservoir contains a layer of mineral oil that floats on the accumulated water. Thus, evaporation is halted and the gauge need only be checked infrequently. For this study, gauges were checked and read approximately bimonthly by a United States Fish and Wildlife Service volunteer. The network is used to monitor seasonal water availability for plants and wildlife (e.g., bighorn sheep, pronghorn antelope), and it comprises about 50 stations spaced 5 to 10 miles apart across this remote and extremely arid area of southwest Arizona, with records from 1989 to present. These data provide an unusual opportunity to examine seasonal precipitation variability at a fine spatial scale for this northern portion of the Sonoran Desert, and we analyze the spatial and temporal variability of this unique set of observations.

The area experiences the summer rainy season between June and September, and the winter rainy season between November and March. April and May have little, if any, precipitation while October is a transition month. We divide the data into the four-month summer season (a brief but critical period) and the eight-month non-summer season ("winter"). We calculate seasonal and annual precipitation totals, standard deviations, and coefficients of variation, and we map the resulting climatologies as Maps of Winter Precipitation (1991-94) and Maps of Summer Precipitation (1991-94).

The data show exceptional variability: annual and seasonal precipitation vary by orders of magnitude (<1 to >10 inches in sequential years at the same station); strong spatial gradients highlight the complex effects of terrain, moisture advection, and the extreme patchiness of precipitation, even at this fine scale. Coefficients of variation show standard deviations to be up to 1.6 times greater than mean precipitation for some areas in summer. Besides providing fine-scale observations of an arid precipitation regime, the results have strong implications for natural resource management in the area, and they highlight the utility of inexpensive, low-maintenance meteorological instrumentation to provide such information.

Read the discussion comments about this topic

U.S. Department of the Interior | U.S. Geological Survey
URL: http://geochange.er.usgs.gov/sw/changes/natural/precip/
Page Contact Information: ESD Web Team
Page Last Modified: Tue 25-Nov-2003 10:48:26 MST
Accessibility, Privacy, and other policies and notices