||<p>While very important from an economical and societal point of view, estimating
precipitation in the western United States remains an unsolved and challenging problem.
This is due to difficulties in observing and modeling precipitation in complex terrain.
This research examines this issue by (i) providing a systematic evaluation of precipitation
observations to quantify data uncertainty and (ii) investigating the ability of the
Ocean-Land-Atmosphere Model (OLAM) to simulate the winter hydroclimate in this region.
This state-of-the-art, non-hydrostatic model has the capability of simulating simultaneously
all scales of motions at various resolutions.</p><p>This research intercompares nine
precipitation datasets commonly used in hydrometeorological research in two ways.
First, using principal component analysis, a precipitation climatology is conducted
for the western U.S. from which five unique precipitation climates are identified.
From this analysis, data uncertainty is shown to be primarily due to differences in
(i) precipitation over the Rocky Mountains, (ii) the eastward wet-to-dry precipitation
gradient during the cold season, (iii) the North American Monsoon signal, and (iv)
precipitation in the desert southwest during spring and summer. The second intercomparison
uses these five precipitation regions to provide location-specific assessments of
uncertainty, which is shown to be dependent on season and location.</p><p>Long-range
weather forecasts on the order of a season are important for water-scarce regions
such as the western U.S. The modeling component of this research looks at the ability
of the OLAM to simulate the hydroclimate in the western U.S. during the winter of
1999. Six global simulations are run, each with a different spatial resolution over
the western U.S. (360 km down to 11 km). For this study, OLAM is configured as for
a long-range seasonal hindcast but with observed sea surface temperatures. OLAM precipitation
compares well against observations, and is generally within the range of data uncertainty.
Observed and simulated synoptic meteorological conditions are examined during the
wettest and driest events. OLAM is shown to reproduce the appropriate anomaly fields,
which is encouraging since it demonstrates the capability of a global climate model,
driven only by SSTs and initial conditions, to represent meteorological features associated
with daily precipitation variability.</p>