WMOD00535

• Title: Regional Model Simulation of an Extreme Precipitation Event in the Colorado River Basin
• Date: 1/5/1992
• Weather Modification - Doc Type: Report

Pt~se,ted at the Third AMS Symposium on Global Change Studies,
_. ~Iluary 5.10, 1992, AtIaDta, GA.
~

REGIONAl; MODEL SIMULATION OF AN EXTREME
PRECIPITATION EVENT IN THE COLORADO RIVER BASIN

by

David Matthews
U. S. Bureau of Reclamation,
P. O. Box 25007, Denver CO 80225

and
Filippo Giorgi and Gary Bates
National Center for Atmospheric Research1, Boulder CO 80301

1. INTRODUCTION

The Global Climate Change Response Program
(GCCRP) of the U. S. Bureau of Reclamation (Reclamation)
is concerned with the possible impacts of global climate
change upon precipitation, evapotranspiration, and
streamflow in the Western United States. Dennis (1991)
has described Reclamation's strategy to develop scenarios of
precipitation under changed climate. Matthews et al. (1991)
described initial efforts to link: a regional climate model to a
local-scale precipitation model (Medina 1991). Our paper
presents an examination of conditions that produced
unusually heavy precipitation in the Colorado River Basin
during the spring of 1983, which produced one of the largest
flows of record and was associated with an unusually strong
El Nino event described by Shea et al. (1991). Meehl
(1990) described a general circulation model's (GCM's)
ability to simulate the seasonal cycle forcing of the El Nino
in a global-coupled ocean-atmosphere model. Recent
experiments by Meehl et al. (1992) using this model indicate
that the El Nino-Southern Oscillation (ENSO) phenomenon
continues to function in doubled-C02 model simulations.
The goal of this study is to evaluate a regional model's
ability to simulate the mesosynoptic and long-wave features
that contributed to the 1983 extreme event, which may be
identifiable in future climate simulations. These features
may be used to describe the frequency and relative intensity
of such extreme events in. altered climate conditions as
estimated from doubled-C02 GCM simulations.
Reclamation's interest focuses on the linkage between large-
scale GCMs used in global climate simulations and a simple
local-scale precipitation model described by Medina (1991)
and Rhea (1977). Such a nested system could provide
objective quantitative distributions of precipitation from
altered climate conditions over individual watersheds that
affect Reclamation reservoirs, hydroelectric power and
irrigation systems.

This paper provides a brief report on the initial
efforts to determine the feasibility of using the nested
regional model outputs to describe regional controls of
precipitation in an extreme. seasonal event. The regional
structure of storms and its effect on precipitation is described
for the Colorado River Basin focusing on the upper reaches
of the Gunnison River in the Colorado Rocky Mountains.

2. REGIONAL MODELING FRAMEWORK

the NCAR 'CCM to provide the large-scale circulation
response to global climate f~rcing. Then the regional-sc3;le
MM4 model is used to descnbe the effect of sub-GCM gnd
scale forcing over a specific region. This forcing includes
the effects of complex topography, large bodies of water and
surface vegetation characteristics that may significantly affect
local climates.

The standard MM4 model is described in Anthes et
al. (1987); however, in this study we have ~sed.results from
the augmented version described by Giorgi and Bates
(1989). The augmented version of the MM4 model includes
an improved sophisticated surface physics-~oil hydrology
package, an explicit bound~ .layer formul~lon, and a more
detailed treatment of radiative transfer in place of the
corresponding standard MM4 formul~tion~. '~e model.is a
compressible and hydrostatic model Wlth pnrmttve eq~attons
written in terrain-varying sigma (s)-vertlcal coordmates,
where S=(P-PI)/P" P'=P. - PI , p. is the surface pressure
and PI is the pressure at the top of the model. The MM4
model design and characteristics are dis(:ussed in two
previous papers in this section, Giorgi et ~., (1992), ~ates
and Giorll:i, (1992); therefore, the model is not descnbed
further here.

In order to evaluate the MM4 performlUlcc, especially
with respect to regional orographic precipitation, the model
was run with European Center for Medium Range Forecast's
(ECMWF) large-scale analysis of data from January 1982 -
December 1983. The ECMWF analysis was used as initial
and lateral boundary conditions for MM4. It had a temporal
resolution of 12 hours and a spatial resolution of 2.80 latitude
and longitude. The MM4 model domain is centered over
the Great Basin at 400 N and 1160 W with a size of 3300 x
3000 lan, having a horizontal grid spacing of 60 Ian and 15
sigma-levels in the simulations presented here. Figure 1
shows the topography used in the model in these
simulations. Giorgi et al. (1992) describe the details of the
MM4 simulations for January 1982-December 1983 and
January 1988-April 1989, which are used in this paper.
They show that the MM4's predictions, of regional
wintenime precipitation and hydrologic. bu~gets were quite
realistic from the perspective of a synoptic climatology.

The next section shows two examples of regional
features analyzed by MM4 which controlled the intensity,
duration and location of precipitation during the spring of
1983. These features were clearly defined and provide a
basis for confidence in MM4's ability to simulate regionalaspects of future climate.

Recently, the climate modeling team at NCAR has
succeeded in nesting the Pennsylvania State
University/NCAR Mesoscale Model (MM4) into the NCAR
Community Climate Model (CCM) for regional climate
simulations. The first stages of developing the nested model
system are described by Dickinson et al. (1989), Giorgi et
al. (1989), Giorgi and Bates (1989) and Giorgi (1990).
The one-way nested technique uses the GCM results from

1 The National Center for Atmospheric Research is sponsored by the U. S. A. National Science Foundation

1