WMOD00263 ‎(2)‎

• Applicant: CWCB
• Sponsor Name: USBR
• Project Name: Response to RFP
• Title: Numerical Simulations of Snowpack Augmentation for Drought Mitigation Studies in the Colorado Rocky Mountains
• Prepared For: USBR
• Prepared By: Joe Busto, CWCB
• Date: 8/20/2003
• Weather Modification - Doc Type: Application

-

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

accumulated precipitation forecasts was generated with both the dump-bucket and
microphysics versions of the forecast model. Both sets of output were compared to a set
of 167 community-based station reports, and to a set of 32 Snotel stations. Climatological
station precipitation forecasts were improved on the average by correcting for the
difference between a station's actual elevation and the cell-averaged topography used by
the model. The model had more problems with the precise timing and geographical
location of the precipitation features, probably due in part to the influence of other model
physics, the failure of the model to resolve adequately winter-time convection events, and
lack of mesoscale detail in the initializations. Wetzel et al. (2002,2003) further
demonstrated RAMS accuracy in predicting snowfall amounts in high-mountain terrain,
specifically the Park Range of Colorado. As in the Gaudet and Cotton (1998) study, the
best agreement occurred at the higher elevation sites and the worst in the valleys. This
could be a related to the inability of the model to represent the valley features correctly
since emphasis is placed on getting the mountain high terrain forcing in the model. In
addition, RAMS exhibited a warm-temperature bias, which may be a consequence of
using ETA model forecast data for initialization and nudging; the ETA model is known to
have such a warm temperature bias.

The current prototype real-time forecast version ofRAMS@CSU is based on version 4.3.
The model is set up on a cluster ofPCs. The forecast model configuration has three
interactive nested grids. Grid #1 has 48 km grid spacing and covers the entire
conterminous U.S. Grid #2 has 12 km grid spacing and covers all of Colorado, most of
Wyoming, and portions of adjacent states, and Grid #3 has 3 km grid spacing covering a
240 km x 240 km area that is relocatable anywhere within Grid #2. Vertical grid spacing
on all grids starts with 150-km spacing at the lowest levels and is stretched to 1000m
aloft, with a total of 36 vertical levels extending into the stratosphere. The model is
initialized with OOUTC ETA model analysis fields and run for a period of 48h, with the
lateral boundary region of the coarse grid nudged to the ETA 3-hourly forecast fields. A
48-h run typically begins about 03 UTC or 8 P.M. MST (when the 00 UTC ETA forecast
data are available), takes 4-5 h of computer time to finish, and is available for real-time
operational use by 2 A.M. MST. Because RAMS has been able to reproduce high-
elevation snow fall amounts with considerable accuracy (Gaudet and Cotton, 1998;
Wetzel et aI., 2002; 2003), we believe that RAMS can be useful in forecasting the effects
of cloud seeding on precipitation for an entire winter season.

The microphysics in RAMS is quite sophisticated. Instead of using continuous accretion
approximations as had been the norm in cloud parameterizations, RAMS uses solutions
to the full stochastic collection equation. Walko et al. (1995) describe the
implementation ofthis approach in RAMS for prediction of hydro meteor mixing ratios.
An important aspect of the implementation strategy was the use oflook-up tables that
enabled fast and accurate solutions to the collection equations. Meyers et al. (1997) then
extended this approach to two-moments of the hydrometeor spectra: mixing ratio and
number concentration. Also in Verlinde et al. (1990) and subsequent implementations in
RAMS, the Kessler-type exponential or Marshall-Palmer basis function for hydrometeor
spectra was abandoned in favor of a generalized gamma distribution function,

11-8