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

()V-I ( )
N, D] D
n(D)=- - -exp--
f(v) Dn Dn Dn

(1)

where n(D) is the number of particles of diameter D, Nt is the total number of particles, v
is the shape parameter, and Dn is some characteristic diameter of the distribution. The
Marshall-Palmer (exponential) and Khrgian-Mazin distribution functions are special
cases of this generalized function. When two moments ofa hydrometeor class are
predicted, all that is needed to completely specify the distribution function given by (1) is
the specification of v. Except for the few cases where observations can be used to
specify v, its value is chosen by trial and error or altered in sensitivity experiments.

Owing to the use oflook-up tables, it became apparent that it is no longer necessary to
constrain the system to constant or average collection efficiencies. Making use of look-
up tables, Feingold et al. (1998) replaced the formerly ad hoc autoconversion
formulations in RAMS with full stochastic collection solutions for self-collection among
cloud droplets and for rain (drizzle) drop collection of cloud droplets. The look-up tables
were then computed using realistic collection kernels rather than constant collection
efficiencies used in the past. The philosophy of bin representation of collection was also
extended to calculations of drop sedimentation. This approach has been extended to all
liquid and ice-phase hydrometeor species. Moreover, cloud droplet concentration is now
predicted from a fully prognosed CCN field (Cotton et aI., 2003). The number of CCN
that activate are a function of air temperature, Lagrangian supersaturation production rate
(related to vertical velocity and other factors) and number concentration ofCCN. Other
factors such as CCN chemistry, mean radius, and spectral width are considered fixed for
a given RAMS simulation. Based on the above CCN characteristics and environmental
factors, the fraction of CCN that nucleates into cloud droplets is accessed in RAMS from
a look-up table that was previously generated from a detailed bin-parcel model.
Prediction of CCN in RAMS includes advective and diffusive transport, source functions,
and nucleation scavenging.

Since Feingold et al (1999) have shown that the presence of GCCN moderates the effects
of CCN on cloud optical properties significantly, we have added a second mode of cloud
droplets (Saleeby and Cotton, 2003). The GCCN nucleates directly into the large droplet
mode which enhances the rate of drizzle formation. In addition, the presence of two
modes provides a more realistic simulation of the Hallett-Mossop secondary ice
formation process.

For ice-phase clouds, background ice nucleation is based on the Meyers et al. (1992)
formula which was derived from published continuous flow diffusion chamber data sets.
The formula used is

Nid = exp { a + b [ 1 OO( Si - 1 ]}

II-9

II