WMOD00447

• Title: Interagency Agreement Funds Usage Period Report - February 2001
• Date: 2/1/2001
• Weather Modification - Doc Type: Report

.

.

.

reflectivites are usually weaker than just above the surface because of the known vertical profile
of reflectivity in snow-producing clouds (Super 1998).

Five copies of the SAA executable file are simultaneously run in 5 different directories on a fast
Sun Ultra 60 workstation purchased by Reclamation for NEXRAD projects. The same effective
reflectivity (Ze)-instantaneous snow water equivalent (SWE or S) rate relationship is being used
at all 5 radars. It is commonly referred to as the Ze-R relationship when rain (R) is involved. For
the SAA, it is referred to as the Ze-S relationship. The~ Ze-S being used with each radar is Ze =
150 S**2.0, developed with special radar and precipitation gage data collected near the
Minneapolis radar during the 1996-97 winter (Super and Holroyd 1998).

A range correction scheme was developed with the same Minneapolis radar data for another GCIP
project (Super 1998), using the vertical profile of Ze measured with the lowest 5 antenna tilts in
a circle of about 35 km radius around the radar. The median vertical profile of Ze was found for
a 9 storm sample and the range correction scheme is balsed on it. A multiplication factor (MF) was
calculated, using the median vertical profile, to compemsate for the reduced Ze viewed by the
radar with increased range because of earth's curvature. Super (1998) showed that use of such
an MF was superior to ignoring the problem and using no range correction.

The range correction scheme applies the MF for SAA-E~stimated SWE using the equation: MF =
C1 + (C2 X R) + (C3 X R X R) where R is range in km and C1, C2 and C3 are empirically
determined coefficients. SAA-estimated snowfall accumulations are multiplied by the MF at and
beyond 36 km range. The coefficient values being used during the 1998-99 winter are: C1 =
+1.04607, C2 = -0.002959, and C3 = +0.0000506. Examples of resulting MF values are 1.29 at
100 km range, 1.74 at 150 km, 2.48 at 200 km and 3.04 at the maximum range of 230 km. The
maximum correction is, of course, at the maximum rangle where it amounts to a tripling of the SAA
estimated snowfall accumulation. However, the lowest tilt radar beam typically overshoots all
clouds by some range closer than 230 km. Moreover, beyond some nearer range the reflectivity
(dBZ) values fall below the minimum allowed for conversion to non-zero snowfall. Of course, any
MF applied to zero snowfall means that no range correction is applied once dBZ values fall below
the minimum value used, currently 4 dBZ.

Once SAA estimat~s have been range corrected for each 1 km X 1 degree range bin, the resulting
snow accumulation values are converted to the Hydrologic Rainfall Analysis Project (HRAP) grid
which has cells of approximately 4 km X 4 km. The NOHRSC requested that this particular grid
system be used. The approach used is to calculate thl3 location of the center of each range bin
and determine which range bins have centers within any particular HRAP cell. This information
is calculated once and thereafter extracted from a "lookup" table. The median precipitation
accumulation value is calculated for those range bins with centers within a given HRAP cell. The
median value is used to represent each HRAP cell. Tlhe only exception to this approach is the
single HRAP cell containing the radar itself. Here accumulations from the 4 range bins closest to
the cell corners are used to calculate the median. However, the SAA considers reflectivity data
to be unreliable within 3 km radius of the radar and sets all such range bins to zero precipitation.
Therefore, calculated precipitation accumulation for the~ single HRAP cell containing the radar is
typically zero or trivial.

Where an HRAP cell is within range (230 km) of more~ than one radar, the snow accumulation
value used in the mosaic or merged distribution is cllways from the nearest radar which is

3