WMOD00283

• Sponsor Name: USBR, Technical Service Center, River Systems 7 Meteorology Group
• Project Name: Use of NEXRAD WSR-88D Radar Snow Accumulations Algorith in the GCIP LSA-NC
• Title: Use of NEXRAD WSR-88D Radar Snow Accumulations Algorith in the GCIP LSA-NC
• Prepared For: USBR Technical Service Center
• Prepared By: Arlin B. Super
• Date: 1/1/1998
• State: MN
• Weather Modification - Doc Type: Report

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

3. ADEQUACY OF LEVEL III REFLECTIVITY
MEASUREMENTS FOR SWE ESTIMATION

Reclamation's SAA was designed to use the greatest resolution reflectivity
measurements available from WSR-88D radars known as Level II data (Crum et al.
1993). These 0.5-dBZ resolution measurements are transmitted from each radar
(RDA) to its Radar Product Generator (RPG), often located in National Weather
Service (NWS) Weather Forecast Offices, where many products are computed.
Level II measurements are also recorded at the RDA to large capacity tapes for
archival.

One of the RPG products is the so-called Level III reflectivity data intended for
graphical displays in up to 16 colors. Level III data likely are restricted to 16 levels
or intervals because more colors would lead to confusion when viewing monitor
displays. The Level III reflectivity resolution must be degraded with only 16
intervals available. Level III reflectivity resolution is 5.0 dBZ when the radar is
scanning in precipitation or severe weather mode and 4.0 dBZ when the clear air
mode is being used.

Both Level II and Level III data have the basic range bin spatial resolution of 1
degree by 1 km. Although the Level III product has degraded Ze resolution, it has
the considerable advantage over Level II of being available in near real time from
NEXRAD Information Dissemination System (NIDS) vendors. Level II data are
rarely available to non-NEXRAD agencies except as taped copies well after the fact.
Moreover, the cost of Level II observations for an extended period can be
substantial ($100.00 for each tape, which covers about 2 days during storms).

Th'e University Corporation for Atmospheric Research (UCAR) has archived Level
III data for the GCIP from a dozen LSA-NC WSR-88Ds during the 1996-97 winter,
and is doing so again during the current 1997-98 winter. Therefore, a snow
algorithm able to use Level III reflectivities could be used to estimate spatial SWE
fields from Upper Midwest snow storms during last winter and the current winter.

Considerable programming and testing was required to (1) modify the Level II
version of the SAA to accept Level III input, and (2) decode the Level III data
received from UCAR as individual files for each beam tilt for each volume scan.
Level III reflectivities are available only from the four lowest antenna tilts, and
UCAR archived only the two lowest tilts for much of the 1996-97 winter.
Fortunately, for the relatively flat teITain around KMPx, the SAA only needed the
two lowest tilts to function as long as the 3 km nearest the radar were ignored.

Software was written to decode each available tilt of Level III data and to combine
all available tilts into pseudo-volume scans. That is, a single file was made for each
volume scan which contained all available reflectivities from any or all of the four

8