On receipt of this memo, the OSF and WSR-88D Adaptable Parameter Working Group are authorizing WSR-88D sites to use three new Precipitation Processing Subsystem (PPS) Z-R relationships to improve radar rainfall estimates during stratiform rain events. Including the new relationships, the OSF and APWG are now encouraging sites to optimize the performance of the PPS by selecting from a total of five Z-R relationships based on season, geographic location, and weather type. Guidance on selecting Z-R relationships is attached.

Although several factors can contribute to degraded PPS performance, we believe the selection of a valid Z-R relationship can provide the most significant improvement in precipitation estimates. The default WSR-88D Convective Z-R relationship (Z=300R^1.4) is most accurate during heavy rainfall associated with deep convection. In 1997, the OSF authorized sites to use a Tropical Z-R relationship (Z=250R^1.2) developed by Rosenfeld, et al. (1993). The Tropical Z-R relationship should be used to improve PPS estimates in tropical convective systems, particularly during land falling hurricanes and tropical storms.

The OSF, the NWS Office of Hydrology (OH), and NWS field offices have continued to investigate radar rainfall estimates during non-convective events. We are recommending that the Marshall-Palmer relationship (Z=200R^1.6), developed by Marshall, et al.(1955), be used to provide the best PPS estimates during general stratiform rainfall events. Several studies [Super and Holroyd (1998), Cairns, et al. (1998), Huggins and Kingsmill (1998), and Quinlan and Sinsabaugh (1999)] have shown that the best Z-R relationship in cool season stratiform rainfall events depends partly on geographic location. Based on these studies, we are authorizing sites to select from two cool season stratiform relationships, Z=130R^2.0 for sites east of the continental divide and Z=75R^2.0 for sites west of the continental divide.

The PPS uses the adaptable parameter CZM to represent the variable "a" in the Z-R relationship (Z=aR^b) and uses CZP to represent the variable "b". The following procedure should be used to modify the PPS Z-R relationship adaptable parameters CZM and CZP:

At the Unit Control Position (UCP)

1. Type AD,(password1),M,(password2),Z

2. Modify CZM - Multiplicative Z-R Coefficient to the selected value.

3. Modify CZP - Power Z-R Coefficient to the selected value.

4. Save the change.

Sites should be aware that changes to the PPS Z-R relationship can directly and significantly impact NWS River Forecast Center (RFC) operations and products. Please coordinate with and/or notify any RFCs that use your PPS products whenever your site changes these parameters.

Please let us know how these changes impact your precipitation estimates. The OSF Point of Contact for radar precipitation is Tim O'Bannon. In addition, the OSF Applications Branch would like copies of any research or studies resulting from the changes. The OSF greatly appreciates your cooperation in this process.

References:

Cairns, M., A. Huggins, and S. Vasiloff, 1998: Precipitation Algorithm improvements in the Eastern Sierra. NWS Western Region Technical Attachment No. 98-08, Salt Lake City, UT, 5 pp.

Huggins, A., and D. Kingsmill, 1998: Improvements of WSR-88D Algorithms in the Intermountain West with Applications to Flash Flood Forecasts and Wintertime QPFs. CIASTA Annual Report on Progress under Task II: Weather Research. Desert Research Institute, Dandini Research Park, Reno, NV, 26 pp.

Marshall, J. S., W. Hitschfeld, and K. L. S. Gunn, 1955: Advances in radar weather. Adv. Geophys. 2, 1-56.

Quinlan, J. S. and E. J. Sinsabaugh, 1999: An evaluation of the performance of the Snow Algorithm at NWFO Albany, NY during the 1997-98 Winter Season. 29^th Conf. on Radar Meteor., Montreal, Quebec, Canada, Amer. Meteor. Soc., 794-797.

Rosenfeld, D., D. B.Wolff, and D. Atlas, 1993: General probability-matched relations between radar reflectivity and rain rate, J. Appl. Meteor., 32, 50-72.

Super, A. and E. Holroyd III, 1998: Snow accumulation algorithm for the WSR-88D radar, Final Report. Bureau of Reclamation Report R-98-05, Denver, CO, 75 pp.

Attachment

December 14, 1999

GUIDANCE ON SELECTING Z-R RELATIONSHIPS

The OSF is authorizing sites to select from five Z-R relationships, depending on the season, geographic location, and expected weather type. Table 1 lists the Z-R relationships currently available and recommendations for selecting the best Z-R relationship for most types of precipitation events.

Table 1. Z-R RECOMMENDATIONS
RELATIONSHIP	Optimum for:	Also recommended for:
Marshall-Palmer (Z=200R1.6)	General stratiform precipitation
East-Cool Stratiform (Z=130R2.0)	Winter stratiform precipitation - east of continental divide	Orographic rain - East
West-Cool Stratiform (Z=75R2.0)	Winter stratiform precipitation - west of continental divide	Orographic rain - West
WSR-88D Convective (Z=300R1.4)	Summer deep convection	Other non-tropical convection
Rosenfeld Tropical (Z=250R1.2)	Tropical convective systems

When mixed precipitation types are present, sites should select a Z-R relationship based on the most significant and/or most widespread type of precipitation.

While selecting the appropriate Z-R relationship should improve radar rainfall estimates, several other factors can limit the performance of the WSR-88D Precipitation Processing Subsystem (PPS). Sites should be aware of these factors:

• The radar must be well calibrated to provide consistent reasonable rainfall estimates.
• Increasing the Precipitation Detection Function (PDF) Nominal Clutter Area (NCA) for light rain (Category 2) impacts the PPS detection of precipitation. Improperly high NCA values can significantly degrade rainfall estimates (see Guidance on Adaptable Parameters Handbook, Volume 4, RPG - available on the web at:

http://www.roc.noaa.gov/ssb/sysdoc/Operations.asp).

• Stratiform precipitation is generally shallow with a fairly steep vertical reflectivity gradient. This will cause the radar beam to sample rain more poorly with increasing range in stratiform events when compared with convective rainfall, and will significantly shorten the effective range for reasonable rainfall estimates.
• Cool season stratiform relationships enhance rainfall estimates from lower reflectivities (see table 2), so clutter contamination may appear to be more prominent in cool season stratiform events than in other stratiform or convective events.