Objective

The Office of the Federal Coordinator for Meteorology and the WMO have established statements recommending the development of tropical cyclone algorithms for use with landbased Doppler radars, and in particular WSR-88D radars:

• Action 93-2.2: a. Request Working Group on Hurricanes and Winter Storms Operations and/or NWS/Office of Meteorology seek approval to implement specialized operational modes or settings to optimize NEXRAD (WSR-88D) utility for tropical cyclones (p. 3, OFCM, 1993)

• Basic Research Priority Recommendations of IWTC-III: C.3 (III) Develop Doppler radar algorithms specifically designed for tropical cyclone situations (p. 11, WMO, 1995).

We are organizing an international working group (Australia, Canada, France, India, Japan, Taiwan, and the United States) to develop and systematically test new and/or existing single Doppler radar algorithms for tropical cyclone detection, tracking, and wind field analysis.

Rationale

A two phase process is envisioned to develop operational tropical cyclone algorithms. Phase I entails the organization of a working group of scientists interested in the problem of developing tropical cyclone algorithms for single Doppler radars. Participation in the working group is voluntary, as we have no funding source ready to sponsor the development of new algorithm ideas. However, in order develop and test ideas for possible tropical cyclone algorithms members of the working group would be given free access to:

• Statement of TPC/NHC's requirements for tropical cyclone description via landbased Doppler radar,
• Common WSR-88D Level II data set on a workstation at TPC/NHC that can be accessed via Internet with which to develop and test algorithms, and
• Documentation and software to read the WSR-88D Level II data.

The working group will meet periodically (every 6 mon.) to discuss progress and make plans for future efforts.

During Phase II proposed algorithms will be put on a common workstation at TPC/NHC and undergo detailed testing using the whole 1995 WSR-88D data set. At the end of Phase II we should have a short list of algorithms for the Operations Support Facility (OSF) to implement. It is hoped that Phase II efforts will be eligible for funding under NOAA or the Office of the Federal Coordinator.

Algorithms

• Center finding-reflectivity (TREC- Tuttle and Gall, Chaser- Griffin et al) and velocity (Wood and Brown, extended VAP- Harasti and List)
  
• Center tracking (combination of centers in time)
  
• Wind field description-various approaches are possible: kinematic properties of the vortex to actual estimate of the winds (GBVTD- Lee et al, TREC- Tuttle and Gall, GBEVTD- Roux et al, extended VAP- Harasti and List, VAD based schemes-Donaldson et al)
  
• Precipitation (need specific one for tropical cyclones)
  
• Loral Data Systems algorithm suite (P. Ray, et al)
  
• Storm Surge (available in Loral suite)

We must also examine existing WSR-88D algorithms (e.g., mesocyclone, TVS, rainfall, etc.) for use in tropical cyclone conditions:

• How well do they perform?
• Can they be improved for use in tropical cyclones?
• Do we need separate versions for use in tropical cyclones?

Dave Sharp (SOO, NWS/KMLB) will coordinate an effort to evaluate current algorithm performance in tropical cyclones with Colin McAdie (TPC/NHC), Stacy Stewart (OSF) and other interested SOOs.

Approach

In 1995 Hurricanes Erin, Jerry, and Opal were observed by WSR-88D radars over Florida and surrounding waters. At the same time a number of complimentary data sets were collected by the NOAA WP-3D airborne radars, NCAR CP-2 radar, and AFRES WC-130 aircraft. These data sets are unique, including measurements from multiple platforms and instruments. It provides an opportunity to systematically test and certify many of the algorithms and provides a basis for developing new algorithms.

In Phase I, 2 h of WSR-88D level II data from Hurricane Erin are available on an FTP site for use in algorithm development. Erin was chosen to take advantage of the excellent WSR-88D coverage of the storm during its lifetime as it made landfall along the east coast and panhandle of Florida.

The first hour of data (10-11 volume scans) are from the Melbourne, FL WSR-88D (KMLB), 0600-0700 UTC, 2 August 1995, during Erin's landfall along the east coast of Florida (MPEGs of KMLB base reflectivity or base velocity for this time are available). The second hour of data (10-11 volume scans) are from the Mobile, AL WSR-88D (KMOB), 1600-1700 UTC, 3 August 1995, during Erin's second landfall along the Florida panhandle . The volume scans are stored in the WSR-88D level II format. Documentation on the level II data format and subroutines to decode the data are also on the FTP site.

KMLB	KMOB

We also plan to include an analog data set of a model wind field sampled from a simulated WSR-88D (e.g., Rankin vortex, divergent flow, combinations, etc.), for which we know the properties of the flow. The model data set would be a useful tool for algorithm development, and for intercomparison of results from the different algorithms. We are constructing such an analog data set and make it available as soon as possible.

Roughly 6 months after the data is made available via FTP the working group will reconvene at a convenient location and time to evaluate algorithm development and performance. The results from all the algorithms will be compared to independent data sources. A nice aspect of the Erin data set is that a number of independent data sources are available to assess the storm's position, track, kinematic structure, and rainfall. These independent data sets include:

• NOAA WP-3D airborne Doppler data sets from Erin
  
• HRD mesoscale surface wind analysis of reconnaissance and other surface data
  
• HRD WSR-74S data set in Erin from Palm Beach, FL NWS Office (PBI)
  
• NCAR CP-2 data set in Erin from Titusville, FL

The first three data sets are available at HRD and can be used to evaluate the algorithm performance.

What's New

The Working Group met most recently during the 22nd Conference on Hurricanes and Tropical
Meteorology, in Ft. Collins, CO. click here for minutes

Minutes of Vail Meeting 11 October 1995 (1 December 1995)

References

Donaldson, R., 1991: A proposed technique for diagnosis by radar of hurricane structure. J. Appl. Meteor., 30, 1636-1645.

Donaldson, R., F.I. Harris, and D. Smalley, 1995: An approach toward estimation of hurricane radial wind speed. Proceedings of the 27th Conference on Radar Meteorology. AMS, Vail, CO, 218-220.

Harasti, P., and R. List, 1995: A single-Doppler radar analysis method for intense axisymmetric cyclones. Proceedings of the 27th Conference on Radar Meteorology. AMS, Vail, CO, 221-226.

Lee, W.-C., B. Jou, B.-L. Chang, 1993: Nowcasting typhoon circulation using Doppler radar - the ground-based VTD (GBVTD) technique. Proceedings of the 26th Conference on Radar Meteorology. AMS, Norman, OK, 83-87.

OFCM, 1993: Record of Actions 93-2 Meeting. AHG/TCR Memorandum 93-4, San Antonio, TX, July 1993.

Griffin, J., R. Burpee, F. Marks, And J. Franklin, 1992: Real-time airborne analysis of aircraft data supporting operational hurricane forecasting. Wea. Forecasting, 7, 480-490.

Ray, P., L. Lemon, D. Miller, and B. Whissel, 1991: Operational Doppler radar in the tropics. Proceedings of the 25th Conference on Radar Meteorology. AMS, Paris, France, 35-38.

Tuttle, J., and R. Gall, 1995: Radar analysis of Hurricanes Andrew and Hugo. Proceedings of the 21st Conference on Hurricanes and Tropical Meteorology. AMS, Miami, FL, 608-610.

WMO, 1995: Proceedings of the Third WMO/ICSU International Workshop on Tropical Cyclones (IWTC-III), Huatulco, Mexico, WMO/TD No. 624, 22 Nov. - 1 Dec. 1993.

Wood, V., and R. Brown, 1992: Effects of radar proximity on single-Doppler velocity signatures of axisymmetric rotation and divergence. Mon. Wea. Rev., 120, 2798-2807.

---