Journal article Open Access

Real-Time Implementation of Single-Doppler Radar Analysis Methods for Tropical Cyclones: Algorithm Improvements and Use with WSR-88D Display Data

Harasti, Paul R.; McAdie, Colin J.; Dodge, Peter P.; Lee, Wen-Chau; Tuttle, John; Murillo, Shirley T.; Marks, Frank D.

The NOAA/NWS/NCEP/Tropical Prediction Center/National Hurricane Center has sought techniques that use single-Doppler radar data to estimate the tropical cyclone wind field. A cooperative effort with NOAA/Atlantic Oceanographic and Meteorological Laboratory/Hurricane Research Division and NCAR has resulted in significant progress in developing a method whereby radar display data are used as a proxy for a full-resolution base data and in improving and implementing existing wind retrieval and center-finding techniques. These techniques include the ground-based velocity track display (GBVTD), tracking radar echoes by correlation (TREC). GBVTD-simplex, and the principal component analysis (PCA) methods. The GBVTD and TREC algorithms are successfully applied to the Weather Surveillance Radar-1988 Doppler (WSR-88D) display data of Hurricane Bret (1999) and Tropical Storm Barry (2001). GBVTD analyses utilized circulation center estimates provided by the GBVTD-simplex and PCA methods, whereas TREC analyses utilized wind center estimates provided by radar imagery and aircraft measurements. GBVTD results demonstrate that the use of the storm motion as a proxy for the mean wind is not always appropriate and that results are sensitive to the accuracy of the circulation center estimate. TREC results support a previous conjecture that the use of polar coordinates would produce improved wind retrievals for intense tropical cyclones. However, there is a notable effect in the results when different wind center estimates are used as the origin of coordinates. The overall conclusion is that GBVTD and TREC have the ability to retrieve the intensity of a tropical cyclone with an accuracy of ∼2 m s-1 or better if the wind intensity estimates from individual analyses are averaged together.

Files (1.7 MB)
Name Size
1.7 MB Download
Views 143
Downloads 225
Data volume 386.6 MB
Unique views 135
Unique downloads 216


Cite as