The 8th Conference on Aviation, Range, and Aerospace Meteorology


M M. Wolfson, MIT Lincoln Lab, Lexington, MA; and B. E. Forman, R. G. Hallowell, and M. P. Moore

The Growth & Decay Storm Tracker* + **

M.M. Wolfson, B.E. Forman, R.G. Hallowell, and M.P. Moore
Massachusetts Institute of Technology
Lincoln Laboratory
Lexington, Massachusetts 02420-9185

A new method for tracking storms that accounts for systematic growth and decay has been developed by MIT Lincoln Laboratory under the FAA Aviation Weather Research Program's Convective Weather Product Development Team. Predicting convective weather is extremely important to aviation since at least half of the national airspace delay is caused by thunderstorms. Accurate one-hour convective weather forecasts meet critical terminal traffic planning needs of the TRACON and ARTCC supervisors and traffic managers.

The new method, dubbed the "Growth and Decay Tracker," extracts and tracks the large scale signal in the radar data and uses the resultant vector field to advect the original weather up to one hour ahead. Wilson (1966) showed that the large scale features in the atmosphere are inherently more persistent with time than small scale features, and he reminded us of this in his invited speech at the Convective Weather Forecasting Workshop (7th Aviation Wx Conference). Browning (1979) used a large scale filtering step in his FRONTIERS forecasting process by degrading the data from 5x5 km pixels to 20x20 km, and reintroducing some major storm cores after the patches had been advected.

Our approach uses a 5x21 pixel (4x4 km pixels) elliptical filter to extract the large scale signal. This allows us to extract long narrow large scale regions (important since storms tend to organize into linear elements) without completely eliminating them while still achieving a high degree of filtering (84 km long-front). We then use a correlation tracker to determine the motion of the resultant large scale over a 6-min interval. This motion field is used to advect the original image one hour ahead. The table below shows the performance of this technique relative to conventional correlation tracking.

       60 Minute Forecast CSI Statistics
    (4 km NEXRAD long-range VIL data)
Line Storms                           Airmass Storms
G&D Tracker 40%                 G&D Tracker 18%
Conventional Tracker 27%    Conventional Tracker 17%
----------------- ----------------
% Improvement +48%           % Improvement +5%

This new technique automatically tracks the storm "envelope" instead of the individual cells, which has been a classic problem in radar meteorology. By effectively tracking the storm forcing, we account for systematic growth and subsequent decay. The technique has proved so successful it is being demonstrated in real time this year at Dallas/Fort Worth International Airport as part of the Integrated Terminal Weather System prototype operating there and in a national demonstration operated for a limited number of airlines by the National Center for Atmospheric Research. Also, the National Severe Storms Laboratory is investigating this technique for possible inclusion in future builds of the NEXRAD software.


Browning, K.A., 1979: The FRONTIERS plan: a strategy for using radar and satellite imagery for very-short-range precipitation forecasting. The Meteorological Magazine, 108, 161-184.

Wilson, J.W., 1966: Movement and predictability of radar echoes. NSSL, Tech Memo No. 28, Norman, OK.

* This work was sponsored by the Federal Aviation Administration. The views expressed are those of the authors and do not reflect the official policy or position of the U.S. Government.

+ Opinions, interpretations, conclusions and recommendations are those of the authors and are not necessarily endorsed by the United States Air Force.

** Patent applied for by Massachusetts Institute of Technology

The 8th Conference on Aviation, Range, and Aerospace Meteorology