The accuracy in predicting the track of tropical storms and hurricanes has increased substantially over the past decades. For example, the average error of the official National Hurricane Center 3-day forecasts has decreased from almost 400 miles in the 1970s to about 170 miles in the 2000's (through 2008). The average error of the 5-day forecasts over the past five years since they became operational is about 280 miles. These gains reflect the combined wisdom of researchers, designers of observing systems, developers of forecast models, and highly skilled forecasters. Not withstanding the advances in accuracy, however, much better is not necessarily good enough, especially for land-falling storms. Enter an additional dose of WISDOM.
The WISDOM program (Weather In Situ Deployment Optimization Method), developed by NOAA's Office of Atmospheric Research (OAR), employs "super-pressure balloons" balloons (balloons that rise to and then remain at a constant predetermined altitude) to obtain atmospheric data around the periphery of a hurricane (not in the storm itself). The idea is for the balloons to probe the winds hundreds of miles around a storm which steer a hurricane and determine its eventual track.
Keep reading for more WISDOM
The winds are obtained by determining the trajectory (change in location over a short period of time) of a small GPS locator carried by the balloons. Individual balloons can stay aloft up to a week or more and obtain data over large areas of the Atlantic basin otherwise poorly observed.
The WISDOM program's objective is to improve the hurricane track prediction in the 3 to 5 day range and enable extending forecasts to 7 days before a possible landfall. In principle, including the balloon data into the initial conditions of numerical forecast models (such as the GSM) will reduce the errors (uncertainty) of the initial conditions and lead to more skillful track forecasts generated therefrom (narrower cone of uncertainty). Maybe, maybe not.
Following a successful feasibility test during the 2008 hurricane season, plans for this year call for a WISDOM "proof of concept" effort to provide balloon observations for researchers to determine track forecast improvements that might be achieved by utilizing WISDOM data. Up to 250 WISDOM balloons are expected to be launched from a variety of locations (here) for two major hurricanes during September. Given the way this season is going, good luck just getting one to sample.
The observations will not be used in real time, but rather used in "hindcasting" experiments where forecast models are run after the fact with and without the WISDOM observations. Unfortunately, for reasons beyond the scope of this article, this is much easier said than done. The difficulties involve the details and idiosyncrasies of forecast models (model dependent results), as well as the methodologies in assessing whatever impact the data might have (I've been there, done that!). Hence, it will likely take several years to determine whether the WISDOM data provides meaningful increases in forecast accuracy.