Wind Model

"Where is this balloon going to go?"


Wind modeling was done using the University of Wyoming, Department of Atmospheric Science, Balloon Trajectory Forecast. The modeler exports a three-dimensional Google Earth KML file or a text file, so the anticipated trajectory was easily modeled. Data from the National Center for Atmospheric Research was used to check the accuracy of the trajectory model and as a better tool for visualizing the wind patterns.


How far the balloon would typically go, and where a good launch and landing location might be located, was something that needed to be identified early in the process. Two test-launch locations, dubbed "Northside" and "Tussey," were selected based because they were a reasonable driving distance from home, and a reasonable distance from restricted air space. These locations were modeled almost daily from 12/12/09 to 05/13/10 at 12 o'clock zulu (7AM EST/8AM EDT.) Click on the images below to download the Google Earth kmz files used to create the images.



Although it had a few acceptable trajectories, Northside often infringed upon restricted airspace in Philadelphia, Atlantic City, and New York; crashed into the ocean numerous times; and was clearly too close to home. Not an ideal launch location.



Tussey frequently stayed away from urban areas except when the wind patterns blew towards the south. Having done this exercise, I understood approximately how far the balloon would travel on a given day based on the wind speeds and direction.



Before looking for a good set of launch locations, I first wanted to identify a few safe landing areas; typically 10 square mile patches of rural area. Eventually I found a 20 by 30 mile area away from airports, restricted air space, and heavily populated areas. Knowing this potential landing area, and knowing about how far the balloon would travel, I panned around Google Earth looking for rural athletic fields and open park spaces in the area where launch might be possible. This provided an excellent sample of potential launch locations that could suit different wind patterns and speeds on the day of the launch.


As the launch date approached, I began modeling the trajectory 84 hours before the launch and began to select the anticipated launch location(s). As the model became more refined closer to launch, the actual launch location was selected.