Bob Simpson was one of the first meteorologists to recognize the potential weather analysis and forecasting benefits of dropwindsonde observations obtained by aircraft in data void areas of the tropical oceans. Dropwindsondes measure vertical profiles of wind, temperature, and humidity as a function of pressure during their fall from an aircraft to the ocean surface and transmit the meteorological information to the aircraft.

In 1982, the Hurricane Research Division (HRD) began to investigate possible improvements of tropical cyclone (TC) track forecasts that could result from the addition of dropwindsonde observations in the TC environment. During aircraft research flights around a TC, the flight crews of the turboprop WP-3D aircraft, operated by the Aircraft Operations Center, obtained the dropwindsonde observations and transmitted them to the National Weather Service before forecast deadlines. The observations helped the Environmental Modeling Center (EMC) and the Geophysical Fluid Dynamics Laboratory (GFDL) improve their model initializations and their TC track forecasts. In 1996, NOAA researchers reported significant reductions in research track forecast errors from operational numerical track models that were initialized in a research mode with and without the dropwindsonde observations.

These accomplishments led to enhancements of the dropwindsonde program that include the development of a new generation of Global Positioning System dropwindsondes and the purchase of a G-IV jet to drop the sondes from higher altitudes. With the acquisition of the jet, the program is now capable of obtaining dropwindsonde observations in support of operations when a TC threatens coastal areas of the Atlantic basin, Hawaii, or southern California. At the same time, EMC and GFDL have steadily improved their model initialization procedures and TC track models, and new types of satellite observations have enhanced the model initializations. Recently, HRD has developed a strategy for identifying potential sonde release locations to optimize the impact of the observations on improving the numerical model forecasts of TC tracks.