Summary: During each hurricane season, NOAA’s Gulfstream-IV (G-IV) Hurricane Hunter aircraft measures wind, temperature, humidity, and pressure in and around hurricanes threatening the United States. An instrument called the dropwindsonde is released from the G-IV to collect the data as it falls to the ocean surface. The plane currently flies where we expect observations from the dropwindsondes will make the best forecasts of where the hurricanes will go (what we call “track forecasts”), but the methods to find these places are more than 20 years old using old computer weather models and methods to put the data into the model. Observing System Simulation Experiments (OSSEs) are used to look at new ways to find where the G-IV should release dropwindsondes to improve forecasts.
In an OSSE, a realistic model of a hurricane (called a “hurricane nature run”) is created. Dropwindsonde measurements are created in the nature run just like they are taken in the real atmosphere, and these data are put into the computer forecast models to see how much they improve the forecasts. In this study, we used the Hurricane Research and Forecasting (HWRF) model to see how they forecast what happened in the nature run, just like it is used to forecast hurricanes during hurricane season. We compare results from data taken at different distances from the center of the hurricane. We also looked at the impact on forecasts of how strong the storm will be (called “intensity forecasts.”)
Important Conclusions:
1. Dropwindsonde data improve track and intensity forecasts.
2. There is a preferred region for hurricane data at a distance from the hurricane’s center that depends on the size of the hurricane. We should gather these observations in a large area around large hurricanes and in a small area around small hurricanes
3. Dropwindsonde data closest to the hurricane most improved the track forecasts.
Read the full paper at https://journals.ametsoc.org/doi/abs/10.1175/MWR-D-18-0157.1.