Impact of CYGNSS Data Assimilation on FV3-GFS Tropical Cyclone Forecasts in October 2018

The Cyclone Global Navigation Satellite System (CYGNSS) provides remotely-sensed surface wind observations over water and is not impeded by cloud cover or precipitation. Therefore, CYGNSS offers a unique opportunity to observe surface winds near the core of tropical cyclones. Here, we present results using three observing system experiments designed to test the impact of CYGNSS on tropical cyclone track and intensity forecasts in NOAA’s FV3-GFS global weather modeling system. The first experiment mimics the operational configuration, but at a reduced horizontal resolution. The second experiment assimilates CYGNSS scalar winds, and a third experiment assimilates wind vectors. The directional wind information is created by merging a pre-existing background wind direction field with CYGNSS wind magnitudes using a variational assimilation method (CYGNSS-VAM). We ran the experiments for one month, cycling every 6 hours. The first two weeks (15-30 September 2018) are used as spin-up with the remaining days used as the verification period (1-15 October 2018). During the verification period, Hurricane Michael formed in the Caribbean, underwent rapid intensification in the Gulf of Mexico, and caused catastrophic damage in the Florida panhandle as a Category 5 hurricane. Our assessment of CYGNSS impact will primarily focus on Michael, while including forecast error statistics for all tropical cyclone forecasts investigated during the period of the study.