Potential Impacts of Radio Occultation Data Assimilation on Forecast Skill of Tropical Cyclone Formation in the Western North Pacific

This study assesses the potential influence of global navigation satellite system (GNSS) radio occultation (RO) data assimilation on the forecast skill of tropical cyclone formation over the western North Pacific in September-October 2019 through a regional model. Data from the Constellation Observing System for Meteorology, Ionosphere, and Climate mission II are applied. The forecast skill considers the hits and misses for nine developing cases and the false alarms and correct negatives for 23 non-developing cases. Forecasts assimilating GNSS RO data reduce the false alarm ratio by 20% and increase the accuracy rate by 19%, compared to forecasts without GNSS RO data. Assimilation of GNSS RO data increases mid-level moisture around the disturbance centers at the initial time of the forecasts. It also increases low-level vorticity for developing cases but decreases vorticity throughout most of the troposphere for non-developing cases. These lead to improved forecast performance for tropical cyclone formation. Tropical cyclone formation is very sensitive to local atmospheric conditions such as moisture, temperature, and winds. Getting these conditions correct is therefore critical for accurate forecasts of tropical cyclone formation. Satellite data are commonly used to characterize atmospheric conditions over the remote oceans where tropical cyclones tend to form. New satellite data, specifically from the radio occultation of global navigation satellite systems, provide fine scale information on moisture, temperature, and pressure. This new information suffers less from contamination by clouds and precipitation than the traditional microwave and infrared satellite observations, and may further improve the representation of atmospheric conditions. In addition to cases of tropical cyclone formation, this study pays particular attention to the quality of forecasts of disturbances that did not develop into tropical cyclones-something that is often overlooked when assessing the quality of tropical cyclone formation forecasts. We find that incorporating radio occultation data into the regional model can improve the local environmental conditions for tropical cyclone formation forecasts. This leads to fewer false alarms and higher hit rates. These improvements may provide critical guidance for tropical cyclone formation forecasters.