The Impact of Geostationary Operational Environmental Satellite-16 (GOES-16) Data on Tropical Cyclone Forecast

The advanced imagers such as ABI onboard GOES-16 and -17, AHI onboard Himawari-8 and -9, AGRI onboard FengYun-4A can provide observations with high temporal (1 – 15 minutes), spatial (0.5 – 2 km) resolutions at 14 – 16 bands ranging from visible to infrared spectral regions, which can help better monitoring, understanding and predicting high impact weather events such as tropical cyclones (TCs). One important application is to improve TC forecast through assimilating the observations into NWP models, especially with information from three water vapor absorption bands reflecting three layered preciptable water (LPW) in the troposphere. In addition, the derived atmospheric motion vectors (AMVs) from high temporal and spatial resolution ABI radiances can further improve the analysis and forecast.

To study the impact of GOES-16 data (Radiances, LPWs and AMVs) on TC forecast, hurricanes Harvey (2017) and Irma (2017) were chosen for case studies, and the experiments were carried out using GSI as analysis system and WRF as forecast model. The experiments with GOES-16 data assimilated were compared with the control run which only assimilates conventional observations and radiance data from polar orbit satellites (AMSU, ATMS, IASI and CrIS). GOES-16 satellite data assimilated include ABI radiances, three LPWs, AMVs, respectively.

The results show that GOES-16 ABI data improve the hurricane track forecast in both cases in which the track forecast RMSEs became much smaller along the forecast hour comparing to the control experiments. For Harvey case, the impact from ABI radiances was most influential, showing that the 72-h track forecast RMSE is reduced by up to 50 %. Meanwhile, LPWs showed the most positive impact in Irma case decreasing 72-h track forecast RMSE by up to 32 % (AMVs and ABI radiances reduce track forecast RMSE by 31 % and 28 %, respectively, in the 72-h forecast). Furthermore, the forecast RMSEs of maximum wind speed (SPD) and minimum sea level pressure (SLP) were reduced in Harvey case. The mean improvement from 6-h to 72-h forecast for SPD were 30 %, 15 % and 7 %, respectively after ABI radiances, AMVs and LPWS, are assimilated. The bias and RMSE of temperature analysis fields against radiosonde were decreased between 350 hPa and 200 hPa after assimilating ABI radiance data. More hurricane cases and results will be presented at the meeting.