Performance of Cloud-permitting Hurricane Prediction through Assimilating High-resolution Airborne Doppler Radar and Satellite-derived Inner-core Observations

Since 2008 a regional-scale ARW-based ensemble Kalman filter (EnKF) analyses and prediction system with airborne Doppler radar-derived winds for cloud-permitting hurricane initialization and forecasting has been creating real-time forecasts as part of the NOAA Hurricane Forecast Improvement Project (HFIP). This system demonstrates very promising performance, especially on hurricane intensity forecasts, through experiments over 73 applicable NOAA P-3 airborne Doppler missions during the 2008–2011 Atlantic hurricane seasons. The mean absolute intensity forecast errors initialized with the EnKF-analysis of the airborne Doppler observations at the 24- to 120-h lead forecast times were 20–40% lower than the National Hurricane Center's official forecasts issued at similar times.

These results are tempered by the reality that NOAA P-3 missions are often limited to Atlantic storms that are close to the North American continent. To potentially address this observational shortcoming, we began using our EnKF data assimilation system to incorporate high-resolution satellite-derived observations in the inner core when airborne radar observations were unavailable. The inner-core observations come from NESDIS's Multi-platform Tropical Cyclone Surface Wind Analysis (MTCSWA), which is a global satellite-based tropical cyclone surface wind analysis that incorporates inputs from multiple satellite platforms and makes use of a variational data fitting method (Knaff et al., 2011). MTCSWA has been produced in real-time since 2009 and provides 10-m winds in polar coordinates with 4.5 km radial and 10º azimuthal resolutions. More importantly MTCSWA can resolve the very strong winds within 200 km or so of the tropical cyclone center. The input data of MTCSWA includes the oceanic wind vectors from the Advance Scatterometer (ASCT) on board the METOP-2A satellite, cloud drift feature track (CDFT) and water vapor (WV) winds from geostationary satellites, balanced flight-level winds estimated from the Advanced Microwave Sounding Unit (AMSU) data from NOAA 15, 16 and 18 (Bessho et al. 2006) satellites, and flight level analog winds created from infrared imagery and operational intensity and position estimates (IRWD; Mueller et al. 2006).

Our preliminary results based upon assimilating the MTCSWA observations for all 2010 Atlantic tropical cyclones initialized every 6 hours with best track wind speed greater than 33 m/s are very promising. Verification based on more than 100 events shows that the ARW forecasts of both hurricane track and intensity with assimilation of the satellite-derived wind are markedly better than the operational GFDL and HWRF regional hurricane prediction models, and are comparable to, or, in some cases, better than the NHC official forecast. Preliminary analysis also shows that assimilation of these high-resolution satellite-based wind observations may be just as effective as assimilating airborne Doppler radar observations when evaluated in terms of the 1-5 days' track and intensity forecasts. The high temporal and global availability of the MTCSWA observations has the potential to extend our success to the vast majority of Atlantic and East Pacific cases and to other tropical cyclone basins that do not have airborne Doppler radar observations.

We are currently exploring the ARW forecast performance through continuous cycling of the hurricane analyses and forecasts with the EnKF assimilation of the airborne radar and/or satellite-derived inner-core observations (i.e., whichever is currently available). These continuous-cycling results, which utilize both airborne radar-based and satellite-based inner core wind information from retrospective runs during the 2010-2011 Atlantic seasons, will be presented at the meeting. Based on preliminary and anticipated results of this experiment, we propose to use this continuous cycling ARW-EnKF system for real-time testing as part of the HFIP demo project during the 2012 hurricane season.