Performance of an Objective Model for Identifying Secondary Eyewall Formation in Hurricanes Matthew Sitkowski CIMSS – University of Wisconsin Jim Kossin NOAA/NESDIS/National Climatic Data Center
Acknowledgments Thank$ –Office of Naval Research –NOAA GOES-R Risk Reduction Jeff Hawkins –NRL Webpage Mark DeMaria –SHIPS Dataset Dave Nolan, Chris Rozoff, John Knaff, Howard Berger, & Chris Velden
Marked changes of the inner core structure Broaden wind field – increase storm surge Linked with rapid intensity changes Landfall blessing/curse “SO SOME ADDITIONAL STRENGTHENING IS POSSIBLE... IF AN EYEWALL REPLACEMENT CYCLE DOES NOT INHIBIT THE INTENSIFICATION PROCESS.” “... AND WE HAVE NO SKILL IN FORECASTING EYEWALL REPLACEMENT CYCLES BEYOND ABOUT 6-12 HOURS...AT BEST Rita (2005) NHC Discussion #18 (Stewart)
Secondary Eyewall Formation (SEF) Objective Model Development Algorithm requires the knowledge of prior SEF events Developed an SEF climatology from Uses environmental features from SHIPS –Shear, MPI, relative humidity … GOES IR features –Improve probability of detection from 22% to 30%
SEF Algorithm 2 classes (Cat 1+ over water) –SEF occurs within the next 12 hours –SEF does not occur within next 12 hours The algorithm is based on the Bayes probabilistic model P (C yes | F) estimates the probability of imminent secondary eyewall formation, given the set F of observed features. P (C yes ) is the climatological probability (~12% in the North Atlantic). “Leave-one-season-out” cross validated
2008 Season 8 Hurricanes Bertha Dolly Gustav Hanna Ike Kyle Omar Paloma 3 SEF Storms Algorithm ran smoothly –Missing key IR features New forecast every 6 hours –t=0,6,12,18, & 24 hr Greatest probability for non- SEF storm was 6% (excluding Gustav) Ike had 18 hr forecast of 65% that verified Sept UTCSept UTC
Hurricane Bertha First major hurricane Never had probability >0% Classic eyewall replacement Features never acted in accord to produce a high probability
Hurricane Bertha Some features were favorable at times –Shear somewhat high –Low MPI for SEF storm
Probability Intensity CubaLA
TXCuba Probability Intensity SEF
Cuba Probability Intensity SEF
2008 Beta Version Run with key features missing – Improvement expected with the addition of the missing IR features No major glitches –Automated - updates every 6 hr –Available online: ftp://ftp.ssec.wisc.edu/pub/matts/ A climatology of recon-based intensity changes associated with secondary eyewall formation would be of great benefit to forecasters