Falling asleep on the job, a possible answer to the low confidence question, and the next storms in line
September 1, 2008
First off, deepest apologies that visited this blog during the wee hours of Sunday morning to read about what the Hurricane Hunters found. Your correspondent got a little too comfortable pondering why the forecaster was compelled to describe his intensity forecast was low confidence and subsequently dozed off. Woke up late to rapidly prepare for a trip to visit my sister (first since coming back from Iraq). Of course the National Hurricane Center, Brendan Loy, Dr Masters, and others were on the job.
To finish the thought from Saturday night, if there was anything to my feeling that the forecaster (Dr. Beven) had a doubt that he wasn’t expressing, it was over the intensity of Gustav at that time. Recall that in the absence of reconaissance data, an estimate of 120 (140 mph) knots was made. The first advisory to have recon data set the maximum sustained 10 knots lower and noted that it could be lower yet (Gustav Discussion 27). The following advisory was another 5 knots lower and yet again noted the possibility of the actual strength being even lower (Gustav Discussion 28). It really wasn’t until 11 PM that the wind speed in the advisory was strongly supported by the winds observed by the Hurricane Hunters.
The situation reminded me of Dennis back in 2005. During that storm, Dr Beven had to make a forecast without recon data while waiting for Dennis to go back over water after traversing Cuba. He went with 95 knots for an initial forecast and 100 knots for 12 hours later. The subsequent forecast, which had recon data, gave a “very generous estimate” of 80 knots. Six hours later it was 85 knots, (which put Beven’s forecast 15 knots over) and forecast a maximum of 95 knots. After that, however, all hell broke loose as Dennis had a deepening (strengthening) rate “that bordered on the insane” and ended up with the storm peaking at 125 knots. Perhaps it was memories of that roller coaster ride that drove Dr Beven to express his low confidence in the intensity forecast for Gustav.
Moving back to the present… we have Hurricane Hanna relaxing near the Bahamas. The forecast currently calls for landfall near the South Carolina/Georgia border on Friday. There’s some disagreement with the models as the UKMET forecasts a south Florida landfall. The track confidence is somewhat low to Hanna being near stationary at the moment and some uncertainty over which way the storm will move once motion resumes. In a general sense, the situation reminds me of Jeanne in 2004; at one point where Jeanne was in a weak steering flow, the forecast was for the storm to come out moving sharply to the north towards North Carolina. The ultimate landfall point, however, ended up being central Florida. I don’t mean to suggest that we are in a repeat situation, I am simply pointing out the trickiness inherent in the forecast in these situations.
Further to the east is Tropical Storm Ike. Unlike Hanna, it’s under a very defined steering regime, which is taking it west/northwest to due west. The forecast takes it to the eastern edge of the Bahamas by Saturday and due to the clear steering mechanism, the confidence in the forecast is, to quote the discussion “relatively high”.
A lively ten days lies ahead.