MargieKieper doesn't have a bio yet.
The Storm at the NHC
By: MargieKieper , 5:05 PM GMT on July 04, 2007
Thursday evening update: Today a large number of the staff at the TPC/NHC signed a petition asking for Bill Proenza to be replaced as NHC Director -- again documented by the Miami Herald. This was done off-site and not all the employees were able to be contacted regarding the petition. However a large majority of the employees who were contacted, about 70%, did sign the petition.
This morning, HRD discussed the matter at an all-hands meeting, and unanimously supports Proenza. Of course, none of the HRD employees work for Proenza, so I am not sure how this vote of confidence would factor into any decision on Proenza's tenure, and it seems that NOAA has already set in motion the actions that will lead to resignation or termination. I may have more information regarding HRD's point of view tomorrow.
It was very interesting that Proenza's own administrative staff signed the petition. Many people at the center have had difficulties working with Proenza, and this factored heavily into some of the decisions. This is unfortunate because this past six months should have been the time to establish a rapport and good working relationship with the staff, and it is telling that not only the Senior Hurricane Specialists and other senior staff, but the administrative staff as well, are willing to go on record as stating that they would prefer a new NHC Director. It does appear that the concerns about whether Proenza can be an effective manager are valid ones, as it is hard to imagine how so many people could take this extreme position without there being quite a bit of substance to their concerns.
I want to try to clarify something. I had concerns with the QuikSCAT numbers from the get-go, I researched the issue with the QuikSCAT data for some time, and I contributed in a major way to the blog Jeff and I co-authored Wednesday, to the point of seeing what I wrote posted in many places on the web in the last day -- but while I do support the Hurricane Specialists, who put their heart into their job on every shift, and while I have come to view Bill Proenza in a very critical light, I am not a meteorologist and in that sense am not part of the tropical met community, and so it is not my place to say anything regarding whether Bill Proenza should step down. That would be overstepping my bounds. Perhaps this is too fine a distinction to draw, but to me it is an important distinction. I want to correct any misimpressions that may have arisen.
Thursday morning update: The third named storm of the West Pacific season formed and subsequently made landfall while in the Gulf of Tonkin. This storm was named Toraji and never strengthened beyond minimal tropical storm strength.
To answer a question on satellites in the news, this was undoubtably referring to the NPOESS program. The scope of the program was reduced and a number of sensors for monitoring climate were cut.
* * * * * * *
Jeff and I have co-authored Jeff Master's WunderBlog, today, which provides a different perspective on the situation at the National Hurricane Center, based on our research into the QuikSCAT statistics presented in the press this past spring, and what we feel is a fair and balanced assessment of the issues raised in the press by NHC Director Bill Proenza.
As time allows over the next couple days I'll be using the blog to post some of the detail and background information from researching the QuikSCAT issue. This will include an analysis of the phone interview the Miami Herald did with Proenza prior to the relase of their Tuesday evening breaking news story, and some interesting information on the history and varied uses of the QuikSCAT satellite, including more detail taken from last spring's workshop.
* * * * * * *
On a Friday in mid-March, I read for the first time Bill Proenza's comments on the reduction in track forecast error accuracy that he believed would occur once QuikSCAT reached the end of its lifespan and failed, in an AP article written by Jessica Gresko, "Hurricane Forecasts May Be Less Accurate" (no longer available online). The article was subtitled, "Aging satellite provides key information, meteorologist says." The numbers, and the emphasis on track forecast rather than intensity forecast, astounded me.
So, bright and early on Saturday morning, I went to the NHC website and pulled the 48-hour and 72-hour average seasonal track error numbers for all years since QuikSCAT became operational -- from 1999 through 2006.
Using the data from these seasonal verification reports available online from NHC, it was a straightforward matter to immediately quantify these track error percentages. Since QuikSCAT data became available, starting in 1999, average track errors for 48-hour and 72-hour forecasts have been reduced by 43 mi and 62 mi respectively. Since 2004, the 48-hour and 72-hour track errors have been approximately 100 and 150 miles respectively. An additional error of 10% for the 100-mile 48-hour track forecast error would result in an error of 110 miles, or one quarter of the 43-mile gain in 2-day track forecasts since 1999 (23%). An additional error of 16% for the 150-mile 72-hour track forecast error would result in an error of 174 miles, or about 2/5 of the 62-mile gain since 1999 (39%).
I then contacted a key meteorologist in the field, who also shared my doubts regards the attribution of such a sizable quantity of the gains in track forecasting to QuikSCAT data.
Much later I reviewed the legislation in Congress to provide funding specifically for a next-generation scatterometer satellite to replace QuikSCAT. One thing struck me in the numbers for this portion of the bill:
(7) Despite its continuing performance, the QuikSCAT satellite is well beyond its expected design life and a replacement is urgently needed because, according to the National Hurricane Center, without the QuikSCAT satellite--I realized that the numbers provided for track forecast accuracy, 136 miles at 48 hours, and 197 miles at 72 hours, were on par with the averaged track errors representative of 2002 (137.5 and 200.1), not current average track forecast errors, which, since 2004, have been on the order of 100 and 150 miles. These numbers, 136 and 197 miles, might come from a different metric perhaps relating to coastal warning areas (the "cone of error"). If so, this may be mixing apples and oranges. No background material is yet available to source these numbers, but the study that provided the basis for the estimated loss of accuracy, 10% at 48 hrs and 16% at 72 hrs, has been shown in the blog Jeff and I posted today to have limitations that preclude reasonably extrapolating the results to landfalling tropical cyclone track errors.
(A) hurricane forecasting would be 16 percent less accurate 72 hours before hurricane landfall and 10 percent less accurate 48 hours before hurricane landfall resulting in--
(i) with a 16 percent loss of accuracy at 72 hours before landfall, the area expected to be under hurricane danger would rise from 197 miles to 228 miles on average; and
(ii) with a 10 percent loss of accuracy at 48 hours before landfall, the area expected to be under hurricane danger would rise from 136 miles to 150 miles on average;
* * * * * * *
I want to go on the record to express my support to the four NHC Senior Hurricane Forecasters Avila, Franklin, Knabb, and Pasch, who publicly expressed their concerns and provided a balancing perspective to what has been, up until now, hidden from the public view.
Things have been more like Stravinsky than Ravel lately at the NHC (shouting matches in the hallways!...and I always associate Ravel with water). Let's hope that these major issues are resolved soon and they can look ahead to the busy portion of the hurricane season with no more than the usual amount of stress that comes with the job.
Comments may also be emailed to viewfromthesurface at wunderground dot com.
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.