Subject: Re: questions for presenters Date: Tue, 6 Aug 2002 13:38:22 -0400 From: "Mark Powell" To: Colin.J.McAdie@noaa.gov CC: Peter Black , James Franklin , Christopher Landsea , "Brian.R.Jarvinen" , "Richard.J.Pasch" , "James.M.Gross" , "John.L.Beven" , "Edward.N.Rappaport" , Frank Marks On Monday, August 5, 2002, at 05:07 , Colin J Mcadie wrote: Presenters, The best-track committee would greatly appreciate your response to several follow-up questions arising from your presentations of last Thursday. Although these questions are posed to individuals, everyone is invited to respond so at we may have the best possible basis for a decision. If you could respond by Wednesday it would be appreciated, as the committee will meet again on Thursday. 1) To Mark: The crux of our debate seems to center upon where the marine-land transition occurs. In your set of '96 W&F papers, we find the following:(part II, p. 332) "According to the analysis in Fig. 3c, Andrew's maximum winds were in the northern eyewall near Perrine where onshore flow reached a maximum of 66 m/s on the coast and then decreased to 62 m/s just inland. The difference between land and marine winds in the vicinity of the eyewall are relatively small due to comparable roughness lengths for marine and open terrain in high winds, as discussed in Part I." (emphasis added) Mark, if you were to write this paper today, would you state this differently? It seems to contradict the discussion on Thursday As I said in my presentation, the results of recent work looking at the mean GPS profiles in the lowest 200 m has me to believe that roughness over open sea is much less than what has been used before (and used in our Andrew study). Therefore the Andrew winds over open sea were probably much higher than what was in our paper. However, our paper focused on landfall and at landfall, the winds were adjusted to the sfc with a PBL model. This model used a Charnock relationship similar to what other models (like GFDL) use to describe how roughness varies with wind speed. Values of roughness from Charnock approach open terrain values over land (as described in the paper). In my opinion, the roughness of a coastal sea with breaking and shoaling waves is probably similar to what Charnock gives in high winds....somewhere between 0.01 and 0.03 m (open terrain). In the Andrew analysis published in W&F, we assumed that the flight level winds were equivalent to mean PBL winds. Now we know that 700 mb winds are typically weaker than the mean PBL wind and H*Wind has a new method to estimate the MBL wind from the 700 mb wind and the flt level Rmax. That method estimates a 90 m/s MBL wind from the 81.6 m/s flt level wind. If we input that number to the PBL model with the air and sea sfc temps at Fowey Rocks for this roughness range we get the following results: Roughness (m) U10 (m/s) U10 (kts) 0.01 69.67 132.4 0.02 67.34 130.6 0.03 65.75 127.6 In my opinion this range of results is appropriate for Biscayne Bay and onshore flow areas of the coastline. In referring to Part I, there is another interesting discussion in 4.a. Determination of Exposure in which there is a discussion of the anemometer (LLWAS) cluster at MIA, with varying exposures, two of which have suburban exposure, one open-terrain (close to the runway). When the wind is out of the north, roughless lengths performed well in normalizing all three to a standard height. "However, when the wind veered to the northeast, open-terrain-adjusted winds at FA2 and FA4......failed to compare with FA1. Apparently, the distant marine fetch (even at 20-30 km upwind) became important, and roughness lengths had to be reduced." If marine effects are felt that far inland, does this argue that marine effects would dominate on the immediate coast? The LLWAS cluster observations indicated that roughness changes as a function of upstream wind direction (this helped launch a USWRP effort to document exposures as a function of direction so that wind observations could be corrected (see www.aoml.noaa.gov/hrd/asos). It is also evident that roughness for any particular direction should depend on upstream fetch. When flow approaches a new underlying surface a new internal boundary layer develops and it takes a while to come to full equilibrium with the new surface. Our paper mentions a couple of studies on this but not enough has been done to give you a firm answer. I believe that when air passes from smooth to rough underlying surface there is some memory of its earlier passage but the distance of the effect is unknown. I think most of the adjustment to new terrain happens over a few km. In our work on the new Hurricane loss model for the State of Florida, we have a virtual roughness for each zip code as a function of wind direction. The upstream fetch influence function is a Gaussian filter in which the half power point is 3 km upstream but the full extent of the upstream fetch extends to 20 km upstream. You also note: "In a hurricane the flow trajectory is complicated by the additional accelerations associated with the pressure field, the radial distance from storm center, and convective components associated with updrafts and downdrafts in the eyewall and rainbands." The above could also be taken to mean that there is very little, or possibly no, transition very near the coast in the core area. Does this agree with your current thinking? I had to put those comments in to appease a reviewer who took issue with the application of surface layer similarity theory to the hurricane. Now that the GPS data support mean log profiles, I believe that surface layer similarity is applicable. 2) To James: Based on your preliminary analysis of dropsonde profiles in coastal regions versus those in a purely marine environment, do you now have some uncertainty in the use of a constant reduction factor for flight level winds in near-coastal regions? I agree with James on this and appreciate the effort he has made to check out the sonde data near the coast. To use one of Pete's quotes: "More data, more data, more now, not later" 3) To Pete: Why did you feel that you could quantitatively convert (or interpret) a surface wind over land from marine observations, but Mark could or would not? That is, your document says "multiplying this by the mean over land to over water ratio of 0.92 +/-0.8..." I dont want to get Pete mad.... 3) To all: How should the apparent strengthening seen in the surface pressure and satellite data after the 0810 UTC 162 kt flight-level wind be interpreted in terms of the wind field? Also, what effect, if any, should the strong cells forming along the coast in the northern eyewall have had on the surface wind field at the shoreline? I know of no definitive work that could answer this question. 5) To Mark: It is noted that the pressure at the time of the 109-111 kt wind at Fowey Rocks was 968 mb. The pressure analysis appearing in Fig 4 of Ed's prelim would seem to indicate that the tightest gradient (about 1 hour later), and presumably the maximum wind, was in the vicinity of BK Corp, at about 945 mb. Even given the fact that the system was deeping over this hour, it seems that the 968 mb is too high to be associated with the maximum wind. This is a concern because our recollection of your presentation (Mark correct us if this is not correct) is that you questioned your analysis because of the apparent mismatch with the Fowey Rock observation - taking it as representative of the maximum wind. In summary, the 968 mb pressure would seem to support some increase in wind after the failure, and perhaps a better match with your analysis. I never said that Fowey Rocks was representative of the maximum wind. I acknowledged that Fowey had not yet reached the inner edge of the eyewall and that the winds could have been higher. How much higher is open to speculation. I said that the 0759 H*Wind analysis wind speed at the location of Fowey Rocks was about 30% higher than what was observed. If the pressure looks too high you should talk it over with NDBC. Dr. Mark D. Powell Atmospheric Scientist AMS Certified Consulting Meteorologist NOAA Hurricane Research Division 4301 Rickenbacker Cswy. Miami, FL 33149