Tropical low near Yucatan straights after 48hrs

This message was edited 2 times. Last update was at 06/06/2006 05:50:20

Two runs in a row now support this solution. Minimum sfc pressure drops to 974mb at 84hrs. I have no basis to say if this appears overdone or not, but it is interesting.

Here's a screen capture...

Steve --

The NAMX has been tending to create spurious tropical cyclones, and it would appear that this will be a feature of this model. I suspect it's the result of the non-hydrostatic nature of the model (buoyancy is predicted, at least at the 12-km/resolved scale of the model). Retrospective runs from last summer handled the development of Hurricanes Dennis, Katrina, and Rita much better than the NAM, and really "went to town" on their deepening into major hurricanes.

The spurious cyclone problem was also a problem with the GFS (and still is to a small extent). I would expect to see some more false alarms out of the NAMX when it's put into operations, but since it's not really used for hurricane guidance (a Hurricane WRF is under development, however, that'll be tuned to handle hurricanes [b]specifically[/b]), it's not a show-stopper for implementation. You just have to know to look out for these false alarms.

If the GFS also develops a tropical system, I'd suggest that would be a good way to confirm development.

Bill

[quote=Bill Bua]Steve --

The NAMX has been tending to create spurious tropical cyclones, and it would appear that this will be a feature of this model. I suspect it's the result of the non-hydrostatic nature of the model (buoyancy is predicted, at least at the 12-km/resolved scale of the model). Retrospective runs from last summer handled the development of Hurricanes Dennis, Katrina, and Rita much better than the NAM, and really "went to town" on their deepening into major hurricanes.

The spurious cyclone problem was also a problem with the GFS (and still is to a small extent). I would expect to see some more false alarms out of the NAMX when it's put into operations, but since it's not really used for hurricane guidance (a Hurricane WRF is under development, however, that'll be tuned to handle hurricanes [b]specifically[/b]), it's not a show-stopper for implementation. You just have to know to look out for these false alarms.

If the GFS also develops a tropical system, I'd suggest that would be a good way to confirm development.

Bill[/quote]

Thanks Bill. I did not know about Hurricane WRF, but ran into it last night while googling.

http://www.emc.ncep.noaa.gov/WRF/menu.html
http://www.wrf-model.org/plots/realtime_hurricane.php

Downloaded the training presentations but did not have a chance to go through them. Is there any recent developments? Would be a shame not to have it running at full steam by late summer.

Its not quite verification time for the above forecast valid 18Z Fri, but everyone around here is talking about the low pressure which developed over the western Caribbean.

http://www.srh.noaa.gov/printable.php?pil=AFD&sid=TAE&date=20060609080531

http://www.srh.noaa.gov/printable.php?pil=AFD&sid=MOB&date=20060608210755

http://www.srh.noaa.gov/printable.php?pil=AFD&sid=LIX&date=20060609083225

A similar type of thing happened with the extratropical cyclone that developed off the Delmarva yesterday and is now in the Gulf of Maine just off the Central Maine Coast as a 992 hPa low 12Z (10 Jun).

NAMX (WRF NMM) was consistent in being too slow with the track and over deepening the low which resulted in a smaller more intense cyclone in the runs 60 to 18 hours leading up to current. Interestingly, the current operational eta NAM was very good except for the 06 and 18Z 8 Jun runs. The eta was a few hPa too high on the pressure, but location and pressure field were subjectively more consistent with verification.

Attached (hopefully) is the 30 hr forecasts of the NAM (eta) and NAMX (WRF NMM) from the 9 Jun 18Z run valid at 10 Jun 12Z that was typical of the runs. The WRF developed the more intense gale center SE of Cape Cod.

As Dr. Buda indicated, I too believe that non-hydrostatic effects are probably the cause here. The convective BMJ precip fields were much more active of the Gulf Stream in this event and thus more latent heat release resulted in the more intense cyclone. While the WRF was only off by 2 hPa in central pressure the sensible weather fields and particularly wind in this case were vastly over predicted.

On a side note about the tropical systems – that NAMX is currently continuing to try and intensify in the Gulf, I had noted AFWA ARW core cases last summer of spurious cyclones in the Arabian Sea. Thus, while we may want to attribute the excessive cyclone spin ups to non-hydrostatic effects I wonder if something else may be happening here since I had seen this in ARW cases too.

Jeff Tongue
Upton, NY

I'm not sure if the hurricane formation and the extratropical storm Jeff Tongue commented on are related but both are frequent problems now.

Regarding the tropical storms, NCEP is looking at this and I plotted hourly BUFR forecast soundings out to 84 hours for four Carribean stations and saw no "smoking gun" but did see a few very bizarre-looking soundings, though most looked ok. The bizarre soundings were in two categories - dry adiabatic lapse rates in saturated environment with extreme CAPE but no convection - this was not in the location near where/when the storm was deeping rapidly and only lasted a few hours, and the other category was odd oscillations in the vertical. Once a storm starts to form, there is very strong feedback between the latent heat fluxes off the sea surface and the wind speed, as faster winds cause more evaporation. This is qualitatively realistic and needed to simulate real hurricanes, but I don't know whether it is excessive or not.

All or nearly all of the precipitation in these tropical storm events this week is being produced by the convective parameterization, which has two major shortcomings that can lead to the type of feedback we are seeing - though the mystery then is why it doesn't do this in the Eta, which also has the same parameterization. First, the BMJ scheme does not alter the subcloud air, so the boundary layer does not get stabilized directly, though it usually causes cooling and drying just above the LCL and this cool air can descend into the boundary layer. This is important because most models with convective parameterizations produce too many tropical storms while cloud-resolving simulations of tropical convection such as by Greg Tripoli found that downdrafts cause local divergence and higher surface pressure in the key areas where the developing storm needs the opposite to consolidate and have positive feedback - the main reason we don't see so much tropical development in nature may be convective downdrafts. Second, it does not have any vertical momentum transport. Several years ago, the GFS was predicting numerous spurious tropical storms as well as real storms, and then momentum mixing was added to its convective parameterization. The result was most, but not all, of the spurious storms were suppressed while real storms were not as much suppressed. However, it is much easier to introduce momentum mixing in a physically meaningful way in a mass flux scheme like that in the GFS than in a scheme that assumes reference profiles and does not compute layer mass fluxes, like in the NAM.

The bottom line is that we can expect this problem to continue in the NAM, but indications from last year are that it will at least make pretty good predictions of real storms - our challenge is to distinguish between the spurious and real.

As for the extratropical cyclones, for the past 2+ months, NAMX (NAM-WRF) has overdeepened shortwave troughs dropping into a longwave trough position tremendously, almost every trough. The very large errors have generally been confined to a small region near the center of the upper low rather than a major overdeepening of the long wave. The resulting too-strong upper low probably had a realistic interaction with the moist baroclinic air mass it ran over, producing the coastal cyclogenesis in the forecast, the error being caused by the overdeepening of the upper system upstream. NCEP has been wrestling with this problem for a long while and has eliminated many possible explanations but nobody knows what is causing this to happen repeatedly.
Unfortunately, we can expect it to continue until the cause is known and a fix can be formulated and tested. A few times it has been correct when it has been deeper than other models, but usually it is wrong.

Stephen

Hmmm...this looked very close to the position of Alberto. No?

Steve-O