Hurricane Isaac, Part 3 - Analysis and Forecasting Thread

[Dsnowx53]

So I am posting this as much to ask if this makes sense: but wouldn't it seem that given how the low the pressure in this thing is, and how good the MW presentation is, if it managed to develop a persist ring of cold convection around the center that it could tighten up and intensify more rapidly than might otherwise be expected? I am not necessarily saying that we should expect that - more saying that theoretically wouldn't it make sense that a storm with an already extremely low pressure could see the winds "catch up" rather abruptly. I believe someone mentioned that Opal exhibited that sort of behavior.

Right, I would think that based on the laws of physics, this intuitively makes sense. The winds increase to fill the void for the lowering pressures, but this storm is still relatively broad with its wind field and does not have a great cold ring core. Thus, the wind compensation for the lower pressures, so to speak, is also spread out well beyond the center with lesser max wind values, as opposed to a tighter core of much stronger winds that would result if the core actually did tighten up.

Irene's pressure was really low, but its strongest winds were well removed from the center by the time it reached NJ/NY.

[LongBeachSurfFreak]

Agree, When Irene hit up this way in the Hudson valley, there was little wind. We had exstensive damage inland damage from flooding.

http://www.huffingto...tural-disasters

NEW YORK

Deaths: 10

Damage: More than $1.3 billion

Power outages: 1.1 million

Irene's eye made its third U.S. landfall in Brooklyn but did not materialize into the big-city disaster many had feared. The city had shut down its subway system for the first time in history and evacuated thousands of coastal residents. Only minor flooding was reported. But inland, the storm caused severe flooding that closed hundreds of roads and bridges. Tourist destinations in the Catskill and Adirondack mountains lost millions of dollars in revenue over the summer as they sought to repair damage to trees, hillsides and lakes that were drained by demolished dams.

Irene certainly want the NYC disaster feared but there were severe coastal impacts east of the center on LI. I was without power for 5 days and had surge in the 5 foot range at high tide aka the worst coastal flooding since the infamous December 92 noreaster (despite top sustained winds of 60ish gusts to 80ish). As said before a large storm with a large fetch will out preform surge wise by a full category. I think we will see a cat 1 80-90mph storm at landfall but easily cat 2 to 3 surge values.

[airmarci]

Here's a screenshot of vorticity from the UW-NMS run at 300 m horizontal resolution (not isaac)...

Don't mean to get off topic but is that a simulation of hurricane eyewall mesovorticies?

[dan88]

000

URNT12 KNHC 280306

VORTEX DATA MESSAGE AL092012

A. 28/02:31:00Z

B. 27 deg 07 min N

087 deg 00 min W

C. 850 mb 1251 m

D. 50 kt

E. 307 deg 62 nm

F. 043 deg 65 kt

G. 308 deg 73 nm

H. 979 mb

I. 16 C / 1521 m

J. 22 C / 1466 m

K. 17 C / NA

L. OPEN S

M. C28

N. 12345 / 08

O. 0.02 / 1 nm

P. AF309 2909A ISAAC OB 06

MAX FL WIND 65 KT NW QUAD 02:07:30Z

MAX OUTBOUND FL WIND 68 KT SE QUAD 02:54:00Z

;

[David Reimer]

000

URNT12 KNHC 280306

VORTEX DATA MESSAGE AL092012

A. 28/02:31:00Z

B. 27 deg 07 min N

087 deg 00 min W

C. 850 mb 1251 m

D. 50 kt

E. 307 deg 62 nm

F. 043 deg 65 kt

G. 308 deg 73 nm

H. 979 mb

I. 16 C / 1521 m

J. 22 C / 1466 m

K. 17 C / NA

L. OPEN S

M. C28

N. 12345 / 08

O. 0.02 / 1 nm

P. AF309 2909A ISAAC OB 06

MAX FL WIND 65 KT NW QUAD 02:07:30Z

MAX OUTBOUND FL WIND 68 KT SE QUAD 02:54:00Z

;

The temperature gradient is rising from previous VDMs.

[IsentropicLift]

Despite all of the pessimistic banter that has been going on in this thread for the past few hours, the IR presentation is continuing to slowly improve. WV loop shows a large increase of moisture in the SE quadrant and this should continue to envelope the center. Sure, this may not be the major Hurricane that it was once forecasted to be, but we're still over 24 hours away from forecasted landfall with very warm gulf waters and increasingly favorable ULL pattern ahead.

The other issue which doesn't seem to be getting much attention is the fact that if this system makes landfall either directly over top, or west of New Orleans, New Orleans will get the onshore flow (Eastern Eyewall) as opposed to the offshore flow (Western Eyewall) that it experienced in Katrina. Should Isaac take this more westerly track, the storm surge flooding could be catastrophic, and the slow movement would greatly prolong the effects. Way to soon to count Isaac out, just my two cent amateur opinion.

[David Reimer]

Means showing signs of getting better organized.

By your definition, yes. A larger temperature gradient is an indicator of strengthening in a tropical cyclone.

[Geos]

Despite all of the pessimistic banter that has been going on in this thread for the past few hours, the IR presentation is continuing to slowly improve. WV loop shows a large increase of moisture in the SE quadrant and this should continue to envelope the center. Sure, this may not be the major Hurricane that it was once forecasted to be, but we're still over 24 hours away from forecasted landfall with very warm gulf waters and increasingly favorable ULL pattern ahead.

The other issue which doesn't seem to be getting much attention is the fact that if this system makes landfall either directly over top, or west of New Orleans, New Orleans will get the onshore flow (Eastern Eyewall) as opposed to the offshore flow (Western Eyewall) that it experienced in Katrina. Should Isaac take this more westerly track, the storm surge flooding could be catastrophic, and the slow movement would greatly prolong the effects. Way to soon to count Isaac out, just my two cent amateur opinion.

Good points. Can't forget about the 16" of rainfall predicted for NO. All that water has to be pumped out of the basin. Will the pumps be able to handle all the "rain water"?

The wind speeds really don't even matter with this storm, it is the storm surge and rainfall tallies that are the serious matter.

[EasternUSWX]

By your definition, yes. A larger temperature gradient is an indicator of strengthening in a tropical cyclone.

Thought so. Was pretty sure it meant there was less humidity in the center meaning the clouds should start to decrease in the eye.

[turtlehurricane]

Center of the storm in view of KEVX, scanning at 41k ft. Closer to the center than KMOB and KLIX.

Don't mean to get off topic but is that a simulation of hurricane eyewall mesovorticies?

Yup, they pop up naturally in the model as well as other small scale features like boundary layer rolls and vortical structures in the eyewall.

[Wild Weather Monger]

GFS brings it just ese of New Orleans then just creeps westward into the city. Looks like a long night in the big easy and surrounding areas.

Feeling a tad better about my daughter down in Houma ATM. Looking like she has a decent chance of staying west of the eye.

[OKpowdah]

so by the way, with a 979mb TS, Isaac has the lowest pressure of a pre-hurricane storm in the tropical Atlantic west of 50W

[Bostonseminole]

70+knots winds on the latest pass at flight level on the NE quadrant.

[Superstorm93]

SSDD

Strong flight level winds with un-impressive surface winds.

Guess it comes with the package of a 979mb Tropical Storm

[40/70 Benchmark]

Very confident in my 80mph LF call from yesterday.....this is not gonna RI.

[turtlehurricane]

They're flying through the non-convective quadrant so that's the distribution of wind you'd expect, SW quadrant should have much higher sfc winds.

[EasternUSWX]

Those FL winds are way more impressive now, large shield of 65-75 mph. Once that starts getting down to the surface probably by morning it will start ramping up.

[turtlehurricane]

Isaac is actually directly over a cold core eddy right now, might explain why convection is weakening. And then it's gonna hit the shelf waters, but there is some better water between the eddy and the shelf so that'll provide another intensification opportunity ( http://www.aoml.noaa.gov/phod/dataphod1/work/HHP/NEW/2012240god26.png ).

[Ed Lizard]

Not exactly sure of station elevation, could be 10 m.

But TS conditions approaching the coast.

I may uload this as a saved image about lunchtime tomorrow when it gets really good.

[Indystorm]

And dew points are rising to the mid 70's from the low 60's ne and nw of the storm. Hopefully the dry air is finally leaving.

[phil882]

Serious question: does the hrrr have known strengths in the tropics? It's usually a third rate model to kill time with when storm chasing. I find it very odd to see it so much tonight. Can only guess folks are having trouble making reaches otherwise.

I think for a meso model..it has a decent cumulus parameters in the tropics as compared to the NAM. I think Phil even said this tonight. However, JMHO..it may be a little overzealous.

I haven't used it in this type of a tropical situation before, but I have to say that overall I have found the HRRR to be a quite useful tool for convective forecasting in the spring and summer for my own needs. I certainly would not refer to it as a third rate model personally. I mean, it isn't perfect by any stretch, but I think it does a pretty good job showing convective mode, strength, etc.

It has explicit convection so no cumulus parameterization. People like its pretty output and it's run every hour, hence it's popularity.

People definitely overuse the HRRR in the winter, especially when they are clinging to hope when a snowstorm is busting, but it's actually a pretty model for spring and summer convection, IMO.

The HRRR is a good meso model but I don't believe it's particularly specialized for the tropics, so other models like the GFDL, HWRF, UW-NMS/other university's models are likely better. I think people are getting the impression in this thread that it's really good since they usually don't see such high resolution model data, but it gets way better than that. Here's a screenshot of vorticity from the UW-NMS run at 300 m horizontal resolution (not isaac)...

It's cool though that the HRRR updates so often, the good mesoscale hurricane models generally have a huge lag time which lowers their usefulness as a forecast tool.

Haha... I saw this conversation happening sooner or later. I start posting about the HRRR and everyone follows suit. Here is my take on it.

The HRRR is still an experimental model being developed by ESRL. At chris87 mentioned, its huge advantage is that it uses no cumulus parameterization, because the resolution of the model is so fine, it is able to properly resolve convection. That means no parameterization (or added unknown) that can be a large source of error (Especially in the tropics where convection is highly prevalent). In order to not use a cumulus parameterization, you need to get down to 3-4 km which is where the HRRR is. Another big advantage (that I didn't see mentioned here) its its data assimilation. Instead of the GFDL and HWRF that use vortex bogusing (I think) to artificially create TCs that they then integrate forward in time, we have actual data that creates the initial conditions, taken both from radar reflectivity and olr from the RAP model. Its not perfect, and the model having to downscale from 13 km to 3 km creates a lot of error, but its superior to the vortex bogusing techniques in my opinion, and perfectly fine for short range forecasts. They folks that work on the HRRR are working to get it down to 3 km for data assimilation, which should improve these forecasts further, but thats still down the road. Notice when you animate the radar and outgoing longwave radiation that you see the model quickly adjust from the 13 km initial conditions to its native resolution... thats the downscaling effect that takes place as the model integrates forward in time. Not all models that have this high resolution are created equal... the power of HRRR is that its an already good model coupled with an ever improving data assimilation scheme.

Those that complained that it does poor with severe weather forecasting are not using it for its correct purpose. Of course if you are expecting the HRRR to nail the location and time of individual supercells or even ordinary thunderstorms you are going to constantly be burned. The model is not perfect. What it does become more useful is in forecasting different convective regiemes, and different convective modes. It won't tell you exactly where a storm is going to develop, but it can usually do a good job telling you what the storm mode for a particular day will be (linear, supercellular, multicell ect.)

For forecasting TCs, it has mainly been unproven (since its still relatively new and its domain only covers where it has access to the RAP, which is mainly over land). However, given how god awful the GFDL and HWRF are with their TC initializations, the HRRR in many ways is likely superior to these models for the time range that it forecasts. The HWRF in particular is what I like to call out, since its pretty much the perfect example of a model that doesn't belong in the tropics, since its model physics were based off the NAM. If they wanted to create a high resolution model for TCs, they should have based it off code other than the NAM which is specifically tuned to work well ONLY in the mid-latitudes.

The most useful tool I think HRRR should be used for is its ability to create dprog/dt output... which lets you pick up on short term trends. Since it does read in real data hourly, it can catch new reflectivity data that shows a system moving off course from a given forecast. For example, it seems to be capturing the short term slowdown of Isaac pretty well, when it slowed down the forward progression of the eye-like feature on the reflectivity animation I showed earlier.

I don't claim to be very knowledgeable on the nuts and bolts of HRRR, but from what I've seen so far (with its performance with Alberto [the ONLY model to pick up on its genesis] and Beryl) I'm lead to believe it has decent utility in the tropics.

For more information go here: http://ruc.noaa.gov/hrrr/

[turtlehurricane]

Pressure is the same as the last recon pass 2 hours ago, which is pretty much a win considering how bad it looks.

[Bostonseminole]

Pressure steady on the last pass at 979, center seem wobble left based on the last fix, these wobbles are going to be critical now as we get closer to the coast.

Using IPad or IPhone or BB so my grammar may suck.

[EasternUSWX]

Thru 54hrs SE LA is under 15-17" of rain. yikes.

[Derecho!]

Recon looks to be wandering the SW quad trying to smoke out a 65 kts SFMR

[dan88]

000

URNT12 KNHC 280515

VORTEX DATA MESSAGE AL092012

A. 28/04:26:50Z

B. 27 deg 17 min N

087 deg 27 min W

C. 850 mb 1253 m

D. 52 kt

E. 065 deg 25 nm

F. 137 deg 60 kt

G. 063 deg 30 nm

H. 979 mb

I. 18 C / 1514 m

J. 22 C / 1519 m

K. 19 C / NA

L. OPEN SW

M. C40

N. 12345 / 08

O. 0.02 / 1 nm

P. AF309 2909A ISAAC OB 10

MAX FL WIND 70 KT E QUAD 03:39:00Z

MAX OUTBOUND FL WIND 62 KT SW QUAD 04:54:00Z

[turtlehurricane]

The northern semicircle has zilch convection so winds are rising gradually at coastal locations rather than in bursts like a healthy tropical cyclone. Also winds are way lower inland since the real strong winds can't penetrate to the ground without convection. HRRR says this will change sometime around sunrise, things would become alot more stormy if it does.

[turtlehurricane]

Isaac has a ton of CAPE by the way, people who believe in CISK would love this storm: http://rapidrefresh.noaa.gov/hrrrconus/jsloop.cgi?dsKeys=hrrr:&runTime=2012082723&plotName=cape_t6sfc&fcstInc=60&numFcsts=16&model=hrrr&ptitle=HRRR%20Model%20Fields%20-%20Experimental&maxFcstLen=15&fcstStrLen=-1&resizePlot=1&domain=t6&wjet=1

[turtlehurricane]

Surge rose 0.1 ft in the last hour at Shell Beach, Louisiana, a spot upstream of New Orleans in the Gulf. 1st column after CDT is predicted, 2nd column is observed, 3rd is the residual (surge). We're heading out of low tide and combined with the increasing surge waters are rising quick, nearly 0.2 ft in the past hour. This is obviously nothing compared to what will happen if the eye tracks right up towards New Orleans... even the eyewall of a strong TS.

08/27/2012 23:30:00 CDT -0.16 1.79 1.95 24 8 27 1008.6 83.8 82.8 -99.9

08/27/2012 23:36:00 CDT -0.15 1.82 1.97 23 9 27 1008.5 83.6 82.8 -99.9

08/27/2012 23:42:00 CDT -0.15 1.84 1.99 24 13 28 1008.6 83.6 82.8 -99.9

08/27/2012 23:48:00 CDT -0.14 1.86 2.00 22 14 27 1008.6 83.6 82.8 -99.9

08/27/2012 23:54:00 CDT -0.14 1.86 2.00 23 19 26 1008.5 83.7 82.7 -99.9

08/28/2012 00:00:00 CDT -0.13 1.87 2.00 25 20 30 1008.5 83.7 82.7 -99.9

08/28/2012 00:06:00 CDT -0.12 1.88 2.00 25 20 29 1008.4 83.7 82.7 -99.9

08/28/2012 00:12:00 CDT -0.11 1.91 2.02 26 21 29 1008.2 83.7 82.7 -99.9

08/28/2012 00:18:00 CDT -0.10 1.92 2.02 26 22 31 1008.2 83.8 82.6 -99.9

08/28/2012 00:24:00 CDT -0.09 1.94 2.03 26 22 31 1008.2 83.7 82.6 -99.9

08/28/2012 00:30:00 CDT -0.08 1.97 2.05 26 20 31 1008.2 83.8 82.6 -99.9

08/28/2012 00:36:00 CDT -0.07 1.99 2.06 27 19 31 1008.0 83.7 82.5 -99.9

[dan88]

978.8mb extrap, 75kt flight level winds, SFMR 60kt in 12mm/hr precip (as well as a 62kt, but in 26 mm/hr)