Hurricane Katia

[phil882]

Right on cue, the 0Z Euro once again teases but doesn't hit the east coast as expected.

Adam brought this up earlier... but he is totally right. There is no way we get this system to go up the coast like Irene because we need an amplified ridge off the east coast instead of a skinny flat ridge like the ECWMF has been suggesting. Thus, the only way we see a United States impact is if the skinny ridge continues to build westward over the Eastern United States. This does not occur because we have the remnants of TD13 over the southeast which is helping to maintain a weakness. Unless this changes significantly, it remains pretty unlikely we will see Katia landfall in the United States. What will happen is that it will reach the western extent of the ridge and do a fairly quick U-turn around the edge and be steered out to sea by a combination of the upper level westerly flow to the north and the upper part of the ridge axis.

[QVectorman]

Looking at the EGRR/UKMET's performance with position so far with Katia it seems to have a handle on it in the deep tropics better than any other models, especially the GFS. Thus, I think it should have a fairly decent shot at having a reasonable solution(compared to the other models) for 120-144 hr forecast period based on its previous performance with this storm. Interestingly enough, 120-144 hrs has the storm just north of Puerto Rico which could lead to more interesting solutions beyond 144 hrs. Either way, it's the best performing so far and its solution is way south of the current ECMWF or GFS. So I think this leaves the possibility we could see it some where in the Caribbean and then who knows where after that!

Here is the 0z average model position error for the life of Katia EGRR is the UKMET and EGR2 is the UKMET 12 hr interpolated continue to be the best performers

Also if we look at the model bias the EGRR=UKMET shows a consistant bias of about 135-157 degrees.

And the trend of the UKMET through 18z so far (trending it south)

Finally, the UKMET at 144 hrs, which from the 120 hr forecast Katia appears to take a jog directly to the SW

[GaWx]

Looking at the EGRR/UKMET's performance with position so far with Katia it seems to have a handle on it in the deep tropics better than any other models, especially the GFS. Thus, I think it should have a fairly decent shot at having a reasonable solution(compared to the other models) for 120-144 hr forecast period based on its previous performance with this storm.

Wasn't the UKMET pretty bad with Irene around the U.S.? I recall it really stinking and being close to the worst medium range global. If so, why should I expect it to do much better this time despite what you said about its performance so far in the deep tropics with Katia?

[phil882]

Definitely some improvement with the inner core this morning... should be upgraded back soon.

[QVectorman]

Wasn't the UKMET pretty bad with Irene around the U.S.? I recall it really stinking and being close to the worst medium range global. If so, why should I expect it to do much better this time despite what you said about its performance so far in the deep tropics with Katia?

I never said to follow it through 168 or 192 hours or through a landfall or recurve...I just said it may have more realistic positions over the next 120 hrs vs. other models. Which the end result is it's still in the region 20 N after 144 hrs.

Every model performs differently with each storm. It's doing well so I feel it may have a more accurate solution than the ECMWF or GFS so it's worth keeping in mind while the storm is in the tropics. Once it begins it's trip out of the tropics I believe another model will probably take the reigns and perform better positionwise than the UKMET. Perfect example, for Irene the AEMI - GFS Ensemble Mean interpolated at 6 hr intervals kicked butt with position and the SHIP was stellar on intensity. Well so far neither one of those have performed well with Katia. Why should I buy the GFS right now, it did well with Irene but for 54 hr position forecasts its averaging 202 nm or 232 miles errors while the UKMET and Canadian are averaging 47nm or 54 mile errors

[HurricaneJosh]

Looking at the EGRR/UKMET's performance with position so far with Katia it seems to have a handle on it in the deep tropics better than any other models, especially the GFS. Thus, I think it should have a fairly decent shot at having a reasonable solution(compared to the other models) for 120-144 hr forecast period based on its previous performance with this storm. Interestingly enough, 120-144 hrs has the storm just north of Puerto Rico which could lead to more interesting solutions beyond 144 hrs. Either way, it's the best performing so far and its solution is way south of the current ECMWF or GFS. So I think this leaves the possibility we could see it some where in the Caribbean and then who knows where after that!

Well, that sure would be a mega-score for the lone, far-left outlier if it happened to verify with this system. I'm sure not banking on it, but it's an interesting point you raise-- I didn't realize it had been outperforming the other models with Katia up to now-- and of course I'd be delighted if you end up being right.

[atownwxwatcher]

Looking at the EGRR/UKMET's performance with position so far with Katia it seems to have a handle on it in the deep tropics better than any other models, especially the GFS. Thus, I think it should have a fairly decent shot at having a reasonable solution(compared to the other models) for 120-144 hr forecast period based on its previous performance with this storm. Interestingly enough, 120-144 hrs has the storm just north of Puerto Rico which could lead to more interesting solutions beyond 144 hrs. Either way, it's the best performing so far and its solution is way south of the current ECMWF or GFS. So I think this leaves the possibility we could see it some where in the Caribbean and then who knows where after that!

Here is the 0z average model position error for the life of Katia EGRR is the UKMET and EGR2 is the UKMET 12 hr interpolated continue to be the best performers

Also if we look at the model bias the EGRR=UKMET shows a consistant bias of about 135-157 degrees.

And the trend of the UKMET through 18z so far (trending it south)

Finally, the UKMET at 144 hrs, which from the 120 hr forecast Katia appears to take a jog directly to the SW

Good information here but NHC disagrees and calls the UK an outlier and is disregarding it...

Also they are not expecting it to strengthen today at all...

[thewxmann]

UKMET has a pretty notorious south bias... I've seen it in various storms ranging from Bill (2009) to Irene, among others.

[BlizzardNole]

Interesting snippet about Katia from the forecast discussion out of the New Orleans NWS office:

GOING TO BE A WHILE BEFORE WE GET RID OF OUR TROPICAL SYSTEM.

STEERING WINDS AND FORCING IS VERY WEAK THROUGH THE NEXT SEVERAL

DAYS. BUT AS IT DOES MOVE INLAND IT BECOMES MORE BAROCLINIC AND

EVEN DEVELOPS FRONTOGENESIS TO ITS SOUTH. THIS SHOULD HELP DRY THE

AREA OUT BY MID NEXT WEEK. THIS WHOLE SYSTEM WILL THEN BE

RESPONSIBLE FOR PICKING UP THE NEXT MAJOR HURRICANE THAT IS NOT

LOOKING TOO FRIENDLY TO THE EAST COAST. HOPE THIS THING TURNS INTO

THE UPPER TROUGH WELL OFFSHORE. BECAUSE INSTEAD OF SKIRTING LAND

LIKE IRENE ON ITS NORTHWARD JOURNEY...THIS ONE IS MAKING A B-LINE

FOR THE EAST COAST WHICH DOES NOT ALLOW IT TO WEAKEN MUCH BEFORE SUPPOSEDLY

TURNING.

[CoastalWx]

06z GFS actually retrogrades that trough into the deep south and that building ridge pushes Katia west.

[VAwxman]

Adam brought this up earlier... but he is totally right. There is no way we get this system to go up the coast like Irene because we need an amplified ridge off the east coast instead of a skinny flat ridge like the ECWMF has been suggesting. Thus, the only way we see a United States impact is if the skinny ridge continues to build westward over the Eastern United States. This does not occur because we have the remnants of TD13 over the southeast which is helping to maintain a weakness. Unless this changes significantly, it remains pretty unlikely we will see Katia landfall in the United States. What will happen is that it will reach the western extent of the ridge and do a fairly quick U-turn around the edge and be steered out to sea by a combination of the upper level westerly flow to the north and the upper part of the ridge axis.

Agree that the most likely path is a curve back out to sea, but I disagree with some of the reasoning why.

The remnants of TD13, as currently modeled, are projected to leave the weakness / ULL more back over the TN valley, which is plenty west enough to allow Katia to hit the East Coast. The problem is the northern stream trough is strong enough in the Northeast to erode the western Atlantic ridge enough to allow the escape north. While not likely, I don't think it would take a significant change to make it more interesting. All it would take is a little faster northern stream, with the next trough into the Midwest stronger / faster, and getting rid of the NE trough faster and allowing the NW Atlantic ridge to come in more than currently projected.

Obviously no model shows this, so we have to favor strongly the recurve.

[Ed Lizard]

Agree that the most likely path is a curve back out to sea, but I disagree with some of the reasoning why.

The remnants of TD13, as currently modeled, are projected to leave the weakness / ULL more back over the TN valley, which is plenty west enough to allow Katia to hit the East Coast. The problem is the northern stream trough is strong enough in the Northeast to erode the western Atlantic ridge enough to allow the escape north. While not likely, I don't think it would take a significant change to make it more interesting. All it would take is a little faster northern stream, with the next trough into the Midwest stronger / faster, and getting rid of the NE trough faster and allowing the NW Atlantic ridge to come in more than currently projected.

Obviously no model shows this, so we have to favor strongly the recurve.

More so the Canadian than GFS, but both would seem to suggest TD 13s remains will impart an almost neutral tilt on the trough and allow Katia to get close enough to at least make it interesting.

[donsutherland1]

Overnight, Katia stabilized. It appears that the storm may have strengthened somewhat toward morning and could be near or at hurricane strength if some of the satellite estimates are accurate. Katia is continuing to head toward an area of diminishing shear.

In addition, Katia continues to somewhat to the west of the NHC's estimates. For example, its 5 am position at 17.0N 51.8W was about 25-35 miles west of the 9/1 5 pm and 11 pm NHC forecast tracks at 17.0N. Prospects that Katia will ultimately reach 70W-75W are continuing to grow.

[OSUmetstud]

latest microwave shows improvement...probably a hurricane given SSD Dvorak of 4.0 and that microwave.

[NJwinter23]

Adam brought this up earlier... but he is totally right. There is no way we get this system to go up the coast like Irene because we need an amplified ridge off the east coast instead of a skinny flat ridge like the ECWMF has been suggesting. Thus, the only way we see a United States impact is if the skinny ridge continues to build westward over the Eastern United States. This does not occur because we have the remnants of TD13 over the southeast which is helping to maintain a weakness. Unless this changes significantly, it remains pretty unlikely we will see Katia landfall in the United States. What will happen is that it will reach the western extent of the ridge and do a fairly quick U-turn around the edge and be steered out to sea by a combination of the upper level westerly flow to the north and the upper part of the ridge axis.

Though let's say TD13 becomes a cutoff low of sorts in the southern U.S. next week with the troughs bypassing to the north, wouldnt this turn into a favorable steering pattern to draw Katia into the East coast?

[VAwxman]

More so the Canadian than GFS, but both would seem to suggest TD 13s remains will impart an almost neutral tilt on the trough and allow Katia to get close enough to at least make it interesting.

Right. Euro is kinda in that camp with the GEM too. All it would take is for the trough diving through west-central Canada to be stronger, and thus building heights more in eastern Canada and New England and it is a different ball game. Again, not forecasting that, but just pointing out that it would not take a whole lot.

[NJwinter23]

Right. Euro is kinda in that camp with the GEM too. All it would take is for the trough diving through west-central Canada to be stronger, and thus building heights more in eastern Canada and New England and it is a different ball game. Again, not forecasting that, but just pointing out that it would not take a whole lot.

Exactly what I was thinking, especially looking at the 6z gfs today. For being a week+ away, what you described is definitely not out of the question, particularly since we don't have a great idea what Lee is going to do just yet.

[LithiaWx]

We may have Hurricane Katia again, this model page refers to her as Hurricane.

http://moe.met.fsu.e...ls/al122011.png

http://moe.met.fsu.e...acevans/models/

Also Lists TD 13 as Tropical Storm 13

[Juliancolton]

We may have Hurricane Katia again, this model page refers to her as Hurricane.

http://moe.met.fsu.e...ls/al122011.png

http://moe.met.fsu.e...acevans/models/

Also Lists TD 13 as Tropical Storm 13

AL, 13, 2011090212, , BEST, 0, 270N, 916W, 35, 1005, TS, 34, NEQ, 100, 100, 0, 0, 1010, 180, 80, 0, 0, L, 0, , 0, 0, THIRTEEN, M,

AL, 12, 2011090212, , BEST, 0, 174N, 520W, 65, 991, HU, 64, NEQ, 30, 0, 0, 0, 1013, 250, 25, 80, 0, L, 0, , 0, 0, KATIA, D,

[GaWx]

I never said to follow it through 168 or 192 hours or through a landfall or recurve...I just said it may have more realistic positions over the next 120 hrs vs. other models. Which the end result is it's still in the region 20 N after 144 hrs.

Every model performs differently with each storm. It's doing well so I feel it may have a more accurate solution than the ECMWF or GFS so it's worth keeping in mind while the storm is in the tropics. Once it begins it's trip out of the tropics I believe another model will probably take the reigns and perform better positionwise than the UKMET. Perfect example, for Irene the AEMI - GFS Ensemble Mean interpolated at 6 hr intervals kicked butt with position and the SHIP was stellar on intensity. Well so far neither one of those have performed well with Katia. Why should I buy the GFS right now, it did well with Irene but for 54 hr position forecasts its averaging 202 nm or 232 miles errors while the UKMET and Canadian are averaging 47nm or 54 mile errors

It isn't just the GFS. The Euro is still not hitting the U.S. even though it it is doing a great job of teasing. I see your point about the UKMET. Perhaps, it will continue to do better for Katia. However, I'm not holding my breath since the UKMET is so far left of the consensus and I believe that it tends to have somewhat of a left bias. With the GFS/Euro still not having a run hitting the U.S. (even though they're getting closer and teasing the weenies), I'm still going with 90% chance of missing the U.S. This is such an atypical pattern for an east coast landfall. Close? Maybe. An actual hit? Very unlikely.

[donsutherland1]

Several additional thoughts:

First, the satellite presenation has continued to improve. I believe Katia has probably reached hurricane strength.

Second, in terms of the FSU-generated analogs (9/2 12z), I believe a number of them are not very good. The worst ones, IMO, are Fox (1950), Holly (1969), Frances (1976), and Irene (1981). The synoptic pattern does not support such early recurvature this time around.

FWIW, my ten analogs at this point would be: H4 (1893), H1 (1927), Gladys (1975), Gloria (1985), Luis (1995), Edouard (1996), Bonnie (1998), Floyd (1999), Bill (2009), and Earl (2010). Those with perhaps the closest 500 mb pattern to what is expected on the 9/2 0z GFS ensembles between 168h and 192h would be, in chronological order, Gladys (1975), Luis (1995), and Floyd (1999).

The mean and median positions for the farthest west those 10 storms came were 73.8W and 73.6W respectively. 80% of the storms reached 70.0W, 70% reached 72.5W, and 40% reached 75.0W. The 9/2 0z Euro comes to around 73.0W, perhaps a little farther when one considers a curved not linear track.

U.S. landfall occurred in 40% of the cases. Once one takes the more similar synoptic patterns (relative to that on the GFS ensembles) into consideration, 33% made U.S. landfall.

70% of the above-mentioned storms reached Category 4. All grew into major hurricanes.

Right now, if I had to venture a guess concerning Katia's track, that guess would be:

17.5N 52.8W

20.0N 56.8W

25.0N 63.9W

30.0N 72.0W

35.0N 69.6W

40.0N 63.5W

Farthest west: probably somewhere between 72.0W and 74.0W. Hence, there is room for a track farther west than what I've outlined.

Finally, I still believe odds favor recurvature. But prospects for a landfall are not insignificant and only a modest change in the evolution of the 500 mb pattern and Katia's response to it could lead to such an outcome. A Floyd-like track would probably be a worst-case scenario. Hurricane #1 (1927) and Gladys (1975) might offer the best analogs. Of course, things can still change.

[OSUmetstud]

hurricane again

000WTNT42 KNHC 021454TCDAT2HURRICANE KATIA DISCUSSION NUMBER 18NWS NATIONAL HURRICANE CENTER MIAMI FL AL1220111100 AM AST FRI SEP 02 2011THE CONVECTIVE DEPICTION OF KATIA HAS IMPROVED THIS MORNING AS BOTHMICROWAVE IMAGERY FROM SSMIS AND CONVENTIONAL VISIBLE SATELLITEIMAGERY INDICATE THAT AN EYE MAY BE FORMING. BECAUSE OF THEIMPROVED STRUCTURE...BOTH TAFB AND SAB DVORAK CLASSIFICATIONS CAMEIN AT A 4.0...OR 65 KT...AS DID A 1053Z CIMSS AMSU INTENSITYESTIMATE. THUS KATIA HAS REGAINED HURRICANE STATUS.THE HURRICANE IS MOVING TOWARD THE WEST-NORTHWEST AT ABOUT 12KT...AS IT IS BEING ADVECTED AROUND THE SOUTHWEST PERIPHERY OF ADEEP-LAYER RIDGE. KATIA SHOULD MOVE TOWARD THE WEST-NORTHWEST ORNORTHWEST FOR THE NEXT FEW DAYS AT A SLIGHTLY SLOWER RATE OF SPEED. ALL OF THE RELIABLE GUIDANCE...WITH THE EXCEPTION OF THE UKMET WHICHHAS A MUCH MORE EQUATORWARD SOLUTION DUE TO A SHORT-WAVE TROUGH OFFOF THE U.S. EAST COAST LIFTING OUT TO THE NORTHEAST MORE QUICKLY...IS IN EXCELLENT AGREEMENT. THE TRACK FORECAST IS VERY SIMILAR TOTHE PREVIOUS ADVISORY AND THE CONSENSUS OF THE GLOBAL AND HURRICANEMODELS LESS THE UKMET MODEL.DESPITE BECOMING A HURRICANE AGAIN...KATIA FACES A RELATIVELYHOSTILE ENVIRONMENT WITH SOUTHWEST VERTICAL SHEAR OF 15-20 KTSUGGESTED FOR THE NEXT DAY OR TWO. BEYOND 48 HOURS...THE SHEARFROM THE GLOBAL MODELS IS ANTICIPATED TO DECREASE...BUT THETHERMODYNAMICS MAY NOT BE IDEAL WITH LOW HUMIDITIES AND A MORESTABLE ENVIRONMENT ANTICIPATED. SOME OF THE UPPER-LEVELTEMPERATURE INCREASES ANALYZED IN THE SHIPS/LGEM STATISTICALMODELS...HOWEVER...MAY BE SPURIOUS...AS THE STRONGER VORTEX FROMTHE GFS MAY BE THE CAUSE OF THESE VALUES RATHER THAN THEENVIRONMENT. THUS THE SHIPS/LGEM GUIDANCE MAY BE TOO LOW AT DAYSTHREE TO FIVE. THE INTENSITY PREDICTION IS SLIGHTLY LOWER THAN ASHIPS/LGEM/HWRF BLEND BUT HIGHER THAN THE PREVIOUS ADVISORY DURINGTHE NEXT TWO DAYS.FORECAST POSITIONS AND MAX WINDSINIT 02/1500Z 17.5N 52.4W 65 KT 75 MPH 12H 03/0000Z 18.6N 53.8W 65 KT 75 MPH 24H 03/1200Z 19.6N 55.2W 70 KT 80 MPH 36H 04/0000Z 20.6N 56.4W 75 KT 85 MPH 48H 04/1200Z 21.7N 57.9W 80 KT 90 MPH 72H 05/1200Z 24.0N 61.0W 90 KT 105 MPH 96H 06/1200Z 26.0N 64.0W 95 KT 110 MPH120H 07/1200Z 27.0N 67.0W 100 KT 115 MPH$$FORECASTER LANDSEA/STEWART

[eurojosh]

I've been waiting for a Met to post, but nothing yet, so let me go... The 1100 guidance on Katia projects a slightly faster strengthening than the 0500, but a slower strengthening than previous days' guidance. Most interesting, though, is the continued west trend in the D5 projected track:

[Eyewall2005]

Jeff masters this morning...

"[Tropical Storm] Talas is expected to hit Japan early on Saturday as a strong tropical storm, then race northwestwards into the Bering Sea off the coast of Alaska. Talas is then expected to transition into a powerful extratropical storm in the waters south of Alaska. This storm will create a ripple effect downstream in the jet stream, all the way to North America, by early next week. The timing and amplitude of the trough of low pressure off the U.S. East Coast expected to potentially recurve Katia out to sea next week is highly dependent upon the strength of Tropical Storm Talas during its transition to an extratropical storm"

[Clinch Leatherwood]

Ridge nosing west a bit more than the 0z but at the same time the remnants of the GOM TS is further along. Losing the weak ridging 87 to 90 hours on the GFS to the NNW of Katia.

[Clinch Leatherwood]

GFS slides the remnant system up towards KY by 130 hours, where it spins, maybe retrogrades a touch. Could be a pretty interesting scenario if it were to hold in place/strengthen with Katia out to the east.

What do the pros think of the new GFS?

[Clinch Leatherwood]

Westerlies trying to dip down with the system coming from Hudsons Bay, but with an UL to the west, it's an interesting situation on the GFS.

[wxsniss]

Westerlies trying to dip down with the system coming from Hudsons Bay, but with an UL to the west, it's an interesting situation on the GFS.

someone on main thread mentioned this

remnants of Lee acting like a closed off low, and consequently westerlies remaining further north

[MJO812]

It looks like she skims Nantucket on this run.

[VAwxman]

Right. Euro is kinda in that camp with the GEM too. All it would take is for the trough diving through west-central Canada to be stronger, and thus building heights more in eastern Canada and New England and it is a different ball game. Again, not forecasting that, but just pointing out that it would not take a whole lot.

GFS actually shows that next trough being stronger, though is so fast that it still kicks Katia out just in time. Close miss for southeast New England though.