Dr Grays 2005 Hurricane Season Verification Report

[cycloneye]

http://hurricane.atmos.colostate.edu/fo ... 5/nov2005/

Very interesting what the report says so read it from top to bottom and any comments about it you can replie here.

[HURAKAN]

Very interesting.

[mike815]

Yeah Dr. Grey does an amazing job. very interesting report.

[x-y-no]

So ... next year Dr. Gray will be stepping back from his PI role ... Guess we all get old eventually.

The whole report is really interesting, but I'm going to focus on a couple of things towards the end (you knew this was coming, right? )

---

In the section on the difference between the 2004/2005 seasons vs other recent active seasons, they say:

These past two seasons have been incredibly active and destructive. Although we believe that we are in an active multidecadal era for Atlantic basin tropical cyclone activity that began in 1995, we do not expect to see as many landfalls in the upcoming years as we have seen in 2004-2005. It has been the unfortunate combination of near-record activity combined with unusually strong onshore steering currents which has made these past two seasons so destructive.

and in the following section on why the steering currents were anomalous, they wrote:

A natural question emanating from the previous discussion is “What has caused the ridge to develop along the East Coast during the Atlantic hurricane season over the past two seasons?” We believe that a primary driver of the ridging along the East Coast has been a considerable warming of sea surface temperatures in the central Pacific. Figure 13 shows the difference in Pacific sea surface temperatures during August-October 2004-2005 from August-October 1995-2003. Central Pacific sea surface temperatures have warmed considerably over the past two seasons, and an anomalous wavetrain emanating from this heat source has propagated downstream across North America and into the Atlantic Ocean. Figure 14 shows the difference in 500 mb heights in the Northern Hemisphere during August-October 2004-2005 from August-October 1995-2003. One can clearly see the development of a ridge in the central Pacific near the center of the anomalous warming, with a downstream trough located in the western part of the United States. Ridging predominates along the East Coast, and another trough is seen near Iceland. This fairly stationary wavetrain setup imparts anomalously strong westward steering flow for most Atlantic basin tropical cyclones, and therefore we have witnessed many landfalls over the past two years.

I find this explanation very plausible. But the question immediately arises in my mind how they can argue that we shouldn't see as many landfalls in upcoming years. This may be true, of course (and regression towards the mean argues for it) but I don't see any reason why this anomalous SST pattern in the central Pacific would be more likely to dissipate than to persist in upcoming years. And they themselves say we can expect this elevated activity for at least another decade, so a decrease in landfalls relative to these years must depend on the expectation that the steering currents will shift away from the pattern of the last two years.

----

I've got a couple of bigger quarrels with the section "Is Global Warming Responsible for the Large Upswing in 2004-2005 US Hurricane Landfalls?"

First of all, they say:

The global warming arguments have been given much attention by many media references to recent papers claiming to show such a linkage.

The don't name the "recent papers", but they can only be talking about Kerry Emanuel's paper in Nature and the one by Wester et al in Science. But this statement badly mischarcterizes what those papers claim. Neither paper claims this year's high landfall rate is in any way caused by global warming, nor have I heard any serious researchers make such a claim.

Yes, some laymen and some members of the press may have drawn this conclusion, but it's not a valid tactic to try and tar this research with the erroneous interpretations on the part of third parties.


Later, they say:

There is no physical basis for assuming that global hurricane intensity or frequency is necessarily related to global mean surface temperature changes of less than ± 0.5oC. As the ocean surface warms, so too does global upper air temperatures to maintain conditionally unstable lapse-rates and global rainfall rates at their required values. Seasonal and monthly variations of sea surface temperature (SST) within individual storm basins show only very low correlations with monthly, seasonal, and yearly variations of hurricane activity. Other factors such as tropospheric vertical wind shear, surface pressure, low level vorticity, mid-level moisture, etc. play more dominant roles in explaining hurricane variability than do surface temperatures. Although there has been a general global warming over the last 30 years and particularly over the last 10 years, the SST increases in the individual tropical cyclone basins have been smaller (about half) and, according to the observations, have not brought about any significant increases in global major tropical cyclones except for the Atlantic which as has been discussed, has multi-decadal oscillations driven primarily by changes in Atlantic salinity. No credible observational evidence is available or likely will be available in the next few decades which will be able to directly associate global surface temperature change to changes in global hurricane frequency and intensity.

Now this argument works very well for frequency of storms, but not so well for intensity. All the factors named play a large role in cyclogenesis, but with the exception of mid-level moisture they don't play as large a role in the case of an established cyclone which is at least in part creating its own environment. Also, they have failed to offer any evidence or rationale for why one would expect these other factors to behave in a statistically significant fashion to counteract the effect of higher SSTs.

I also find the claim than incremental increases in SSTs would not have any relation to incrementally increased cumulative storm intensity very peculiar when it emanates from a group who's seasonal forecast methodology is so heavily dependent upon SST distribution in the Atlantic basin as driven by the AMO. I understand the theoretical possibility (described in the introduction) that individual factors of their multi-parameter forecast may show a negative correlation when considered in isolation, but I have never seen (nor do they offer) any evidence that this is true in the case of SSTs.


Finally, it seems to me they're making a curiously selective use of statistical argument when they say:

It is rare to have two consecutive years with such a strong simultaneous combination of high amounts of major hurricane activity together with especially favorable steering flow currents. The historical records and the laws of statistics indicate that the probability of seeing another two consecutive hurricane season like 2004-2005 is very low. Even though we expect to see the current active period of Atlantic major hurricane activity to continue for another 15-20 years, it is statistically unlikely that the coming 2006 and 2007 hurricane seasons, or the seasons which follow, will have the number of major hurricane US landfall events as we have seen in 2004-2005.

This is, of course, merely a restatement of the regression towards the mean argument. I have no problem with that in principle, but what they're ignoring is that if indeed the overall level of tropical cyclone activity is increasing (as the Emanuel and Webster papers indicate) then absent some related mechanism causing a statistical shift away from landfalling tracks, one must expect a long-term increase in intensity of landfalls.

[DoctorHurricane2003]

These past two seasons have been incredibly active and destructive. Although we believe that we are in an active multidecadal era for Atlantic basin tropical cyclone activity that began in 1995, we do not expect to see as many landfalls in the upcoming years as we have seen in 2004-2005. It has been the unfortunate combination of near-record activity combined with unusually strong onshore steering currents which has made these past two seasons so destructive.

Let's not jinx ourselves now... :\

[WindRunner]

Sad thing is that this is already out of date.

[dizzyfish]

Sad thing is that this is already out of date.

Yup. I'm thinking this may have to be updated......at least once.
God........I am so over this and last years season!

[dhweather]

He wisely left himself an "out", about the very rare development after the 18th.

Either way, I've only skimmed it over, but so far so good.

I have a lot of respect for him and his staff - they've done tremendous work over the years.

[donsutherland1]

X-y-no,

Just a quick note on U.S. major hurricane landfalls. Prior to 2005, the previous record for most landfalls in a season was 3. The most such landfalls in consecutive seasons was 5.

2005 saw 4 major hurricanes make U.S. landfall (beating the old record by 33%). 2004-05 saw 7 major hurricanes make U.S. landfall (beating the old record by 40%, which is even more impressive than the new single season record, in my view).

The record goes back to 1851 and research goes back even farther, but sparse populations in hurricane-prone coastlines likely understates some of the landfalling data from the 1800-1850 period.

[x-y-no]

X-y-no,

Just a quick note on U.S. major hurricane landfalls. Prior to 2005, the previous record for most landfalls in a season was 3. The most such landfalls in consecutive seasons was 5.

2005 saw 4 major hurricanes make U.S. landfall (beating the old record by 33%). 2004-05 saw 7 major hurricanes make U.S. landfall (beating the old record by 40%, which is even more impressive than the new single season record, in my view).

The record goes back to 1851 and research goes back even farther, but sparse populations in hurricane-prone coastlines likely understates some of the landfalling data from the 1800-1850 period.

Yes I understand that and acknowledged the statistical argument.

But here's my point: they offered two causes for this extraordinary number of landfalls - the peak in activity (related in large part to the AMO, but potentially also with climate change) and the anomalous warm area of SSTs in the north central Pacific. By their own reasoning, the former is likely to cause quite a few more very active seasons over the next 10 to 15 years (although getting all the way to Gamma again is of course statistically unlikely). So the question then becomes do we have any physical reason for believing that we won't be seeing the same anomalous warm SSTs in the cantral Pacific. And I don't see that they offer any reason other than the statistical argument for that either.

Now statistics is a fine tool, but there are two big problems with their statistical reasoning in this case:

First, when one is looking at a natural cycle that's 50 to 60 years long, 150 years of data (less than 3 cycles) is not a very good baseline for seeing what the limits in amplitude of that cycle are.

Second, expecting a regression towards historical means is only valid if there is no long-term trend in the data. And here I think their reasoning is kind of baffling. I'm familiar with Dr. Gray's criticisms of the methodology in Emanuel's and Webster's work (and I agree that a couple of his points merit serious consideration) but the fact that he's considered their work carefully enough to make such criticisms raises the question of how he could possibly misunderstand their results to the degree exhibited by their mischaracterization quoted above. And I keep coming back in my mind to the apparent inconsistency between their own methodology (which is heavily dependent on SSTs in the tropical Atlantic) and their argument that an incremental global increase in SSTs cannot have an impact on cumulative cyclone intensity. I just can't reconcile the two.

[Jim Hughes]

X-y-no,

Just a quick note on U.S. major hurricane landfalls. Prior to 2005, the previous record for most landfalls in a season was 3. The most such landfalls in consecutive seasons was 5.

2005 saw 4 major hurricanes make U.S. landfall (beating the old record by 33%). 2004-05 saw 7 major hurricanes make U.S. landfall (beating the old record by 40%, which is even more impressive than the new single season record, in my view).

The record goes back to 1851 and research goes back even farther, but sparse populations in hurricane-prone coastlines likely understates some of the landfalling data from the 1800-1850 period.

I have only read part of Gray's paper but I read his discussion part at the end of his paper as well some of Jan's comments and I have to kindly disagree with some of his groups perceptions.

This recent activity, regarding percentages, is not some statistical anomaly. There are reasons why things tend to happen. At least from the point of cause and effects. There is a fairly good chance that this land falling of the majors could continue for another 2-3 years.

I believe we tend to throw the word chaos around because we lack the true understanding of why things occur. Now I have no idea if you or Jan checked out my discussion regarding the cyclical nature of the AMO and the possible stratosphere relationship with it.

http://www.storm2k.org/phpbb2/viewtopic ... 28#1141428

I was hoping to have some open dialog about it but it seems to be unimportant to some people. This is to bad. I know from reading most of your discussions that you are aware of most of the feedbacks between the oceanic and atmospheric teleconnections. This is very important here.

I stayed away from certain things in my discussion and I barely touched base on why we may have been seeing this considerable increase in both land falls and majors the past few years. I did mention the 50 mb QBO directional change but that was really nothing. I just threw a bread crumb out there.

The point of the discussion was to basically point out as to how or why the cyclical nature of the AMO had flip flopped. Or what the AMO had made flip flop if it is the driving force.

I may write up a simple version of my previous discussion later today and stay away from certain other things.

Jim

[Dr. Jonah Rainwater]

Landfall activity during 2004 was concentrated along the United States East Coast and the Florida Peninsula; whereas, landfall activity was concentrated along the United States Gulf Coast during 2005. During 2004, one landfalling hurricane (Ivan – Category 3) was recorded along the Gulf Coast, while four landfalling hurricanes were recorded along the East Coast and the Florida Peninsula (Charley – Category 4, Frances – Category 2, Gaston – Category 1, and Jeanne – Category 3). During 2005, three major hurricanes (Dennis – Category 3, Katrina – Category 4, and Wilma – Category 3) made landfall along the Gulf Coast, while two hurricanes (and only one major hurricane) (Katrina – Category 1 and Wilma – Category 3) made landfall along the East Coast and Florida Peninsula.

How is it that Charley isn't a Gulf Coast landfall, but Wilma is?

[x-y-no]

Landfall activity during 2004 was concentrated along the United States East Coast and the Florida Peninsula; whereas, landfall activity was concentrated along the United States Gulf Coast during 2005. During 2004, one landfalling hurricane (Ivan – Category 3) was recorded along the Gulf Coast, while four landfalling hurricanes were recorded along the East Coast and the Florida Peninsula (Charley – Category 4, Frances – Category 2, Gaston – Category 1, and Jeanne – Category 3). During 2005, three major hurricanes (Dennis – Category 3, Katrina – Category 4, and Wilma – Category 3) made landfall along the Gulf Coast, while two hurricanes (and only one major hurricane) (Katrina – Category 1 and Wilma – Category 3) made landfall along the East Coast and Florida Peninsula.

How is it that Charley isn't a Gulf Coast landfall, but Wilma is?

They meant Rita as the third Gulf Coast landfall, obviously. They name Wilma twice - Gulf Coast and then properly as hitting the Florida peninsula as well.

[x-y-no]

I was hoping to have some open dialog about it but it seems to be unimportant to some people. This is to bad. I know from reading most of your discussions that you are aware of most of the feedbacks between the oceanic and atmospheric teleconnections. This is very important here.

I'm not ignoring that post, Jim, but there's a lot of material there to respond to and I really haven't had the free time to try and do it justice. I'll try to post a reply in that thread tomorrow.

Jan

[Jim Hughes]

I was hoping to have some open dialog about it but it seems to be unimportant to some people. This is to bad. I know from reading most of your discussions that you are aware of most of the feedbacks between the oceanic and atmospheric teleconnections. This is very important here.

I'm not ignoring that post, Jim, but there's a lot of material there to respond to and I really haven't had the free time to try and do it justice. I'll try to post a reply in that thread tomorrow.

Jan

Well I welcome your critique but like I said earlier I might write a much shorter version and just use parts of the original discussion for a data base regarding URL's/papers. I can go back and edit the original discussion with letters or an asterisk to make it easier. Then someone could go back and look at just certain parts because I know it is very long.

These would be the parts of the discussion that I gave URL's for certain papers that were the most important regarding graphs and time tables between the AMO and the 30 & 50 hPa temperature trends. No space weather.


Jim