The last (and my favorite) of the major hurricane seasonal forecasts came out yesterday. Phil Klotzbach and Bill Gray from Colorado State are sticking with their April forecast. Rather anti-climactic, but perhaps it’s a measure of the confidence in the forecast. I didn’t have a chance to write up something in April, so I’ll do it now before normalcy, if it sets in, brings six weeks of inactivity.
http://hurricane.atmos.colostate.edu/Forecasts is used in most of the following. My comments are in italics.
EXTENDED RANGE FORECAST OF ATLANTIC SEASONAL HURRICANE ACTIVITY AND LANDFALL STRIKE PROBABILITY FOR 2011
We continue to foresee well above-average activity for the 2011 Atlantic hurricane season. We are predicting the same levels of activity that were forecast in early April due to the combination of expected neutral ENSO conditions and very favorable atmospheric and oceanic conditions in the tropical Atlantic. We continue to anticipate an above-average probability of United States and Caribbean major hurricane landfall.
Note – in the table below, I changed K&G’s June numbers to the range that include ±1 standard deviation of their hindcast error and added a column with NOAA’s forecast.
| Forecast Parameter and 1950-2000
Climatology (in parentheses) |
Issue Date
8 December 2010 |
Issue Date
6 April 2011 |
Issue Date
1 June 2011 |
NOAA |
| Named Storms (NS) (9.6) | 17 | 16 | 12.3 – 19.7 | 12-18 |
| Named Storm Days (NSD) (49.1) | 85 | 80 | 59.9 – 101.1 | |
| Hurricanes (H) (5.9) | 9 | 9 | 6.9 – 11.1 | 6-10 |
| Hurricane Days (HD) (24.5) | 40 | 35 | 24.8 – 45.2 | |
| Major Hurricanes (MH) (2.3) | 5 | 5 | 3.4 – 6.6 | 3-6 |
| Major Hurricane Days (MHD) (5.0) | 10 | 10 | 4.7 – 15.3 | |
| Accumulated Cyclone Energy (ACE) (96.1) | 165 | 160 | 112 – 212 | |
| Net Tropical Cyclone Activity (NTC) (100%) | 180 | 175 | 127 – 223 |
PROBABILITIES FOR AT LEAST ONE MAJOR (CATEGORY 3-4-5) HURRICANE LANDFALL ON EACH OF THE FOLLOWING COASTAL AREAS:
- Entire U.S. coastline – 72% (average for last century is 52%)
- U.S. East Coast Including Peninsula Florida – 48% (average for last century is 31%)
- Gulf Coast from the Florida Panhandle westward to Brownsville – 47% (average for
last century is 30%)
PROBABILITY FOR AT LEAST ONE MAJOR (CATEGORY 3-4-5) HURRICANE TRACKING INTO THE CARIBBEAN (10-20°N, 60-88°W)
- 61% (average for last century is 42%)
Acknowledgment
This year’s forecasts are funded by private and personal funds. We thank the GeoGraphics Laboratory at Bridgewater State College (MA) for their assistance in developing the United States Landfalling Hurricane Probability Webpage (available online at http://www.e-transit.org/hurricane).
The forecast is a blend of a statistical analysis, a search for analog years with similarities to April-May conditions and the expected August-October conditions, and “qualitative adjustments for other factors not explicitly contained in any of these schemes.”
As they often do, the June forecast is based on a new June statistical model and uses just four predictors (the geographic ranges are in the image at the top):
| Predictor | 2011 Forecast Value |
| 1) April-May SST (15-55°N, 15-35°W) (+) | +0.3 SD |
| 2) April-May 200 MB U (0-15°S, 150°E-20°W) (+) | +1.6 SD |
| 3) ECMWF 1 May SST Forecast for September Nino 3 (5°S-5°N, 90-150°W) (-) | +0.3 SD |
| 4) May SLP (20-40°N, 30-50°W) (-) | -0.9 SD |
Decoding key: SST: Sea Surface Temperature; SLP: Sea Level Pressure, SD: Standard deviation; 200 MB U: zonal wind at atmospheric pressure 200 mb, an altitude of about 40,000 feet.
Other notes:
ENSO
The moderate-to-strong La Niña conditions of this past winter have continued to weaken since early April. According to the Climate Prediction Center (CPC), SST anomalies in the central and eastern tropical Pacific have now warmed to the point where ENSO is in its neutral state. While the moderation of tropical Pacific SSTs have continued over the past couple of months, we do not expect to see a transition to El Niño conditions during the next several months.
Current Atlantic Basin Conditions
Conditions in the Atlantic remain favorable for an active season. May SSTs across the tropical Atlantic remain at above-average levels (Figure 17). The anomalously strong trades during the latter part of the winter that caused some anomalous cooling in the tropical Atlantic have now weakened (Figure 18). Sea level pressure anomalies during the month of May were also well below-average throughout the tropical Atlantic (Figure 19). Altogether, tropical Atlantic conditions currently reflect an environment conducive for an active hurricane season.
Forthcoming Updated Forecasts of 2011 Hurricane Activity
We will be issuing a final seasonal updates of our 2011 Atlantic basin hurricane forecast on Wednesday, 3 August. We will also be issuing two-week forecasts for Atlantic TC activity during the climatological peak of the season from August-October. A late-season forecast for the Caribbean will be issued in early October. A verification and discussion of all 2011 forecasts will be issued in late November 2011. All of these forecasts will be available on the web at: http://hurricane.atmos.colostate.edu/Forecasts.
I will likely report on the Aug. 3rd update, but will probably not cover the two-week forecasts. Last year, several were dominated by current activity, low or high.
Wade says:
June 3, 2011 at 10:00 am
I don’t have time now to check the words in the forecast, but in 2009 Klotzbach and Gray forecast 90% of average. See my note above from 6:19 am. That season wound up with only 69% of average. Of course, several forecasts before the positive AMO arrived in 1995 had forecasts for below average activity.
@jon
Hardly.
But if that’s the level at which you prefer to engage, I’m sure my grandson would oblige.
Ric Werme says:
June 3, 2011 at 6:19 am
Yes, a couple years were way over forecast,
========================================================
Ric, you only gave examples of the past five years…..
..and four of the past five were over forecast, three way over forecast
With the new equipment the hurricane forecasters are using – naming any two clouds within sight of each other, or any cloud that has circulation if for only 10 minutes – there is no more “normal” or “average” any more.
I thought the Sahel dust served as nucleation sites helping to start the thunderstorms that eventually give birth to Atlantic hurricanes.
John F. Hultquist says:
June 2, 2011 at 10:20 pm
“I went to a lecture at the local university tonight. The speaker noted that scientists are often wrong but there is not anything wrong with that. What is wrong, is when they don’t learn anything from being wrong. The commenters here (and elsewhere) are mostly just having a bit of fun with the early-stage learning curve of hurricane prediction.”
Very Well Said, Sir. If these so-called “hurricane season predictions” are to be treated as predictions then what is one to do with them. I can imagine some things as follows:
1. Do not hold a sailboat regata of 1,000,000 or more sailboats during August or September in the area southwest of the Dominican Republic.
2. Do not plan a family reunion that would draw more than 10,000 people during August or September in the area of Homestead, Florida. However, if you are willing to move a few miles to the northwest, you will be perfectly safe in Sebring.
3. If you are starting a new cruise line that travels from the Dominican Republic to West Africa, do not schedule the maiden voyage for August or September.
I think you get the idea. Really, can someone tell me how these so-called “hurricane predictions” are more useful than what I suggest above?
crosspatch: Enhanced CO2 is probably playing a role in the Sahara greening. CO2 helps plants use water more efficiently, which all other things being equal would allow them to survive in dryer climates.
@jon
My grandson wearies of this O/T engagement.
Au’voir
Averages and even “normals” are misleading. In many cases one should probably use the median, not the “average” when figuring for things like rainfall amounts or number of hurricanes. That should put you in a ballpark range. Averages tend to be completely useless, particularly on things that have a large standard deviation. A median is often better.
Latitude says:
June 3, 2011 at 11:29 am
Ric Werme says:
June 3, 2011 at 6:19 am
Given the size of the standard deviation of the current state of the art, predicting 90% of average in 2009 and getting 69% isn’t too bad. The current statistical approach applied as a hindcast predicts 75%.
The five examples are all they list in the end of the forecast, I didn’t have time to dig up more. What I should have done anyway is quote from their table of hindcasts using the new forecasting scheme. Note, though, the data I did quote was the final forecast, made up from the statistical computation, the analog years, and their “qualitative” (seat of the pants?) numbers.
Here’s the table how well the new scheme does. For all years, “climatology” is 100 – as in 100% of the 1950-2000 average of NTC – Net Tropical Cyclone Activity.
Note how much NTC varies from year to year, and note the big change in 1995 when the AMO switched. Not all years are better forecast, for example, the 2010 scheme predicted 195%, very close to the 196% observed.
Sorry but I’m with savetheshark. Some of the comments at the start were particularly brainless and quite nasty – just empty sarcasm. People with nothing constructive and intelligent to say shouldn’t be mouthing off and putting down a guy trying to do his best at a very difficult job.
Some people seem to come here purely in order to make negative comments about CAGW. I can understand the urge to vent sometimes. But in this case I suspect many did not even read the article beyond a quick glance to see that it was about hurricanes so that they could theme their vitriolic and sarcastic `witty’ comments appropriately.
I don’t want all skeptics to be associated by proximity with this kind of mindless negativity. Anthony often posts information simply because it interests him. The articles are not simply topic headings along the lines of “what we are going to moan about today”, which is how some seem to be treating them.
Ian H says:
June 3, 2011 at 7:18 pm
My goal in posting here is to explain and defend scientific method. My posts have been intended to challenge the idea that “hurricane forecasts” are sufficiently rigorously formulated that they are deserving of the word ‘prediction’ or the word ‘science’. If we are going to call this science, then isn’t everything that is based on meticulous research to be counted as science? Obviously, these “hurricane forecasts” are not falsifiable and that is for a very specific reason. There are no hypotheses used in the forecasts that must be rejected as falsified if this season has no hurricanes at all. Now, my claim might be false, all you have to do is show and explain the falsifiable hypothesis. In plain speech, how do these forecasters challenge themselves when they make a forecast? How can they be wrong?
Theo Goodwin says:
June 4, 2011 at 8:22 am (Edit)
In 2006 the June forecast was for NTC of 195, the observed was 85, several standard deviations away from their usual performance. In “Met speak,” the forecast did not verify. One of the hypotheses behind the forecasts is that El Nino depresses NTC.
In http://hurricane.atmos.colostate.edu/Forecasts/2006/nov2006/ they report:
While the prediction summary was for 195%, the hypothesis remains El Nino will suppress NTC, and hence the hypothesis survived that test. Whatever hypotheses go into El Nino predictions may not have.
BTW, that year also provided a lot information about the role of Saharan dust.
If you want, you can consider this part of the genesis of a new hypothesis governing Atlantic Tropical Storms. That’s one of the reasons I like the K & G forecasts, you just don’t get this sort of detail from any other seasonal forecast.
You stated “In plain speech, how do these forecasters challenge themselves when they make a forecast? How can they be wrong?”
Their forecast didn’t verify. (Their words “was not successful” – a euphemism for your “wrong.”) That by itself is a challenge (your word and theirs) to make a better forecast the next time.
Ric Werme says:
June 4, 2011 at 10:16 am
Theo Goodwin says:
June 4, 2011 at 8:22 am (Edit)
Thank You for your very informative and very interesting response. My concern is to safeguard scientific method. From the information that you provide, I conclude that these forecasters were unable to make a prediction that would challenge one or more of their hypotheses by putting it up for falsification. So, their work remains at the level of pre-science. They have incredibly good hunches and they really should pursue experimental research into the dust from Africa and into the behavior of La Nina. Of course, those undertakings are very expensive and will not be made until some national governments commit to them.
Consider the following:
“Therefore, we are inclined to believe that a large part of this mid-tropospheric dryness in August was due to frequent outbreaks of African dust associated with more frequent incursions of the Saharan Air Layer (SAL) (Evan et al. 2006).”
Notice two things. One, they were unable to predict this matter. In addition, they are not able to identify some one or few of their hypotheses for revision. Therefore, they are working on hunches. No doubt the hunches are extremely brilliant, but they are still hunches and should not be treated as physical hypotheses of science.
Consider the following:
“One of the extraordinary features of the 2006 Atlantic basin hurricane season has been the rapid onset of El Niño conditions in the tropical Pacific.”
They were unable to predict this matter. And, as above, their system of hunches are not rigorously formulated so that they can identify one or a few as culprits and reject them. There is a bigger lesson here. These folks cannot predict the behavior of La Nina. I believe no one can. I believe no one has done the legwork necessary to identify and describe the natural processes that constitute the La Nina phenomenon. Until that is done in a rigorous way, there will be no prediction of La Nina onset or whatever.
Describing phenomena such as La Nina in physical hypotheses is something that must be done before climate science can get off the ground. To think otherwise is just to fool ourselves about the extent of our knowledge about climate.
Thanks again. I enjoyed your post.
Jim Clarke says:
June 3, 2011 at 9:45 am
Thanks for pointing out the flak that Bill Gray has endured. Don’t forget who was vice president when Hansen was scooping up all the funding for his more-or-less useless climate models. It was Gore.
The money would have been more wisely spent on the things Bill Gray was curious about. It would have given us real data, rather than a computer generated dream-land.
One thing Gray wanted studied more was thermohaline circulation. To this day much of how it actually works remains a mystery. We have a general idea, but specific details remain unreserched. It apparently is not a steady flow, like a river or garden hose, but holds waves and/or pulses of some sort, which may-or-may-not influence the rate cold water upwells. Right now we have a good idea of how trade winds influence upwelling, but, if another factor also contributes to upwelling, then forecasts based only on trade winds will get messed up.
The data Gray thirsted for would have increased our understanding of how the world actually works. Instead he was rudely and brutally told, “Stick to hurricanes, Bill.” (Hansen is the one who should have been told to stick something some place. )
It is sad and ironic that Gray now gets the backlash Hansen has earned. It shows you the horrible effect pseudo-science has had on the public psyche.
Caleb says:
June 5, 2011 at 2:42 am
“The money would have been more wisely spent on the things Bill Gray was curious about. It would have given us real data, rather than a computer generated dream-land.”
Thanks for this interesting post. I am no critic of Gray. I believe that what he wants to reserach is what should be researched. Shut down the Hansens to feed the Grays.
You write:
“One thing Gray wanted studied more was thermohaline circulation. To this day much of how it actually works remains a mystery. We have a general idea, but specific details remain unreserched. It apparently is not a steady flow, like a river or garden hose, but holds waves and/or pulses of some sort, which may-or-may-not influence the rate cold water upwells. Right now we have a good idea of how trade winds influence upwelling, but, if another factor also contributes to upwelling, then forecasts based only on trade winds will get messed up.”
Yes, let us fund this. This can produce genuine physical hypotheses. However, let us all recognize that all climate science is in its infancy and has not produced one physical hypotheses that goes beyond Arrhenius and that can be used to predict and explain climate phenomena. Forecasts of coming hurricane seasons illustrate clearly that more data must be collected and more phenomena must be understood if a science is going to emerge in the “nearish” future.