Warm Atlantic & Trop. Cyclone Landfall Tendencies

[donsutherland1]

So far, it appears increasingly likely that the April-June period will see the Western Hemisphere Warm Pool (WHWP) come in at >+1.00 and the Tropical North Atlantic Oscillation (TNA) > 0. This is a particularly ominous signal. The following data is taken from seasons that met these criteria (1950-2004):

• Total storms: 61
• Major hurricanes: 16 (26%)
• Landfalling storms: 23 (38%)
• GOM landfalling storms: 20 (87% of landfalling storms)
• East Coast landfalling storms: 3 (13% of landfalling storms)

Even if one relaxed the WHWP criterion to > 0.000, here's the data:

• Total storms: 172
• Major hurricanes: 44 (26%)
• Landfalling storms: 57 (33%)
• GOM landfalling storms: 44 (77% of landfalling storms)
• East Coast landfalling storms: 13 (23% of landfalling storms)

Finally, 6/15 (40%) of the seasons that met the wider criteria saw at least one major hurricane make U.S. landfall. The major hurricanes that made U.S. landfall during those seasons were as follows:

• Donna (1960)
• Camille (1969)
• Allen (1980)
• Alicia (1983)
• Opal (1995)
• Charley (2004)
• Ivan (2004)
• Jeanne (2004)

With the exception of Jeanne, all of the above major hurricanes made U.S. landfall in the Gulf of Mexico region.

[ticka1]

Don - very interesting analysis. Why does the 87% and 77% keep jumping out at me for the GOM. I was hoping we would have another non-busy year here for the residents of Texas. Your analysis just shows that everyone needs to be prepared.

[jfranklin]

Don,

It is a little hard to gauge the significance of these results, because you haven't compared the number of landfalls in years that meet your criteria with the number of landfalls years that do not. This, I think, would be the most relevant comparison, no?

James

[donsutherland1]

James,

Here's the data for the other combinations:

WHWP>0; TNA<0:
• Total storms: 67
• Major hurricanes: 11 (16%)
• Landfalling storms: 28 (42%)
• GOM landfalling storms: 20 (71% of landfalling storms)
• East Coast landfalling storms: 8 (29% of landfalling storms)

WHWP<0; TNA>0:
• Total storms: 144
• Major hurricanes: 49 (34%)
• Landfalling storms: 41 (28%)
• GOM landfalling storms: 28 (68% of landfalling storms)
• East Coast landfalling storms: 13 (32% of landfalling storms)

WHWP<0; TNA<0:
• Total storms: 186
• Major hurricanes: 39 (21%)
• Landfalling storms: 59 (32%)
• GOM landfalling storms: 37 (63% of landfalling storms)
• East Coast landfalling storms: 22 (37% of landfalling storms)

[donsutherland1]

Ticka,

I very much hope that the U.S. can escape landfalling storms. Unfortunately, the reality is that such a risk exists. The National Hurricane Center has an outstanding Hurricane Preparedness page that can be found at:

http://www.nhc.noaa.gov/HAW2/english/intro.shtml

Now that Hurricane Season has begun, one should pay very close attention to what the NHC is saying. The U.S. is really privileged to have such an outstanding organization with among the world's top talent available to help keep its people safe from hurricanes/tropical storms.

[jfranklin]

James,

Here's the data for the other combinations:

WHWP>0; TNA<0:
• Total storms: 67
• Major hurricanes: 11 (16%)
• Landfalling storms: 28 (42%)
• GOM landfalling storms: 20 (71% of landfalling storms)
• East Coast landfalling storms: 8 (29% of landfalling storms)

WHWP<0; TNA>0:
• Total storms: 144
• Major hurricanes: 49 (34%)
• Landfalling storms: 41 (28%)
• GOM landfalling storms: 28 (68% of landfalling storms)
• East Coast landfalling storms: 13 (32% of landfalling storms)

WHWP<0; TNA<0:
• Total storms: 186
• Major hurricanes: 39 (21%)
• Landfalling storms: 59 (32%)
• GOM landfalling storms: 37 (63% of landfalling storms)
• East Coast landfalling storms: 22 (37% of landfalling storms)

Don,

OK, still need one more piece of information! How many years go into each of your three scenarios? That way we can calculate the average number of landfalls per season for each stratification. If they are different in a meaningful way, the next step would be to test the means for statistical significance.

James

[donsutherland1]

James,

The following are the number of years for each of the four scenarios:

WHWP>0; TNA>0: 15
WHWP>0; TNA<0: 8
WHWP<0; TNA>0: 13
WHWP<0; TNA<0: 19

Having said this, the sample sizes are small (WHWP & TNA data only go back to 1950) and the data is not statistically significant at a 95% level of confidence. Nonetheless, there did seem to be a somewhat higher proportion of storms making landfall in the Gulf States among the total that made U.S. landfall in seasons where the WHWP>0 and the TNA>0 and the standard deviation of ~0.2 was smaller than that for some of the other scenarios concerning the WHWP and TNA.

[jfranklin]

James,

Here's the data for the other combinations:

WHWP>0; TNA<0:
• Total storms: 67
• Major hurricanes: 11 (16%)
• Landfalling storms: 28 (42%)
• GOM landfalling storms: 20 (71% of landfalling storms)
• East Coast landfalling storms: 8 (29% of landfalling storms)

WHWP<0; TNA>0:
• Total storms: 144
• Major hurricanes: 49 (34%)
• Landfalling storms: 41 (28%)
• GOM landfalling storms: 28 (68% of landfalling storms)
• East Coast landfalling storms: 13 (32% of landfalling storms)

WHWP<0; TNA<0:
• Total storms: 186
• Major hurricanes: 39 (21%)
• Landfalling storms: 59 (32%)
• GOM landfalling storms: 37 (63% of landfalling storms)
• East Coast landfalling storms: 22 (37% of landfalling storms)

Don,

Thanks. So, doing the division yields the following average number of landfalls per year for the three above scenarios, respectively:

1) 2.5 lpy GOM, 3.5 overall
2) 2.2 GOM, 3.2 overall
3) 2.0 GOM, 3.1 overall

So it looks to me as though it really doesn't matter what the value of WHWP and TNA are, because you get 2-2.5 landfalls in the Gulf of Mexico, and 3-3.5 landfalls overall, regardless.

I think the bottom line is that there is no proven technique to predict the likelihood or (even more difficult) the locations of landfall at the beginning of the season. The only serious study I have seen that addresses this issue is the one recently published my Mark Saunders in Great Britain, who found some correlations between US landfalls and synoptic patterns AT THE END OF JULY! I've never seen anything to make me think we can say anything intelligent about landfalls at the beginning of June.

[donsutherland1]

James,

You wrote, "I think the bottom line is that there is no proven technique to predict the likelihood or (even more difficult) the locations of landfall at the beginning of the season." I agree with that statement. Until one can accurately forecast the synoptic pattern that will predominate, among other factors, such skill will likely remain elusive.

[mizzou]

is don southerland a professional meteorologist? this is better than the nws