• Sailing is all about the Weather.

    Big into the exploration of Atlantic Hurricanes since Katrina came uninvited into his world, James (Jim) Gurley (JamesG161) has followed every Tropical Storm birthed in Atlantic, Gulf and Caribbean waters since. Being a boater, he knows that we often need more time to prepare than we get from the TV weather folk. Jim relies on the science of storm development to share early warning info with friends and fellow boaters.

    Early in 2018, Jim and John Shepard, (JSSailem) started to chat about the weather data available. John asked Jim to help forecast Pacific NW storms, and this morphed into discussions on weather forecasting.

    For John, sailing in the PNW is sometimes hit and miss. One day is ugly, then a string of beautiful days but no wind, followed by a series of blue-sky days and 12 knot breezes. Being ready for those great sailing days means you need to look to the Pacific Ocean and what is brewing. John has been into Pacific NW Weather since the 1970’s when his first PNW November storm hit bringing more than 40 days and 40 nights of continual rain.

    Together we want to share information, new APPs, safety, and thoughts about letting the weather help you. Identify some of the resources for sailors and help prepare you for your next sailboat outing.

    It is far better to go out on the water knowing what to expect in weather terms, than to be out on the water and see dark ominous clouds suddenly appear, unprepared.

Hurricane Season 2023

D

dlochner

.
Jan 11, 2014
12,348
Sabre 362 113 Fair Haven, NY
shemandr said:
What does that pie chart mean? Seriously.
Click to expand
What's lacking is reference data. What's the mean and median number of Named Storms, Hurricanes, and Major Hurricanes? That would help to put the assessment in context.

One of the challenges of working with small sample sizes is the outsized effect of one or two storms. For discussion purposes, let's say that the mean number of Major Hurricanes is 2. If three storms hit it is a 50% increase over predictions, but it was only one storm.

So, @JamesG161 can you provide us with the reference data, specifically the mean and median for each of the categories?
 
BigEasy

BigEasy

.
Jun 21, 2004
2,662
Beneteau 343 Slidell, LA
Read a couple days ago that El Niño should be firmly in place by the time the most active part of hurricane season starts. Normally a strong El Niño reduces the number of storms or prevents them from becoming major hurricanes, so that is a good thing. BUT, this year, the Atlantic basin will have historically high water temps, so it’s yet to be seen how those two factors play out.
 
  • Like
  • Helpful
Likes: jssailem and JamesG161
JamesG161

JamesG161

SBO Weather and Forecasting Forum Jim & John
Feb 14, 2014
7,717
Hunter 430 Waveland, MS
Davidasailor26 said:
The press release leaves a lot to be desired.
Click to expand
I bet you got to that link via the Press Release;)

Their forecast is a compilation of a lot of input, exception may be from the Farmers Almanac.;)

Jim...
 
  • Ha
Likes: jssailem
D

dlochner

.
Jan 11, 2014
12,348
Sabre 362 113 Fair Haven, NY
JamesG161 said:
Ok, here is more information.

View attachment 216082

Do you know what ACE means?

Jim...
Click to expand
Thanks Jim, this is a much more informative image. The Atlantic is pretty much a crap shoot with equal chances of above, below, and normal. Pretty hard to be wrong on that prediction. The Eastern Pacific is much more interesting, with a 55% chance of above normal weather. If you're a gambler don't bet on the EP having a mild hurricane season. Comparing the averages to the predictions for the EP, it's clear they expect more hurricanes of greater intensity. On the other hand, the Atlantic predictions suggest the usual amount of named storms and hurricanes, but a lower likelihood of major storms. But it is early yet, too early to start betting. ;)

How did I interpret the predictions? Take a look at the predictions for Atlantic hurricanes, the average number (7) is dead center in the predicted range suggesting the range is probably accurate. However, the average number of major hurricanes (3) is at the higher end of the predicted range for major hurricanes (1-4) suggesting the model expects fewer major hurricanes.

The opposite seems to be the case for the Eastern Pacific. The average number of hurricanes (8) is at the low end of the predicted range (7-11) as is the average number of major hurricanes (4) when compared to the predicted range (4-8).

This analysis is done without any consideration of the metrology of hurricanes, it is simply a way of interpreting the predictions made by those who have run the prediction models. The basic concepts are the same for any prediction model.
 
JamesG161

JamesG161

SBO Weather and Forecasting Forum Jim & John
Feb 14, 2014
7,717
Hunter 430 Waveland, MS
dlochner said:
This analysis is done without any consideration of the metrology of hurricanes, it is simply a way of interpreting the predictions made by those who have run the prediction models. The basic concepts are the same for any prediction model. [:badbad:]
Click to expand
Ok Dave, John and I did this for 2022, please read the entire Post, since at the top, is the key for NOAA assumptions.

Hurricane season 2022 Analysis

If your read the first link in post #1, then you can compare 2022 and 2023.
Jim...

PS: @jssailem is starting a SBO group race right now, thus only monitoring our posts.
 
D

dlochner

.
Jan 11, 2014
12,348
Sabre 362 113 Fair Haven, NY
JamesG161 said:
Ok Dave, John and I did this for 2022, please read the entire Post, since at the top, is the key for NOAA assumptions.
Click to expand
I wasn't trying do anything with the assumptions made in the prediction models. Those who be the domain of researchers who are trying to improve the models they use.

The question I was answering, is how to make sense of the numbers produced by the prediction models because on the surface they don't make much sense, which is why I asked for the extra information. I came by this knowledge prediction models from 18 hours of doctoral level statistics and 30 years of applying models to predict academic achievement.

For those who are interested, all prediction models are derived from the linear regression formula we learned in basic algebra, y = mx + b. The models used by meteorologists and others are obviously more complex and for meteorology incorporate millions of data points which require supercomputers to solve. All models have this basic form "Predicted Event = (a bunch of variables) + error term. These more complex prediction formulas are known as multi-variant regression formulas.

The error term is important, because it will tell you the range of possible answers. Going back to the Atlantic prediction, the range was 5-9 hurricanes. Since the error term is always symmetrical, i.e., expressed in a =/- form, the model likely predicted 7 hurricanes +/- 2. Thus NHC reports it will be a "Normal" season because 7 is the average number of hurricanes.

Looking at the Named Storms predictions, the range is 12-17. The midpoint of that range is 14.5, which slightly higher than the average 14.

Finally looking at the Major Storms, the predicted range is 1-4 and the historical average is 3. Since 3 is slightly higher than the midpoint (2.5) there is a slightly lower chance of a major storm.

If all three conditions were equally possible, above, below and average, they would all have a 33.3% probability. In this case the Normal gets the edge because the prediction and history align quite well, so it has a 40% chance. While the others get 30% chances.

So long as the data and model have been developed on sound scientific principles, the basic interpretation is the same, doesn't matter if you are predicting hurricanes, drought, academic achievement, corn yields, the principles are the same. It is all rooted in the linear regression formula we learned in algebra and the statistical methods were developed more than 100 years ago to predict crop yields on farms.
 
JamesG161

JamesG161

SBO Weather and Forecasting Forum Jim & John
Feb 14, 2014
7,717
Hunter 430 Waveland, MS
All the Tropical Season forecasts are predictions.

If you read our End of 2022 analysis, you should have seen their errors.

NOAA never admits they are wrong, but move on to 2023.
__________
All forecasts are Statistical and compared against a Median.

The worst in 2022, on ENSO and African Monsoons.
_____

The real issue is the Extrapolations to project landfall points. Their best was 48 hours ± 25 miles.

I also know Statistics.

But here we try to use a pure Science, not Statistics, to help fellow Sailboaters.

The ECMWF models are done for crops, etc.
GFS = NOAA

Jim...
 
  • Helpful
Likes: jssailem
You must log in or register to reply here.

Top Bottom