3
$\begingroup$

A formula was given by Brian Norcross but I wanted to know what others thought of this formula:

Hurricane Sustained Winds (mph) = 9.615 * (1015.8 - central pressure [mb])0.6143

New contributor
Jeremiah Leiden is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
$\endgroup$
6
  • $\begingroup$ Who is Brian Norcross & from where did you get the formula? $\endgroup$
    – Fred
    Aug 5 at 18:36
  • $\begingroup$ @Fred here in the US, Brian Norcross is a well known meteorologist from his local coverage of Hurricane Andrew (1992) and the Weather Channel. I realize that the name might be foreign to those not in the field or the US, but for those here in the US, saying his name would probably be considered sufficient :) $\endgroup$ Aug 5 at 18:56
  • 1
    $\begingroup$ @Jeremiah: what is the question? What we think of it is just discussion, there needs to be something you want to ascertain rather than just opinion :-) $\endgroup$ Aug 5 at 18:57
  • $\begingroup$ I was asking if the formula was sufficient or a good one to follow. I found the discussion here: wxinfinity.com/viewtopic.php?t=263 $\endgroup$ Aug 6 at 16:04
  • $\begingroup$ @njuffa Jeremiah did give a link to the forum he found it on in comments... the post there immediately references Norcross' book "Hurricane Almanac: The Essential Guide to Storms Past, Present, and Future". Doesn't seem easy to verify the ref (unless you happen to have a copy!) given it's now second and third hand to all of us. But seems fair enough to me for a user not to have a full scientific reference in their question, as they just came across it... $\endgroup$ Aug 7 at 10:23

1 Answer 1

3
$\begingroup$

Calculating the formula's average absolute error percentage with the best-track CSV archive for the Atlantic basin gave values of:

formula errors

Which shows both that it is somewhat reasonable, and that its skill is still fairly limited.

Such a simplistic formula would fail to account for surrounding pressures, storm structure, and things like extratropical transition. If I remember right, there is usually a time lag, as wind speed changes more quickly (particularly after landfall!) compared to pressure, as the storm and its warm core structure take time to respond to local changes. Also, storms that are subtropical or undergoing extratropical transition have an increased chance of having dynamics-caused wind maxima far removed from the MSLP center. Those reasons line up with the greater error in the columns for subtropical systems and tropical depressions (and it also looks like in skimming through the full results that there is probably larger errors later in a storm's life)

The applicability of this formula would be challenging... because we can't know the true minimum central pressure without direct observations anyways. You could use it when looking at models... but they're already just models, so it doesn't seem there's good reason to get precise with windspeed guesses. The models forecast wind as well. Model wind speed forecasts often won't line up with reality due to resolution and continuity constraints, but that's true of MSLP estimates as well.

And we already have solid, well-refined options for analyzing storms when direct observations (like NOAA aircraft with their scatterometers, aircraft radar, and dropwindsondes to much better investigate reality... or stations/buoys/land-based radars) are limited: the Dvorak Technique, which uses satellite presentation and trends to estimate the intensity. An Evaluation of Dvorak Technique–Based Tropical Cyclone Intensity Estimates by Knaff et al. shows Dvorak errors (by comparing to storms with recon data) continue to average less than 10 knots.

So in the end the formula is just a novelty exercise to try to find a way to hyper-simplify complex things (similar to this SE question on forecasting the weather), and simply verifies that lower local pressures do tend towards greater winds (as would make sense from the full Navier-Stokes equations, as they have the pressure gradient force terms, and lower pressures tend to mean more pressure gradient). But not sure it has extensive applications.

$\endgroup$
1
  • $\begingroup$ (Do note that in a given year there can be significant independence problems in calculating the estimate's errors, as each 6 hours is a separate data point, but significantly related to other data points on the storm... and even separate storms [due to regional factors like surrounding pressures]. And also that limited observation counts can give wilder inconsistency/statistically insignificant values in the error... see the ET storm column! That's why I put 10 years here, to get a more representative total... checked 50 years, it shows basically the same averages +/- a couple percent or so) $\endgroup$ Aug 7 at 10:09

Your Answer

Jeremiah Leiden is a new contributor. Be nice, and check out our Code of Conduct.

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.