NHC releases 86-page report on Hurricane Michael -- officially upgraded to Cat 5

[Darrell KSR]

Only the 4th category 5 hurricane ever to hit the U.S.

Here's a link to the report.

https://www.nhc.noaa.gov/data/tcr/AL142018_Michael.pdf

[dan_bgblue]

The following is a very detailed description of what they think occurred. The last statement says this is our best guess at this point, but things could change based on some more obscure math models, and more massaging of suspect data already collected.

I am saying that they could be as accurate if they strapped Jim Cantore to a vertical I beam that was driven 20 feet into the beach sand, and let him hold a child's pinwheel that was attached to a large bicycle inner tube and held facing the wind for 10 seconds. The number of twists of the inner tube times the diameter of the pin wheel divided by the coefficient of friction between the sand in the air and the blades of the pin wheel would not only make Jim a very happy camper but it would also make for great theater on the weather channel. Would not tell them squat about the wind speed, but it would be a good exercise in absurd comedy.

[PedroDaGr8]

The following is a very detailed description of what they think occurred. The last statement says this is our best guess at this point, but things could change based on some more obscure math models, and more massaging of suspect data already collected.

I am saying that they could be as accurate if they strapped Jim Cantore to a vertical I beam that was driven 20 feet into the beach sand, and let him hold a child's pinwheel that was attached to a large bicycle inner tube and held facing the wind for 10 seconds. The number of twists of the inner tube times the diameter of the pin wheel divided by the coefficient of friction between the sand in the air and the blades of the pin wheel would not only make Jim a very happy camper but it would also make for great theater on the weather channel. Would not tell them squat about the wind speed, but it would be a good exercise in absurd comedy.

Your synopsis is 100% incorrect. By phrasing it "what they think happened", you imply they are just pulling it out of their ass, when you couldn't be any MORE wrong. Ignoring your hyperbolic statements about Jim Cantore, this analysis is QUITE sound. There was no "massaging" the data, instead they were revising an analysis based on a much larger data pool. Every decision on data sources were data-based analyses on why said source is more or less reliable. Furthermore, the previous real-time data provides a "minimum" for how bad things they are and they provide error bars for their revised analysis. At the end of the day, the refinement is VERY minor (from 135knt to 140knt) and the change from Cat 4 to Cat 5 is entirely due to the boundary. The difference between a 135knt storm and a 140knt storm is low.

For those that care about science, here's some snips of what the report ACTUALLY says and a more layman's translation:

While these observations are well below both the operational and final best track intensities, the observing sites were likely not optimally located to sample the maximum winds, which is typical during landfalling hurricanes.

Translation: In reviewing the data, they can see that they were not at the location to get the actual peak wind speed information, which is common for hurricanes that make landfall (they don't say why). As such, the real-time data shows the wind-speed and pressures as being LOWER than the true maximum. In the most basic speech: We know how bad it was where we were at and we know that some places were worse.

Michael’s estimated intensity at landfall in Florida is 140 kt. While the real-time operational estimate was 135 kt, the final best track intensity estimate was determined by a detailed post-storm analysis review of the available aircraft winds, surface winds, surface pressures, satellite intensity estimates, and Doppler radar velocities – including data and analyses that were not available in real time. It should be noted that the NHC best track intensities typically have an uncertainty of around ±10%

Translation: The real-time result of 135kt has already been shown by the real-time and non-real-time data to be lower than the maximum. As such, they used all of the data sources available, including ones which are not available real-time, to calculate an estimated surface wind speed at the location of the maximum. They know how bad it was away from the maximum, so any estimate is already lower bounded by the real-time data. They also provide the error bars for ANY best track estimate (±10%).

Parts 1-5 under Winds and Pressure are them going through each data source, what the data source says and why it is more or less accurate. These data sources are what they use to calculate the revised values.

From there, it then goes on to analyze how well the various forecasting models did, etc.

A verification of NHC official track forecasts for Michael is given in Table 5a. The average official track forecast errors were smaller than the mean official errors for the previous 5-yr period at all forecast times.

Translation: In evaluating how well the NHC forecast models did at predicting the path of the hurricane: the mean errors were smaller than usual, meaning the NHC models did a good job.

Average official intensity forecast errors were significantly larger than the mean official errors for the previous 5-yr period. Table 6b has a similar verification for the NHC Michael intensity forecasts including the forecasts issued before it became a tropical cyclone, and the forecast errors were again significantly large. A homogeneous comparison of the official intensity errors with selected guidance models is given in Table 6c. While the official forecasts were poor compared to the
5-yr mean, they were better than almost all of the intensity guidance forecasts, trailing only the HWRF model (HWFI) at one time, 48 h.

Translation: In evaluating how well the NHC forecast models did at predicting the intensity of the hurricane, the models did a BAD job compared to their historical averages. To get an idea of HOW bad, they compare it to a number of other models used to predict intensity. Based on that analysis, ALL of the models did badly, implying that there is something about the development of this hurricane not captured in any model. When comparing to the various models, only one model at one time point did better. As a scientist, this is the kind of thing that you live for; when there is something new and unusual when you don't expect it. It happens regularly and often shapes how the scientific viewpoint evolves. Keep in mind, it does not render the previous models and frameworks wrong; any new revised model has to predict the "normal" hurricanes just as well as the previous model but must do better on these unexplained cases.