Category 4 and 5 hurricanes

[MiamiensisWx]

I think the number of historical Category 4 and 5 hurricanes is underestimated. Look at the discrepancies between the Monthly Weather Review and HURDAT.

MWR:

1919 (Florida Keys/Texas) - 927 mb (ship)
1924 (Cuba) - 924 mb (peripheral pressure from ship on north coast of Cuba)
1932 (Cuba) - 915 mb (ship)
1933 (Storm 8) - 948 mb (ship east of NE Caribbean)
1933 (Storm 15) - 929 mb (ship east of Yucatan)
Inez (1966) - 927 mb (recon)

HURDAT:

1919 - 120 kt (140 mph)
1924 - 105 kt (120 mph)*
1932 - 115 kt (135 mph)
1933 (Storm 8) - 105 kt (120 mph)**
1933 (Storm 15) - 95 kt (110 mph)***
Inez (1966) - 130 kt (150 mph)

*This storm was an average sized TC from the deep tropics; Cat 3 is too low
**This peak was several days after the peripheral 948 mb measurement
***This one was a west-moving TC in the Caribbean; winds may be much higher (see Dean and Felix)

My estimates:

1919 - 140 kt (160 mph)
1924 - 140 kt (160 mph)
1932 - 140 kt (160 mph)
1933 (Storm 8) - 120 kt (140 mph) in the Atlantic
1933 (Storm 15) - 135 kt (155 mph) off Yucatan and major hurricane at both Mexican landfalls
Inez (1966) - 140 kt (160 mph) because of small size

[MiamiensisWx]

There are also several Category 5 storms that should probably be downgraded.

Fort Lauderdale (1947)
Dog (1950)
Easy (1951)
Cleo (1958)
Ethel and Donna (1960)
David (1979)

I need more data with respect to ambient pressures for David and Carla/Hattie (1961) in order to judge those storms. I know Crazy will disagree with me, but I also believe that Carrie 1957 is also overestimated. I would maintain its Category 4 status, but 135 kt (155 mph) seems too high.

[RL3AO]

I'm sure they are underestimated. Thats why the tropics cannot be judged and used to argue for or against man made global warming IMO. Reliable data only goes back 30-40 years for all basins except the Atlantic.

[MiamiensisWx]

I'm sure they are underestimated. Thats why the tropics cannot be judged and used to argue for or against man made global warming IMO. Reliable data only goes back 30-40 years for all basins except the Atlantic.

I think that's too generous even for the Atlantic, in my opinion. In reality, accurate intensity data only extends back to the early 1980s in the EPAC and WPAC. In the Atlantic, reliable intensities extend back to the late 1970s. All historical data for the Indian Ocean should be discounted as unreliable, and there have been several studies that suggest storms in the 1990s (well within modern technology) have been underestimated in the Indian Ocean. Look at Sidr 2007 as one example. It was likely a Category 5 storm, but it wasn't upgraded. The same case can be said for the South Pacific prior to the 1990s.

Revolutionary storms like Felix, Katrina (Cat 3 at 920 mb), Rita (Cat 3 with 930s pressure), and Erin (over OK) create further complications.

[KWT]

I habve few doubts that some of the storms are stronger then what is suggested from the past however at times you have to wonder about how much you can trust those ship reports from the past given they can sometimes not be well calibrated and lead to misguiding pressure reports.

I know off topic but I'd love to know how many storms have been missed in NE Atlantic, I'd bet there would be at least several dozen over the space of 100 years that have been missed, esp in late Autumn.

[MiamiensisWx]

I habve few doubts that some of the storms are stronger then what is suggested from the past however at times you have to wonder about how much you can trust those ship reports from the past given they can sometimes not be well calibrated and lead to misguiding pressure reports.

You can compare those ship reports with other background pressures and reports from adjacent land stations or vessels. That can offer a good idea of the tropical cyclone's size (i.e. outermost 1005 mb isotherm). Another indicator is the storm's formative region or track. That can offer clues about the upper-level pattern. Historical accounts can also offer suggestions about the possible RMW, forward motion of the storm, etc. All these data are crucial for reanalysis, and although nothing is definite, it can be utilized to document a storm's history. A reasonable intensity estimate can be calibrated based on our knowledge of more recent tropical cyclones like Dean, Felix, Hugo, Katrina, et al.

By the way, I think many storms have been missed in the deep Atlantic during the off season, including subtropical storms in the 1990s and previous years. We only started classification within the past few years.

Finally, ships may have reported a tropical cyclone or hurricane, but they were dismissed as non-tropical storms; thus they were not included in HURDAT.

[MiamiensisWx]

I know off topic but I'd love to know how many storms have been missed in NE Atlantic, I'd bet there would be at least several dozen over the space of 100 years that have been missed, esp in late Autumn.

I definitely agree; see above.

The current historical database is so frustrating; it is littered with so many inaccuracies, overestimated intensities, underestimated intensities, undercounted seasons, et al. So many current intensities contradict our latest knowledge. For example, thew 1935 Labor Day hurricane is listed as 140 kt (160 mph) in HURDAT, but it was clearly stronger based on the ambient pressures, small wind radii, and rapid intensification. We can only solve the climate change debate if we fix our historical database.

[Ptarmigan]

I thought of the same thing too and made a thread about it.

viewtopic.php?f=31&t=93773

I too believe some are underestimated and overestimated. I kow with Carla, a high pressure system caused it to move over Port O'Connor. If it was not for that high pressure, it would of ended up over Galveston.

As for Hurricane Dog, I believe it was a 5 because it was over the same area as Isabel in 2003. I think it had winds of no more than 165 mph at most, most likely 160 mph. Like I said before, if we had satellite, it probably would resemble Isabel, an annular hurricane because it maintained being a major hurricane for sometime.

I think the 1886 Indianola Hurricane was a Category 5 at some point because it made landfall at 925 millibars, one of the most intense to make landfall on record and in Texas. It was described as a large hurricane too. Other hurricanes I think are Category 5 are:
Great Hurricane of 1780
Racer's Storm of 1837
Great Havana Hurricane of 1846 (902 millibars)
1893 Cheniere Caminada Hurricane
San Ciricao Hurricane of 1899

I also believe that many storms did go unrecorded. We recently started to name and track subtropical systems. It would not surprise me if they went undetected. In fact, I think 1886, 1887, 1893, 1933, and 1936 were more active, especially with 1933, which I think was more active than 2005. I think 2005 type seasons are not as rare as many like to think.

[Ptarmigan]

I definitely agree; see above.

The current historical database is so frustrating; it is littered with so many inaccuracies, overestimated intensities, underestimated intensities, undercounted seasons, et al. So many current intensities contradict our latest knowledge. For example, thew 1935 Labor Day hurricane is listed as 140 kt (160 mph) in HURDAT, but it was clearly stronger based on the ambient pressures, small wind radii, and rapid intensification. We can only solve the climate change debate if we fix our historical database.

The 1935 Labor Day Hurricane was small. It had hurricane force winds extending up to 15 miles from the eye. Also, it may have had a pressure as low as 880 millibars. I think the wind was actually 200 mph in that storm. People were sandblasted to death.

[MiamiensisWx]

I think the 1886 Indianola Hurricane was a Category 5 at some point because it made landfall at 925 millibars, one of the most intense to make landfall on record and in Texas. It was described as a large hurricane too. Other hurricanes I think are Category 5 are:
Great Hurricane of 1780
Racer's Storm of 1837
Great Havana Hurricane of 1846 (902 millibars)
1893 Cheniere Caminada Hurricane
San Ciricao Hurricane of 1899

I also believe that many storms did go unrecorded. We recently started to name and track subtropical systems. It would not surprise me if they went undetected. In fact, I think 1886, 1887, 1893, 1933, and 1936 were more active, especially with 1933, which I think was more active than 2005. I think 2005 type seasons are not as rare as many like to think.

A recent study has revealed that the 902 mb reading in Havana was incorrect. It was actually 940 mb. I think it is difficult to estimate the 1846 hurricane's intensity, but I believe it may have been a strong Category 4 or 5 hurricane in the Caribbean. It was also likely a major hurricane west of the Florida Keys.

http://www.aoml.noaa.gov/hrd/Landsea/Partagas/impacthurrhist.pdf

I think 200 mph is unrealistic for the 1935 hurricane's sustained winds, but 165 kt (190 mph) is not out of the question. My best estimate (160 kt/185 mph) is supported by a recent study. The unofficial 880 mb reading should be discarded, as it was taken several hours prior to the 1935 hurricane's landfall.

[Ptarmigan]

A recent study has revealed that the 902 mb reading in Havana was incorrect. It was actually 940 mb. I think it is difficult to estimate the 1846 hurricane's intensity, but I believe it may have been a strong Category 4 or 5 hurricane in the Caribbean. It was also likely a major hurricane west of the Florida Keys.

http://www.aoml.noaa.gov/hrd/Landsea/Partagas/impacthurrhist.pdf

I think 200 mph is unrealistic for the 1935 hurricane's sustained winds, but 165 kt (190 mph) is not out of the question. My best estimate (160 kt/185 mph) is supported by a recent study.

I think 902 millibars may have been when it was over the Caribbean. Then again the pressure over Cuba was due to the barometer instrument being out of calibration. It appeared to form in the same area where Wilma did in 2005. As for the Labor Day Hurricane, I wonder how they were able to get the size of the hurricane. The eye was 8 miles in diameter from what I have read.

Speaking of that, I remember Super Typhoon Marge being recorded with winds of 215 mph and 888 millibars, which is overestimated. 888 millbars for typhoons would be around 155 to 165 mph due to lower ambient pressure or around 160 mph with Dvorak. Gilbert had 888 millibars and 185 mph winds and it was a monster.

[CrazyC83]

There are also several Category 5 storms that should probably be downgraded.

Fort Lauderdale (1947)
Dog (1950)
Easy (1951)
Cleo (1958)
Ethel and Donna (1960)
David (1979)

I need more data with respect to ambient pressures for David and Carla/Hattie (1961) in order to judge those storms. I know Crazy will disagree with me, but I also believe that Carrie 1957 is also overestimated. I would maintain its Category 4 status, but 135 kt (155 mph) seems too high.

David IMO was correct, I think the pressure was lower than estimated (I think that it was more like 908mb) - also David's pressure at the US landfall was underestimated based on surface reports (I think it was more like 955mb although its Cat 2 intensity is correct I think). Donna was probably also Cat 5 at one point, and I believe Dog had the 160 kt measured by recon (although that might have been flight level but at higher level so 150 kt is my estimate). As for Dog's pressure, probably around 915mb in Isabel range.

Ethel was certainly no Cat 5 (I think Cat 3 at the most), and Cleo and Easy may or may not have been. (Easy probably was as it is listed in the BT as Cat 5 for 18 hours, so there had to be consistency there, unless that was flight-level winds)

Other storms I believe were Cat 5 at one point:

1886 Indianola Hurricane (I can't see how a storm, even a small one, could be strengthening to 925mb on the NW Gulf Coast)
1910 Hurricane #3 (listed at 924mb/130 kt and that was after landfall)
1919 Atlantic-Gulf Hurricane (I think it was one of the low ambient, large gradient storms like Katrina and Rita, and its pressures were unusually low for its intensity - I think it peaked sub-900mb)
1924 Hurricane #7 (based on a later report)
1932 Hurricane #10 (also based on a later report, confirmed by ship report)
1957 Carrie (I think she inched up to 140 kt in that 24 hour period at 135)
1961 Esther (Unusually low pressure for so long)
1999 Floyd (Dropsonde data supported Cat 5)

My guess for the peak pressure and winds for those:

1886 Indianola - 914mb/145 kt
1910 #3 - 916mb/145 kt
1919 Atlantic-Gulf - 894mb/155 kt
1924 #7 - 921mb/140 kt
1932 #10 - 913mb/145 kt
1957 Carrie - 925mb/140 kt
1961 Esther - 922mb/140 kt
1999 Floyd - 921mb/145 kt

[CrazyC83]

A recent study has revealed that the 902 mb reading in Havana was incorrect. It was actually 940 mb. I think it is difficult to estimate the 1846 hurricane's intensity, but I believe it may have been a strong Category 4 or 5 hurricane in the Caribbean. It was also likely a major hurricane west of the Florida Keys.

http://www.aoml.noaa.gov/hrd/Landsea/Partagas/impacthurrhist.pdf

I think 200 mph is unrealistic for the 1935 hurricane's sustained winds, but 165 kt (190 mph) is not out of the question. My best estimate (160 kt/185 mph) is supported by a recent study.

I think 902 millibars may have been when it was over the Caribbean. Then again the pressure over Cuba was due to the barometer instrument being out of calibration. It appeared to form in the same area where Wilma did in 2005. As for the Labor Day Hurricane, I wonder how they were able to get the size of the hurricane. The eye was 8 miles in diameter from what I have read.

Speaking of that, I remember Super Typhoon Marge being recorded with winds of 215 mph and 888 millibars, which is overestimated. 888 millbars for typhoons would be around 155 to 165 mph due to lower ambient pressure or around 160 mph with Dvorak. Gilbert had 888 millibars and 185 mph winds and it was a monster.

I think Gilbert was a bit under 888 actually, since the best track shows 888 after rapid intensification then 889 six hours later when he started weakening. I think 885mb is the best estimate for Gilbert.

As for the 1846 hurricane, there is very little data to prove any intensity, but I think the 902mb was when it hit the Florida Keys, not Havana (where 917mb was disproven and 940mb is correct). I think that makes sense as it could have easily gone into rapid deepening as it entered the western Florida Straits. 917mb measured (probably lower in reality) as a storm is emerging off of Cuba after several hours makes little sense unless it was one incredibly strong storm; it was probably around 922mb when it made landfall in Cuba, weakened to about 935mb (since the minimum pressure was probably not recorded) then deepened to 902mb in the 24 hours afterward. That is just my guess.

[HurricaneBill]

I think somebody said for every storm that was probably overestimated, there was likely one that was underestimated.

One storm I'm curious about is the 1928 San Felipe/Lake Okeechobee Hurricane. Sustained winds of Category 5 strength were actually RECORDED as it moved over Puerto Rico. Supposedly the recorded winds were a bit far from the center. I read that it was a very large hurricane and the diameter of hurricane force winds was more than 200 miles.

[CrazyC83]

I think somebody said for every storm that was probably overestimated, there was likely one that was underestimated.

One storm I'm curious about is the 1928 San Felipe/Lake Okeechobee Hurricane. Sustained winds of Category 5 strength were actually RECORDED as it moved over Puerto Rico. Supposedly the recorded winds were a bit far from the center. I read that it was a very large hurricane and the diameter of hurricane force winds was more than 200 miles.

That probably explains why it was so catastrophic...that had to have had a really low pressure at peak intensity...probably in the 900-910 range...

[wxman57]

Let's take a look at all the recorded Cat 5s from 1851 through 1950. In 100 years, only 6 Cat 5 hurricanes were observed in the entire Atlantic Basin:

http://myweb.cableone.net/nolasue/Cat5early.gif

Now let's take a look at the period from 1951-2006, about half the time:

http://myweb.cableone.net/nolasue/Cat5now.gif

From a quick comparison, I'd say we're seeing about 6 times as many Cat 5s now as we were 50 to 150 years ago. Clearly, there's a trend there, right? Nope, we cannot make any such conclusion. Prior to the era of modern recon, the only direct measurements of central pressure and surface wind came from the passage of a hurricane over land or over a very unlucky ship at sea. Even today, new equipment is allowing for more accurate measurement of surface winds, like the SFMR. Without SFMR, it's possible that Dean or Felix might have not been classified as Cat 5s.

The same goes for all tracks at sea for which there were no direct measurements in the past - we just don't know how strong the hurricanes were for sure, or how long they maintained peak intensity. Yes, there were ship reports, but ships generally steer AWAY from approaching hurricanes, and measurements from a hurricane's perimeter aren't an accurate way to estimate central pressure and surface wind at the core.

I saw a talk at the AMS meeting in New Orleans 2 weeks ago by Greg Holland. He actually claimed that the HURDAT (hurricane database) was accurate to within 2.5 m/s (5 kts)! I was sitting next to Chris Landsea at the time and asked him about it. He just rolled his eyes and said no way. Perhaps using the most modern technology today we can get CLOSE to measuring the current intensity of a hurricane within 2.5 m/s, but without recon our estimates are probably quite a bit farther off. So how could estimates of hurricanes well out to sea from early last century be accurate to within 2.5 m/s? They're not. The tracks are just best guesses based on a couple of ship reports well away from the core for the most part.

[Derek Ortt]

In most basins, HURDAT only uses the Dvorak technique (and for the East Atlantic)

Dvorak numbers for Wilma were 6.0/6.0 6.5/6.5 at its record intensity. HURDAT may be accurate to within 50KT, but most certainly not 5

[wxman57]

In most basins, HURDAT only uses the Dvorak technique (and for the East Atlantic)

Dvorak numbers for Wilma were 6.0/6.0 6.5/6.5 at its record intensity. HURDAT may be accurate to within 50KT, but most certainly not 5

Yeah, I almost had to reach over and hold Chris Landsea down while Greg Holland was speaking. Greg Holland said the HURDAT database was accurate to within 5 kts, so there was no question that his trends showing increasing major hurricane activity in recent decades was real. Chris sounded like he was going to explode.

[Ptarmigan]

I think Gilbert was a bit under 888 actually, since the best track shows 888 after rapid intensification then 889 six hours later when he started weakening. I think 885mb is the best estimate for Gilbert.

As for the 1846 hurricane, there is very little data to prove any intensity, but I think the 902mb was when it hit the Florida Keys, not Havana (where 917mb was disproven and 940mb is correct). I think that makes sense as it could have easily gone into rapid deepening as it entered the western Florida Straits. 917mb measured (probably lower in reality) as a storm is emerging off of Cuba after several hours makes little sense unless it was one incredibly strong storm; it was probably around 922mb when it made landfall in Cuba, weakened to about 935mb (since the minimum pressure was probably not recorded) then deepened to 902mb in the 24 hours afterward. That is just my guess.

I read that recon measured Gilbert as low as 882 millibars. I am not surprised that Gilbert had lower pressure. It could be lower than Wilma.

[Ptarmigan]

In most basins, HURDAT only uses the Dvorak technique (and for the East Atlantic)

Dvorak numbers for Wilma were 6.0/6.0 6.5/6.5 at its record intensity. HURDAT may be accurate to within 50KT, but most certainly not 5

So much for accuracy. Wilma having a Dvroak number of 6.0, well gees, it shows it does not take into account of pinhole eyes. That same thing nearly killed the Hurricane Hunters in 1989 with Hugo when it was listed as a 3, which was a 5.