Comments on the proposed revisions to the 1930 Hurricane Season
From the Best Track Supervisory Committee
[Replies to the comments are indented and in bold face. CWL - September 2010]
General comment: The caption on the track map has thee cyclone #2’s. Please correct this.
[Corrected.]
Storm #1, 1930:
1. The new proposed HURDAT starts the storm 10 kt stronger than the original? What is the basis for this? The ship report the next day? If so, please state that. In light of this change, it might be desirable to extend the track 6-12 hours earlier if possible.
[The 50 kt/1006 mb observations at 05Z on the 22nd is indeed the reason for beginning the system 10 kt stronger than originally. This is now so stated in the metadata. However, we'd prefer not to begin the system earlier as very little data is available on the 21st (or 20th). Starting the cyclone, say, at 00Z on the 21st as a 30 kt depression would be making the assumption that it actually began on that date, when in reality it could just as easily have formed on the 20th (or 19th or earlier) over the void of the eastern tropical Atlantic. The metadata now clearly states that the exact date of genesis is uncertain.]
2. The proposed new track on 22-24 August is more jumpy, with a notable slowing down followed by a speeding up. Does the data justify this? If not, the committee recommends smoothing out this part of the track.
[Agreed to smooth the track.]
3. The proposed 105 kt peak intensity on 25-26 August looks reasonable in light of the data. However, since the 960 mb pressure ob was accompanied by hurricane force winds, the actual central pressure may have been in the lower 950’s. Might the data justify a slightly higher intensity?
[Agreed. Winds are boosted to 110 kt for a peak intensity. Certainly, the system could have been a Category 4 hurricane. Without additional evidence, the intensity late on the 25th and early on the 26th is at the high end of a Category 3.]
4. The committee has issues with the extratropical transition on 27 August, as it is unclimatologically far south and may be 12-24 hours too soon. Please re-examine the data to better refine the time of ET, and to see if a second baroclinic low was present near the hurricane on 27 August.
[It is agree that the original HURDAT of extratropical transition by 12Z on the 27th is too early. The system is clearly extratropical by 12Z on the 28th. Unfortunately, the density of observations is too sparse for a 00Z 28th analysis. It is estimated that extratropical transition occurred around 06Z 28th, around 41N. There does not appear that a second baroclinic low was present in the vicinity of the hurricane on the 27th.]
5. The 12Z HURDAT positions for 31 August and 1 September do not match the “extratropical low was centered at” position in the daily metadata descriptions. Please correct or clarify this.
[These statements have been removed from the daily writeups and the relevant description of significant changes have been included in the summary paragraph.]
Storm #2, 1930:
1. The committee feels that the evidence for starting this system as early as 25 August is too thin. Instead, it would probably be better to start the system on 29 August, when observations show westerly winds to the south of the likely cyclone.
[Agreed.]
2. In the daily descriptions, there are normally references to “available observations”. On 27-28 August, the maps suggest there are *no* available observations. You might want to come up with an alternate term for those situations.
[These statements have been removed from the daily writeups and the relevant description of significant changes have been included in the summary paragraph.]
3. The two westerly wind reports on 29 August are near 11N. Given the small size of the circulation seen later, this suggests it might have south of the proposed track at that time. Please consider revising the track southward on 29-31 August. Is there a time series of data from Barbados and St. Lucia that may help determine how close the center passed to those islands?
[Agree to move the cyclone farther south by a degree and a half on the 29th and adjusting the track southward from the 29th through the 31st. Unfortunately, no additional observations for these islands were available. (Mike Chenoweth did collect Barbados observations, but there was a gap between 1921 and 1942 in his compilation.)]
4. In the metadata summary, the RMW at Santo Domingo is stated to be about 10 n mi. The data from the steamship Coamo suggests that the center took roughly 90 minutes to cross the ship from RMW to RMW (150 mph estimated at 12 PM and 1:25 PM). The proposed track suggests the storm was moving 8-9 kt, which in turn suggests the maximum wind ring was 12-13 n mi in diameter. This would yield a RMW of 6-7 n mi. Is there information to suggest the ship did not sample the full diameter of the eye, or that there is something wrong with the peak wind times or the track speed? If not, the committee suggests re-evaluating the intensity using the smaller RMW.
[Agreed to use the 6-7 nm RMW at landfall in Santo Domingo. Winds at landfall are estimated to be 135 kt (5 kt higher for the 18Z 3rd value), making this a high end Category 4 hurricane in Dominican Republic. However, it is quite possible that this was a Category 5 hurricane at landfall. The metadata writeup now contains this discussion.]
5. Given the detailed pressure record from the Coamo in the Monthly Weather Review, is it possible to calculate the pressure gradient and the winds?
[Possibly, but given that we already have a central pressure and a fairly accurate representation of the RMW of the hurricane, this information will give a more reliable estimate of the maximum wind than a pressure time series at a single point.]
6. The committee concurs with the southward shift in the track on 4-5 September. Are there data from Port-au-Prince and/or the Cayman Islands that would help further refine the track?
[The only Port-au-Prince observations available are the 12Z data shown on the Historical Weather Maps. For some reason, the Cayman Islands were not providing observations before and during this hurricane.]
7. Is the data over the Gulf of Mexico on 7-8 September dense enough and conclusive enough to justify the downgrade to a tropical depression?
[Agreed that the available observations may not have been definitive on reducing the system to a tropical depression over the Gulf of Mexico. Peak winds were 20 kt N on the 7th and 30 kt W on the 8th. Original 35 kt from HURDAT retained for these dates.]
8. There are several comments and questions about the Florida landfall. First, the Tampa OMR records state that the center passed south of Tampa, while the track chart shows it over or north of Tampa. Please change this part of the track. Second, are there observations from Titusville, Daytona, or anywhere else on the Florida east coast that might define an exit point? Third, it appears likely that the 1006 mb pressure in Tampa was not a central pressure. This opens the possibility that the cyclone was stronger than 35 kt at its Florida landfall, albeit not too much stronger given the seeming lack of impacts.
[It is agreed to move the landfall point to south of Tampa as a 40 kt tropical storm. Unfortunately, the only U.S. Weather Bureau sites with anemometers and barometers for Florida in 1930 were Apalachicola, Jacksonville, Key West, Miami, Pensacola, and Tampa.]
9. Please check the longitude of the ob from the steamship Magmeric in the 12 September metadata.
[Corrected.]
10. Please review the proposed changes in intensity on 14-15 September. Is the available data close enough to the center to justify the proposed decrease?
[Agreed that there is not enough data to justify a decrease in the intensity on those dates.]
11. The HWM have some really awful frontal analyses in 1930, as shown by the analyses as this storm crossed Florida. It is suggested that you do your own frontal analyses instead of relying on the HWM in this year. This may not be as big of an issue for this system as for system #3 and Additional Notes #1.
[Agreed to provide this for other systems as needed.]
12. Please explicitly state the final fate of this system on and after 17 September in the metadata summary – extratropical, dissipated, or absorbed by another system.
[The system dissipated after 18Z on the 17th.]
Storm #3, 1930:
1. The committee has mixed feelings about adding this system. Even when the central winds and temperatures suggest it may have tropical characteristics, the cyclone was embedded in a baroclinic zone with a continuous stream of cooler air coming from the eastern United States. In addition, during extratropical transition the system seems to merge with the same frontal system that spawned it. This suggests the possibility that the cyclone never fully lost its frontal character. However, one committee member notes that the data does not confirm that the system was extratropical on 1-4 October and therefore a longer tropical or subtropical phase may be needed. Please give the data on this system another going over to better make the case that it was subtropical or tropical.
[The data have now been replotted on a separate map, which does not have the perhaps misleading frontal analyses from the Historical Weather Maps. The proposed life cycle - beginning as an extratropical cyclone, evolving to a tropical storm for about two days, then continuing as an extratropical cyclone for several more days – is maintained.]
2. Please re-examine the position at 1200 UTC 2 October, as an ob at 28.3N 72.4W suggests the center was northwest of the proposed position.
[Examination of this ship’s pressure time series suggests that it was 3-5 mb too low. Position of the cyclone is based more on another ship at 12Z that also had a NE wind with 1000 mb. The bias of the former ship is now noted in the daily writeup.]
3. There are places in the daily metadata (2, 3, 14, 15 October) where the position in the “Available observations” statement does not match the 12Z HURDAT position. Please correct this.
[These statements have been removed from the daily write-ups.]
Storm #4, 1930:
1. The committee concurs with the decision to add this system. Please check to see if the plot on the track map is correct, as the 1200 UTC position for 18 October seems to be missing.
[The missing 12Z 18 October label was due to the clunky plotting package we were originally using. The new map provided by Joan is much better.]
2. Since the 997 mb ship ob on 19 October was not a central pressure measurement, this suggests that the central pressure was possibly in the 990-995 mb range. Is it possible that the maximum winds were a little higher than 50 kt?
[Agreed to adjust the maximum intensity of this system to 60 kt on the 19th.]
3. The committee questions the rather quick demise of the system. Is it possible it drifted southward toward the coast of Mexico as pressures rose to the north? Please re-examine the data to see if this was the case and if anything was left on 21 October.
[Data is strongly suggestive that the system was completely dissipated by 12Z on the 21st as moderate northeasterlies cover the Gulf and the coastal Texas and Mexico stations show pressure rises of several millibars from the 20th. However, the data is somewhat sparse on the 21st in the southwestern Gulf of Mexico. The system is extended an additional 12 hours, with decay now occurring after 06Z on the 21st.]
4. A complication is that the October 1930 Monthly Weather Review (page 433) mentions a tropical cyclone in the eastern Pacific that died out over the Bay of Campeche on 20 October. Is the data good enough to tell if the Gulf system actually originated in the eastern Pacific or instead was a separate development (as currently proposed)?
[The Northeast Pacific tropical cyclone was likely a hurricane – 983 mb reported with a calm, two other reports of 50 kt max winds – that made landfall around 12Z on the 18th near 19.5N 105.0W. The newly added tropical cyclone in the Gulf of Mexico had its genesis at about the same time, 600 nm to the east. It is highly unlikely that the Gulf system originated from the strong tropical cyclone in the Northeast Pacific.]
Additional Notes:
System #1: See the comment above about the quality of the HWM frontal analyses. It is likely this system was extratropical and should be left out of HURDAT. However, the data should be re-checked in case this was a tropical cyclone in close proximity to a front, especially since an extratropical transition at 34N 75W in late August would be unclimatological.
[All available observations have been utilized for 12 hourly analyses of this system. While having a cyclone undergo extratropical transition near 35N 75W in late August is unusual, the observations and analyses do support that this did occur.]
System #2: Are any data available from St. Lucia or the other islands closer to the time of the aircraft report? The data presented does not conclusively rule out that a small tropical storm passed near St. Lucia or St. Vincent after the 8 AM St. Lucia ob, although a “100 miles an hour” system appears unlikely. Please provide more detail on other obs from this area and some charts for the binder.
[Unfortunately, no additional observations are available for this system. It is agree that the data presented does not preclude that a small tropical storm passed near St. Lucia or St. Vincent on the 6th. The Historical Weather Maps for the 7th and 8th are now included in the binder.]
System #3: The committee concurs with leaving this system out of HURDAT.
Other comments: The Monthly Weather Review indicates two systems that should be investigated to see if they should be included in HURDAT, or for possible inclusion in the Additional notes section.
1. The May Monthly Weather Review mentions a low that crossed the Gulf of Mexico 29-30 May and affected Florida. The HWM show a non-frontal low over the western Gulf of 29 May and a 1000 mb low merging with a front over the eastern Gulf on 30 May. There is one reported gale wind on 29 May in the Gulf. This system should be investigated to see if it was a subtropical or tropical cyclone prior to merging with the front, and how good the frontal analysis was.
[The Historical Weather Maps, Monthly Weather Review and COADS ships data show that a cyclone formed in the central Gulf of Mexico late on the 29th of May, moved northeastward and making landfall in Florida on the 30th, and merged with a strong baroclinic low on the 31st east of the U.S. mid-Atlantic states. While the system was organizing late on the 29th, a frontal boundary had moved into the Gulf of Mexico. It appears that the system was extratropical throughout its short lifetime. Highest winds observed were 30 kt from a ship and lowest pressure was 1001 mb, both on the 30th (though the environmental pressures were also quite low). (Note that the HWM chart wind barbs are plotted in mph, not kt. Thus the 3 ½ barbs are 35 mph/30 kt, not 40 mph/35 kt. This is confirmed in COADS, as this ship’s observations are available in there as well.) No Florida U.S. Weather Bureau land stations observed tropical storm force winds. Given that the system did not have tropical storm force winds and that it was extratropical, the cyclone will not be added into HURDAT.]
2. The June Monthly Weather Review tracks chart shows a low that meandered across the Florida Peninsula on 14-16 June. It is likely that this was a tropical depression or less, but it should be examined just in case.
[The Monthly Weather Review and COADS indicate that tropical depression formed in the eastern Gulf of Mexico on the 14th of June, made landfall over South Florida on the same day, and meandered northward over the Florida peninsula during the next two days. No gale force winds were reported from ships in the COADS data set, nor did any U.S. Weather Bureau stations report any tropical storm force winds. Lowest pressure reported was 1010 mb on the 14th. It is possible that this system reached tropical storm intensity, but with no observed winds higher than 25 kt, the cyclone will not be added into HURDAT.]
One additional comment was received from the NHC Best Track Change Committee in December 2010:
We've finished deliberating on the proposed new storm #3 of 1930. The committee vote (not counting mine), was two votes in favor of including the system in HURDAT as a tropical or subtropical storm, two votes in favor of not calling it a tropical or subtropical storm, and one abstention. After examining the data, I think the evidence that the cyclone shed its frontal character is not conclusive enough, and based on this the committee recommends that this system be left out of HURDAT.
This system may be more along the lines of a frontal hybrid - developing an inner wind core and/or central convection while still retaining some frontal character. Whatever the case, it developed enough tropical characteristics that the discussion that you've put together should be included in the possible other systems section for 1930.
[Agreed. The system is included in the Additional Notes of the 1930 metadata
writeup.]