[Atlantic Oceanographic and Meteorological Laboratory]

HRD Home
About AOML
About HRD
Data Sets
 yellow triangle bulletMission Catalog
 yellow triangle bulletSfc. Wind Anal.
 yellow triangle bulletSyn. Surveillance
 yellow triangle bulletRe-Anal. Proj.
 yellow triangle bulletBasin-Wide Data
 yellow triangle bulletData Formats
Weather Info
What's New

OAR Logo
NOAA Oceanic Atmospheric Research

NOAA Aircraft Operations Center Logo
NOAA Aircraft Operations Center

Site Map

Staff Data Center Contact  Information

Research Divisions

Hurricane Research Division


Comments on the proposed revisions to the 1931 Hurricane Season

From the Best Track Supervisory Committee


Replies are in bold face from Andrew Hagen and Chris Landsea – June 2011



General comments and questions:


1. When does the National Climatic Data Center start having records for surface data from the various other countries of the Atlantic TC basin?  Is any of it available prior to 1931?  The committee notes a Swan Island OMR for storms #2-3 and wonders how far back that record extends?


Within the last year, NCDC added additional raw observations from Mexico and Swan Island to their EDADS website.  Unfortunately, because of the loss of funding, this website is semi-permanently down so it is not possible to check exact dates at this time. 


2. The committee is concerned about inconsistency in the interpretation of 60 kt winds.  For example, the 60-kt observation for storm #7 is used to justify hurricane strength, while similar observations from storms #2 and #6 are not.  Please explain the difference in this application.


Storm 7 contains more evidence of hurricane intensity than both Storms 2 and 6.  In Storm 7, in addition to the 60 kt wind observation (with a 1001 mb pressure) we also have an observation of 45 kt simultaneous with 991 mb.  Assuming the central pressure is less than or equal to 987 mb, this yields a wind speed of at least 66 kt according to the Landsea et al. north of 35N pressure-wind relationship.  Having both the 991 mb pressure and the 60 kt wind is enough evidence that this system was a 65 kt hurricane.  For Storm 2, there is one single ship observation of 60 kt late on the 14th.  The lowest observed pressure from any ship for the entire storm was 1006 mb.  15 hours later, data indicates the cyclone made a U.S. landfall with a 50 kt intensity.  Although the HURDAT intensity is increased from 50 to 60 kt at 18Z the 14th – 00Z the 15th, it is difficult to say that there is enough data to make this a hurricane in HURDAT.  For Storm 6, it is possible that it could have been a 65 kt hurricane briefly before the landfall in Jamaica on 12 September.  However, without further evidence it is suggested to boost this system to a 60 kt tropical storm (up from 55 kt originally) right before landfall in Jamaica.


Storm #1, 1931 (June 24-28):


            1. Adding June 24 looks reasonable, as the Historical Weather Maps rather conclusively show a closed circulation on that day.  This should be explicitly noted in the metadata daily write-up and the summary.


Ok.  Added to metadata.


            2. The metadata summary states that the highest observed land station winds were 35 kt from Corpus Christi, and that this is the basis for the Texas landfall intensity of 40 kt.  When was this observed and what is the source of the data?  Please note here and in the appropriate daily section.  It is noted that the Monthly Weather Review Monthly Summary Table for June (page 255) shows the maximum wind for the month at Corpus Christi was 36 mph.  This seems to match the OMR in the binder.  If it turns out that the 35 kt winds are not correct, please re-examine the Texas landfall intensity.


The Monthly Weather Review Monthly Summary Table maximum wind values are the 5-min maximum values.  The Corpus Christi OMR lists both the 5-min maxima and the extreme velocity which means “fastest mile” observed.  The fastest mile at Corpus Christi was 40 mph (35 kt).  This is equivalent to about a wind average over a period of 1 minute and 40 seconds.  If a conversion factor of 1.04 is used to convert the 1 min 40 sec wind to a 1-min wind, 36.4 kt is obtained.  The anemometer was located at a height of 78 ft (24 m).  The 36.4 kt times a factor of .945 yields a 10m 1-min wind of ~34 kt.  The height of the station was obtained from column 4 of the MWR Table of maximum wind.  The Corpus Christi station listed in the MWR table is the same station as the Corpus Christi OMR because both had identical values of maximum 5-min wind.  Since 34 kt is yielded after converting to 10m and 1-min, it is recommended to make the landfall intensity 40 kt.  The wind occurred from the NE at 15Z on 27 June according to the OMR and the pressure at that exact time was 1010.5 mb and falling slowly.


The following new observations are added to the June 27th daily metadata paragraph as highlights:

35 kt NE (max w/1-min) with 1010 mb at Corpus Christi at 15Z (OMR)

20 kt E with 1009 mb (min p) at Corpus Christi at 23Z (OMR).


In the metadata summary, the sentence “The highest observed wind from a land station was a fastest mile wind of 35 kt at Corpus Christi” is changed to: “The highest observed wind from a land station was a fastest mile wind of 35 kt at Corpus Christi (34 kt after converting this observation to a 10m 1-min value).  This observation was obtained from the Corpus Christi Original Monthly Record of observations.”


Storm #2, 1931 (July 11-17):


1. As with storm #1, if the data appears to support a closed circulation on 10 July please state that in the daily section.


In July 10 daily metadata paragraph, the following sentence is added after the sentence “HURDAT did not previously analyze the system on this day” :

“Although available observations including a west wind at 12Z at Cabo Bracias indicate the possibility of a closed circulation on the 10th, the west wind may have more likely been due to a land breeze.”


            2. On a related note, the committee is not sure the data actually justifies adding 10 July. Can it be determined if the observation at Cabo Gracias was an actual synoptic-scale wind or a mesoscale (landbreeze?) effect?


After analyzing the 12Z observation from Cabo Gracias from 24 hours earlier and 24 hours later, it appears that the west wind at Cabo Gracias may indeed have been more due to a land breeze.  The original start in HURDAT at 06Z on the 11th is retained.


            3. Is the data coverage sufficient to justify the downgrade to a tropical depression before landfall in Yucatan?


Assuming that the newly analyzed position on 12 July at 12Z is correct, then the data coverage is sufficient to downgrade to a tropical depression.  From 12Z on the 11th to 12Z on the 12th, the pressure at Belize City rose by 0.7 mb.  The analyzed distance of the cyclone from Belize City at 12Z on the 11th is 320 nmi and at 12Z on the 12th is 80 nmi.  Another observation- a ship at 12Z on the 12th located approximately 80 nmi south of the analyzed center has a northeast wind, which means that is not part of any cyclonic circulation located to its north.  If the position is correct, this cyclone had to have been extremely weak, and it is recommended to be a 30 kt tropical depression at landfall in Yucatan at 17Z on the 12th.


The following sentence is added to the metadata summary after the sentence “Therefore, the cyclone is analyzed to…” :

“This downgrade is due to sufficient observational coverage, which indicates a weaker cyclone.”


            4. See general point #2.  The 60 kt observation on 14 July suggests the possibility that the system briefly reached hurricane strength.


Explanation from general point #2: For Storm 2, there is one single ship observation of 60 kt late on the 14th.  The lowest observed pressure from any ship for the entire storm was 1006 mb.  15 hours later, data indicates the cyclone made a U.S. landfall with a 50 kt intensity.  Although we increased the HURDAT intensity from 50 to 60 kt at 18Z the 14th – 00Z the 15th, it is difficult to say that there is enough data to make this a hurricane in HURDAT.


            4. There are several issues that require a re-analysis of the track after landfall in Louisiana:


                        4a. The Morgan City OMR:  It is asserted that the cyclone made it closest approach to the station at 20Z, some 7 hours after the minimum pressure.  However, the winds at the time of the minimum pressure were westerly, and remained westerly for several hours afterward according to the data on page 6 of the OMR.  This is inconsistent with the assertion that the 1001 mb pressure was 70 n mi north of the center position.


After additional analysis, it appears that the center passed very near Morgan City around 12Z on the 15th.  At 13Z on the 15th, Morgan City OMR reports 10 kt W with 1001 mb.  HWM on the 15th shows calm at Morgan City.  A central pressure of 1000 mb is analyzed from this data and added into HURDAT at 12Z on the 15th.  The following additional changes are recommended to be made to the HURDAT latitudes: 7/14 18Z: 26.0N, 7/15 00Z: 27.2N, 7/15 06Z: 28.4N, 7/15 12Z: 29.6N, 7/15 18Z: 30.4N, 7/16 00Z: 31.0N, 7/16 06Z: 31.6N, 7/16 12Z: 32.2N.  Landfall is now shown to have occurred at 7/15 10Z at 29.2N, 91.0W.  Based on the revised track, pressure data from Morgan City, and the timing of certain gale force observations from ships, the following additional changes are made to the intensities: 7/15 10Z landfall: 50 kt; 7/15 12Z: 45 kt; 7/15 18Z: 40 kt.  Since the central pressure at 12Z was 1000 mb, the central pressure at landfall was likely a few mb lower than that.  A 50 kt landfall intensity in the Brown et al. (2006) pressure-wind relationships for north of 25N corresponds to a 996 mb landfall central pressure for 10Z, which makes sense.


4b. Are there OMRs available from other stations in Louisiana that would help refine the track?  For example, both the original track and the proposed revision pass near Shreveport, which reported a northwesterly wind at 12Z on 16 July (no pressure shown on the map).  Is the OMR for that location available?


No other OMRs are available at this time.


4c. It is noted that Austin’s 40 mph winds on 15 July were from the northwest, while Houston had a maximum monthly velocity of 35 mph from the west on 15 July.  Also, the Monthly Weather Review chart for the morning observations (July issue, chart IX) shows northwest winds at Galveston.  Is it possible that after landfall this system turned more westward than shown in the current track (or in the Monthly Weather Review track), or were these winds associated with a separate system?  Could such a turn be made consistent with the wind directions from the Port Arthur OMR?  Is such a turn consistent with the observations in Shreveport on 16 July?


These winds are not associated with Storm #2.  The winds appear to be part of a separate system – not a tropical cyclone.


5. The summary section of the metadata tends to jump around chronologically and is confusing to follow.  Please re-write it for clarity.


It has been re-written.


            6. This storm seems to have had an unusual structure during and after landfall.  The Morgan City OMR does not mention strong winds, while the Severe Local Storms table in the Monthly Weather Review (pages 283-284) mentions significant rain and wind over southwestern Alabama (not strong enough to produce gales at Mobile).  It is also noted that Pensacola had a maximum monthly velocity of 38 mph from the south on 16 July.  Have the Mississippi. Alabama, and Florida OMRs been checked for additional information?  Is there a chance this system had a subtropical structure? 


The Mississippi, Alabama, and Florida OMRs have not been checked for additional information.  OMR website is now semi-permanently down.  The system does seem to have an atypical structure for a tropical cyclone and may have been subtropical.


Storm #3, 1931 (August 10-19):


            1. There are issues with the origin of this system on 10 August.  The first is that the pressure obs do not support the westerly winds at Port Castries, with lower pressures to the south.  Has this wind been checked to make sure it did not result from an orographic effect?  The second is that there are two ships farther south with lower pressures and SW-SSW winds, suggesting the initial development could have been south of St. Lucia.  If the obs are good enough to determine a closed circulation on 10 August, please mention this and the obs used to locate it in the daily section.


Available data indicates that the wind at Port Castries 24 hours earlier was southeasterly rather than southwesterly like on the 10th.  The pressure there also dropped by 1.3 mb from the 9th to the 10th.  It is agreed to move the position on the 10th farther southeast.  With the more southeastern solution, then it is better not to change the timing of genesis in HURDAT from the 18Z starting time because those two ship observations, which are much farther south, are not enough evidence of a fully closed circulation to warrant beginning the cyclone six hours earlier than originally. 


            2. On a related note, are OMRs or anything similar available for the Lesser Antilles for this system?  These might better reveal a small system than the daily snapshots that are currently available.  If the initial center is farther south than currently thought, it might be detectable in the Barbados obs, for example.


No additional observations are available from the Lesser Antilles.


            3. Daily section for 17 August: There are multiple ship reports near 18.5N 92.6W around 0500 UTC with wind of NW 35 kt, NW 45 kt, and S.  Are these all correct?  Note that a NW wind 45 kt could mean the system was stronger than currently analyzed as it crossed Yucatan.


It was found that there was a mistake in the ship highlights for August 17th.  In addition, the highlights were out of chronological order.  The corrected ship highlights for August 17th are as follows: “Ship highlights: 35 kt NW and 1007 mb at 05 UTC at 18.5N, 92.6W (MWR); 40 kt NW (no time) in the vicinity of 18.5N, 92.6W (MWR); 45 kt SW and 1002 mb at 18 UTC at 18.7N, 92.7W (COA); S wind on the 17th (no time) (MWR).”


A 35 kt NW wind was observed at 05Z.  A 40 kt NW wind was likely observed some hours after 05Z.  A 45 kt SW wind (with 1002 mb) was observed at 18Z.


Since the analyzed position of the cyclone at 06Z on the 17th is basically right on the beach about to emerge back over water after having been over land for a long time, an intensity increase from 30 kt (at 00Z) to 35 kt (at 06Z) is shown.  The analyzed intensity value at 06Z is equal to the highest observed wind at that time.  Given the distance from the analyzed cyclone center to the ship at the time (40 nmi), it is possible that could have been near the RMW at the time given that land likely disrupted the circulation which was likely to be poorly organized.  It is tough to show a sharp intensity increase before the system emerges over water unless there is data to do so.  The intensity is boosted to 50 kt by 12Z on the 17th in accordance with the other observations of 40 and 45 kt winds and the 1002 mb pressure.  Therefore, no additional changes were made to HURDAT.


            4. Daily sections for 17 and 18 August: There are ship reports of northwest 35 kt and 1007 mb near 18.6N 92.6W at 05Z on each of these days.  Is this correct?


This observation only occurred on 17 August.  The one on the 18th is incorrect and has been deleted.


            5. “A 35-kt intensity is analyzed at 06Z (down from 40 kt originally, although the highest observed wind at that time was 45 kt at 05Z” (the ship Morazan).  If there was a 45-kt ob one hour before the synoptic time, please justify why 35 kt was chosen for the synoptic time intensity.  As written, this makes no sense. (See point 3.)


The metadata summary is incorrect.  The above quote is changed to: “A 35-kt intensity is analyzed at 06Z (down from 40 kt originally).  The highest observed wind at that time was 35 kt at 05Z.”


A 40 kt ob occurred sometime likely a few to several hours later, and a 45 kt ob occurred at 18Z on the 17th, as described in point 3.  The reason why 35 kt was chosen at 06Z instead of going 5 kt above the highest ob is explained in the answer to point 3.


Storm #4, 1931 (August 16-21):


            1. There are significant position changes at just about every HURDAT time during 16-19 August, not just on 17 August.  Please comment on that in the summary and in the appropriate daily sections.


The following sentences are added to the metadata summary: The position on the 16th is adjusted by about 1.5 degrees northwest of the previous HURDAT position due to a 5 kt SSW wind near 14.2N, 60.8W along with a few other observations.


Later, the following sentence is added: The position is also adjusted by over a degree on the 18th and 19th in accordance with available observations.


            2. Is there any San Juan OMR (or any OMRs from the Virgin Islands) that would give more details about the observed pressures?


No OMRs appear to be available from San Juan or the Virgin Islands.


            3. The committee is concerned that there is no conclusive evidence that this system was ever a tropical storm, but concurs there is not enough evidence currently to downgrade it.  However, the possibility that it wasn’t a storm should be mentioned in the metadata summary.


The following sentences are added to the metadata summary: “Although observations indicate that the system was likely at least a tropical depression, there were no observed winds of tropical storm force or low pressures.  This system may not have been a tropical storm since there is no conclusive evidence of tropical storm intensity.  However, there is currently not enough evidence to downgrade this system either.  It is maintained as a 35 kt tropical storm from 06Z on the 17th through 06Z on the 18th.”


Storm #6, 1931 (was #5 originally – September 6-13):


            1. Do the observations from the Windward Islands conclusively support the existence of a closed circulation on 6 September?  The surface map for 12Z more resembles an open wave, which suggests the genesis time should be delayed from that of both the original HURDAT and the proposed re-vision.  Is any time sequence of data available from Port of Spain, Trinidad?


Agreed that genesis should be delayed 24 hours because of the lack of a closed circulation on the 6th of September.  No additional observations were available from Trinidad.


            2. Should the 30 kt observation near 16N 67W on 7 September be mentioned in the daily section?  As currently written, the section gives the impression nothing significant was occurring that day.


Yes, it can be, though usually these daily highlights include gale force winds and/or pressures 1005 mb or lower.  Added to September 7th highlights: “30 kt SE and 1013 mb at 12Z at 15.8N, 67.0W (HWM).”


            3. What is the source of the 948 mb pressure reading given in the 10 September metadata?  Is this extrapolation from the 952 mb ob in Belize?  Is there a separate observation that has come to light?


The 948 mb pressure was a rough estimate of the central pressure utilizing the 952 mb peripheral pressure reading with simultaneous 52 kt winds.  However, given the uncertainties involved and that the measurement was from a coastal station not over the open ocean, the reanalysis focuses upon the 952 mb value itself.


            4. Can a RMW be estimated from the account of the center passage in Belize (Monthly Weather Review page 365)?


It appears as though the city was inside the RMW for at least a period of 50 minutes, and possibly slightly longer.  The speed of the storm was 13.5 kt.  This means that the RMW had to have been at least 11 nmi.  If the exact center of the eye passed over the city, the RMW may have been as small as 11 nmi, but it could have been larger than that.  The climatological RMW (Vickery et al. 2001) for a cyclone at that latitude with a 950 mb central pressure is 12 nmi.  Therefore, the information indicates that the RMW was not smaller than the climatological RMW value even though other information indicates the likelihood that the horizontal extent of the cyclone (radius of hurricane force winds) was likely rather small in area.


            5. The 952 mb pressure reading from Belize looks strange since it was two hours before the eye arrived in what was a rather small cyclone.  Does the Meteorological Service of Belize have any more information on this ob?


This does look strange, but it is ambiguous what actually occurred.  The 952 mb observation may have been a central pressure and it may have been a peripheral pressure.  The author of the article of the person giving the quote could have meant that 952 mb was the lowest pressure observed even though that was not explicitly stated.  That would make much more sense and with the wind observations that were given.  We know that the eye passed over the city because the wind dropped to 11 kt (and possibly lower than that) inside the RMW.  The 115 kt observation appears to be a reliable, measured observation from an anemometer.  If 952 mb was a central pressure value, that yields 109 and 111 kt according to the southern and southern/intensifying Brown et al. relationship.  Adding 5 kt and to this relationship matches the highest wind recorded and was the value chosen.


The Meteorological Service of Belize has been contacted, but as of yet has been unable to provide any additional information.


            6. A typo: “This was follow by highest wind recorded at Belize City of 115 kt N between 1950-2000Z:”


Typo fixed.


            7. It should probably be stated in the metadata summary that the intensities in the Gulf of Mexico are uncertain due to an apparent less-than-average amount of data.


Ok, a sentence has been added to the metadata.


Storm #7, 1931 (originally storm #6 – September 8-16):


            1. The major track adjustments on 8-9 September: What is the basis for these, given the lack of observations on 8 September?  Is it the data on 9 September?  Please be more specific as to why a three degree position change was made when no data seems to support it.


The position chosen at 18Z on 8 September is primarily based on data on 9 September.  Agreed that there is very little data on the 8th to justify that change and positions are adjusted northward late on the 8th and early on the 9th accordingly.  Data on the 9th though does suggest positions near those analyzed.  System was moving west-northwestward, so extended similar direction of motion and speed to point at 18Z on 8th since there is no evidence to change timing of genesis. 


2. 10 September metadata: Is the pressure entry of 966 mb at St. Thomas correct?  The metadata summary says 1000 mb, and the Monthly Weather Review (page 347) says 29.57 inches or 1001.4 mb.


The pressure entry of 966 mb at St. Thomas is not correct and should be 1001 mb. 

3. The barometer reading from San Felipe of 973 mb in the metadata summary:  Is this taken from Figures 2 on page 348 of the Monthly Weather Review?  If so, what is the chance that this refers to a pressure reading from the San Felipe *hurricane* and not the San Felipe *station*?  It is noted that the wind speeds shown in the figure drop to 10 mph, suggesting that the San Juan station was in the southern edge of the eye when the 988 mb pressure was reported.  This calls into question whether there could be a 15 mb pressure gradient between there and the center of the eye, as well as the revised peak intensity of 90 kt.


Agreed, the MWR was likely referring to the San Felipe hurricane.  The 973 mb central pressure recommended to be added to HURDAT at 00Z on the 11th is instead replaced with a 987 mb central pressure due to the 10 kt winds occurring simultaneously with a 988 mb min pressure at San Juan.  A central pressure of 987 mb yields 68 kt according to the Brown et al. southern pressure-wind relationship.  Although winds of 78 kt were estimated to have occurred at San Juan (left side of the cyclone), 75 kt is chosen for the intensity at 00Z on the 11th (down from 80 kt originally).  A 70 kt intensity is chosen for 18Z on the 10th (down from 85 kt originally).  Re-revised intensities at 06Z and 12Z on the 11th are changed to 70 and 65 kt, respectively.


4. Can the intensity decrease on 11 September be changed to a rate that better matches the known interaction with Hispaniola?


The cyclone interacts with Hispaniola from around 14Z on the 11th until 12Z on the 12th.  During this time, the cyclone is essentially moving due west along the southern coast of the island.  The intensity is reduced at 18Z on the 11th to 55 kt, 45 kt at 00Z on the 12th and bottoming out at 40 kt at 06Z.  A 60 kt ship at 19Z indicates that the system quickly rebounded once it left the coast of Hispaniola, so the intensity is bumped up to 45 kt just as the system reached the coast at 12Z and 60 kt at 18Z over the Caribbean.


            5. Are any observations available from Santo Domingo or from Jamaica for this storm?  Any account from Jamaica would be very important, as the 60 kt ship observation on 12 September suggests the possibility the system was a hurricane at landfall (see general point #2).


There are not any additional observations from those locations that we are aware of.


            6. Is the 83 kt at Veracruz shown in the Mexican record a sustained wind or a gust?


Although it is not completely certain what the averaging time of that wind is, it is likely to be a 5 min wind since this was typical of meteorological observations of the era reported in the official Mexican observation sheet.


Storm #8, 1931 (was originally storm #7 – September 23-29):


            1. Please note that the 29th is also new to HURDAT in the HURDAT section.




            2. The revised intensity of 50 kt at 12Z 25 September was based on a 45 kt ship in the metadata summary.  However, the strongest wind mentioned in the 25 September daily metadata is 40 kt.  Please correct or clarify this.


There was in fact a 45 kt ship at 12Z on 25 September.  This observation is added to the ship highlights in the 25 September daily metadata.


            3. “there were likely no ships near the storm that day”.  The chart in the binder shows several ships in the circulation, but none in the maximum wind area.  Please re-phrase this.




            4.  See general point #2.  The committee generally feels that this system was a hurricane, but other supporting evidence may need to be played up.  For example, the 991 mb ship report with coincident 45 kt winds suggests a central pressure in the upper 980 mb range, which would help support hurricane strength.


Agreed, the following is added to the metadata: “Using the 10-kt per mb rule (assuming 991 mb with 45 kt is inside the RMW), the central pressure was likely near 987 mb.  A central pressure of 987 mb yields a wind speed of 66 kt according to the Landsea et al. (2004) north of 35N pressure-wind relationship.  This observation, combined with the 60 kt wind observed three hours earlier, provide good evidence that this system was a 65-kt hurricane.”


Storm #10, 1931 (was original storm #8 – October 18-22):


            1. “The system is analyzed to have made a Cuban landfall at 18Z on 19 October as a 40 kt tropical storm”.  The original HURDAT intensity at that time was 35 kt.  What is the basis for the stronger intensity at landfall?  Ship report?  Data from Cuba?  Is the change in intensity due to the change in landfall time?  Please state the basis in the appropriate locations.


The following is added to the metadata summary: “The reason why the HURDAT intensity at 18Z on the 19th is increased from 35 to 40 kt is because the landfall time was changed due to the track changes.  HURDAT originally showed a 40 kt intensity at the point before Cuban landfall, and no changes were made to this 40 kt intensity for Cuba.”


            2. The committee concurs with the extratropical transition at 00Z 21 October.  The Historical Weather Maps show a frontal wave over the eastern Atlantic on 23 October east of the final position, so there is a possibility the cyclone dissipated later than currently thought.  Please look into this.


On the 22nd at 15Z, a single observation with an easterly wind component confirms the circulation is still closed – 10 kt NE with 1001 mb at 32.7N, 51.4W.  Thereafter, there are no more observations containing easterly wind components (even after obtaining the COADS for the 23rd).  There is no evidence of a closed circulation on the 23rd.  Please see observations plotted on a map of 10/23 00Z and 06Z.


Storm #11, 1931 (new storm – November 1-5):


1.  While the committee is in favor of adding this system, it is not ready to do so just yet.  The data suggests that many of the strong wind observations were due to a cold surge moving southward from the Gulf of Mexico, which calls into question how tropical this system was.  There is a need for additional analysis on the thermal structure of the cyclone, and how far from the center were the strongest winds.


Isotherms in increments of 5°F have been analyzed on the maps for 1-4 November.  Observations indicate rather isothermal structure on all of the days this system exhibited tropical storm force winds.  On the 2nd, peak winds of 45 kt N from one ship was about 120 nm from the center, while a second ship with peak winds of 35 kt NNW was about 75 nm from the center.  On the 3rd, the peak 50 kt WNW was about 200 nm from the center and a 45 kt NW wind was about 160 nm from the center.  On the 4th, peak SW, NW and WNW winds of 35 kt were about 150, 90 and 125 nm from the center, respectively.  These generally large distances from the center suggest a subtropical storm structure.


2. There are also concerns about the track during the dissipating phase on 4-5 November, as a more eastward position at 12Z 5 November would be more consistent with the westerly winds at Cabo Gracias.  Please re-examine this.


Agreed, the position at 12Z on the 5th is changed to 18.0N, 84.0W.  The following positions are also changed.  11/4 18Z: 17.5N, 85.0W.  11/5 00Z: 17.6N, 84.7W.  11/5 06Z: 17.8N, 84.4W.


Storm #12, 1931 (new storm – November 11-16):


            1. The committee concurs with adding this system to HURDAT.




            2. There are a couple of caveats about the genesis.  The data for starting it on 6 November looks less than conclusive, and the rising pressures in the southwestern Caribbean on 10-11 November make it unclear whether this is one system or two separate systems – a low from 7-9 November followed by a tropical storm from 11-16 November.  Please re-examine this aspect of the system.


It appears that there is not enough evidence to connect the low from 7-9 November to the low from 11-16 November to say conclusively that they are the same system.  Some of the data from the 7-9 low looks spurious (like the erroneous 1005 mb ship ob) so this one appears only to have been a low or a tropical depression.  This low may have moved off to the north-northeast and dissipated, and a new system may have become the tropical cyclone of the 11-16th.  Or the original system may have hung around and then became the tropical cyclone.  It is decided to maintain the tropical cyclone from 11-16 November with the same positions and intensities, but the 7th through the 11th at 06Z are deleted.


Storm #13, 1931 (originally storm #9 – November 22-25):


            1. The committee concurs with the proposed changes in track and intensity.




            2. Was this possibly a subtropical storm or go through a subtropical phase?  If there is evidence to support this possibility please mention it in the metadata.


Agreed.  It is possible that the system could have been a subtropical storm for some of its lifetime.


            3. The committee would like to see additional analyses for a couple of days before and after what is in the binder to we can better understand the genesis and dissipation of the system.


Ok.  Additional analyses were done for two days before (November 20 and 21) and one day after (November 26).  Nothing too interesting was revealed from these analyses before.  After, on November 26 at 00Z the general area of the remnant of the circulation can be determined to be between 26-27N, 76-77W.


Also of note is that toward the back of the 1931 binder, in the email from Mike Chenoweth  there is information discussing flooding rains in St. Lucia on the 21st (may or may not be related to precursor of system).  A portion of this has been added to the metadata writeup.


Additional Notes:


            System #1: Two points on this system.  First, the committee feels there may be enough evidence that this system was a short-lived tropical storm before it hit Hispaniola.  Not only was there a gale-force ship, but the 1007 mb ship had a simultaneous winds of 30 kt, suggesting a central pressure of 1004-1006 mb.  Given the pressures of 1015 mb in the Turks and Caicos Islands, this would also support tropical-storm intensity.  What were the westerly wind speeds in Puerto Rico (to help infer what the winds were to the north of the center)?


After reassessing this system, it is agreed that there is enough evidence that this was a weak, short-lived tropical storm.  It is now added into HURDAT as new storm #5 from 1 to 4 September.


            Second, during the passage across western Cuba and Florida, has that data been analyzed in detail (especially the Florida OMRs)?  It appears unlikely that this was a tropical storm in western Cuba and Florida, but it won’t hurt to re-examine the data – especially if this system is added as a tropical storm.


The Miami, FL OMRs were obtained from the NHC library and the maximum wind at Miami was 26 kt S during the day on the 7th.  A few hours later, the minimum pressure at Miami was 1010 mb.  The wind then shifted to SW and pressure rose a few mb by 00Z on the 8th.  A few inches of rain fell on the 7th in Miami.  No other OMRs were able to be obtained.  It is more likely that the system never attained a closed circulation again after landfall in Hispaniola.  There are no observed westerly winds on the 4th and 5th, and available observations on the 6th and 7th indicate a trough.


Other comments: The Monthly Weather Review indicates one other system that should be investigated to see if they should be included in HURDAT, or for possible inclusion in the Additional notes section.


            1. A low pressure system formed over the Bahamas on 12-13 October and moved generally northward off the southeastern U. S. coast for a couple of days, followed by a turn toward the northeast ahead of an oncoming frontal system.  This low may have formed from a decaying frontal boundary, but the Historical Weather Maps analyze it as non-frontal.  Although the Monthly Weather Review states it was not very strong, it should be analyzed to see if it (at the least) should be included in the Additional Notes.


After obtaining additional observations and reassessing the data, this system should be added into HURDAT as a new tropical storm (new #9).  Data was obtained from 12-17 October for this suspect.  The data indicates that this system was a tropical cyclone from approximately 13 October around 12Z until it became extratropical on 16 October at either 06 or 12Z.  There was one observed gale with this system while it was likely a tropical cyclone – 35 kt S on the 15th from the ship Maravi (MWR October gale chart).  There was one observed low pressure while this system was likely a tropical cyclone – 1003 mb from the ship Maravi on the 15th at 22Z (observed simultaneous with 30 kt SW winds) at 28.9N, 73.3W (MWR).  Temperatures all around the cyclone were isothermal and in the 70s.  Although observations indicate that the wind field of cyclone may have been somewhat broad, the structure is compact and symmetric enough to be considered a tropical cyclone.  The system is analyzed as a 30 kt tropical depression from 13 October at 12Z through 15 October at 00Z.  A 35 kt tropical storm is analyzed for 06-12Z on the 15th, increasing to a 40 kt tropical storm at 18Z on the 15th – 00Z on the 16th.  06Z on the 16th is chosen for the time this cyclone became extratropical.  Although there was only a slight temperature gradient across the low by early on the 16th, the structure started to became increasingly asymmetric rather quickly and the strongest observed winds were located far from the center by 12Z on the 16th with light winds and low pressures dominant within 100-150 nmi of the center.

[Horizontal Rule]

[OAR/DOC/NOAA Logos] Atlantic Oceanographic and Meteorological Laboratory Logo [United States Department of Commerce] [Atlantic Oceanographic and Meteorological Laboratory] Atlantic Oceanographic and Meteorological Laboratory Logo [National Oceanic and Atmospheric Administration] [Office of Oceanic and Atmospheric Research] Department of Commerce Logo National Oceanic and Atmospheric Administration Logo Ocean and Atmospheric Research Logo

  Disclaimer | Privacy Policy