Printer Friendly Version
Response to inquiry from Georgia Institute of Technology

A series of questions was submitted to the committee through the Earth and Atmospheric Sciences Department at the Georgia Institute of Technology concerning the assignment of Saffir/Simpson category for pre-1915 landfalls. The committee took as a starting point a draft response prepared by Chris Landsea (as noted) but has differed on some points. The questions, preceded by year and storm number, are unedited. We note that the track revisions themselves stand (i.e. position and intensity) but the implications (effects as expressed by the Saffir/Simpson category) are in some cases revised. While direct measurements are the best basis for this determination, they are often lacking, especially during the 19th century. A key factor, storm structure, is quite variable and can sometimes only be assumed. The committee has sought to make its best judgment in this somewhat subjective process.

Several of the questions concerned the inclusion of a Saffir-Simpson designation for hurricane-force winds occurring inland, for example, in Georgia following landfall along the Florida panhandle. The committee reviewed this issue. It was decided that in well-documented cases, inclusion would enhance the dataset, as long as these cases could be identified as such. It was agreed that an ?I? prefix would serve this purpose. Thus, for example, AFL2, IGA1, would clearly indicate a Florida landfall, with hurricane-force winds subsequently experienced in Georgia. The category describes the strongest effects observed in that state, analogous to usage for coastal landfall. An Atlantic landfall along the Georgia coast will continue to be noted as GA1. It was also decided that given the peninsular nature of Florida, and its existing subdivision into four quadrants, the use of the ?I? designator would typically not be necessary. That is, coastal effects are likely in almost all instances. However, its use will not be completely ruled out.

1852 / Storm 1 ? The current designation for SW FL is that of Category one impact. The data point used to make this designation appears to be that of 12Z on 08/22, based on Table 8. This data point has a placement of 23.8N and 81.1W with 90kt winds in the best track data and places the storm center approximately 50 nautical miles from the FL keys. The two following data points of 18Z on 08/22 and 0Z on 08/23 place the storm first within 25 nautical miles of Key West with 90kt winds and then passing over Dry Tortugas with the same 90kt winds. The recommendation is to either adjust BFL1 to BFL2 or change the best track data so that winds for BFL will be category one strength during this period.

Response : The track is not adjusted. Given that the track does pass over the Dry Tortugas, which are within unincorporated Monroe County, the landfall designation is changed to BFL2.

1857 / Storm 2 ? NC impact is designated as category one and supported by a reading of 35.2N 75.7W with 80kt winds in Table 8. However the following best track reading on 09/11 12Z located at 35.4N 75.5W has 90kt winds and is within the outer banks of NC. The recommendation is to either change the designation NC1 to NC2 or adjust the winds of the best track data over this stretch to category one strength.

Response (from Landsea, with further inquiry, committee concurs): After further consulting additional references including Tannehill (1938), Dunn and Miller (1960), and Barnes (2001), none of these mention this hurricane or indicate any significant impacts in North Carolina. While the Outer Banks were sparsely populated in the 1850s, Category 2 conditions along the coast would likely have been strong enough to have caused significant impacts and at least some note in the historical summary books. It is likely the case that the center of the Category 2 hurricane passed over Cape Hatteras. However, because of the asymmetry in the wind field induced by the motion of the hurricane, the maximum winds of 90 kt will be on the right side of the storm ? which would be offshore. So for a location that experiences the eye of a hurricane that has a substantial forward speed, that location will likely not get the impact of the strongest winds in the hurricane. Please refer to Powell and Houston (1996) and Landsea et al. (2004) regarding Hurricane Andrew for a more recent example of this phenomenon.

1867 / Storm 7 ? Southern TX (ATX) has a designation of category one. Best track data shows a position of 25.9N 97.1W with 90kt winds on 10/02 at 18Z. This places the storm less than 4 nautical miles from the TX coast. The Metadata related to the reanalysis of this storm states, ?moved track closer to Brownsville, Texas, as it appears that the town was in the western eyewall.? So even though the eyewall appeared to be inland over 20 nautical miles from the tracking data point, a designation of only category one was made for ATX. The recommendation is for either ATX1 to be changed to ATX2 or for the best track data to be revised.

Response : The track is not adjusted. Given the configuration of the coastline and direction of motion, ATX1 is retained.

1871 / Storm 6 ? NW FL (AFL) currently has a designation of category one. The Metadata suggests that the designation was made from a reading on a ship (Lizzie M. Merrill) approximately 94 nautical miles away from the impact location of 29.2N 83.0W where an estimate of 70kt winds is given and is the same as best track winds given for the reading just prior to landfall. The wind estimate for the ship in the raw data was also 70kt. The ship is also located just south of the North/South divider for FL. The recommendation is for BFL to obtain a category one designation and for AFL to be upgraded to category 2 given the likely increase in winds over the 100 nautical mile distance.

Response (from Landsea, committee concurs): It is agreed that southwest Florida should be included as a Category 1 impact as well. This is now added into HURDAT. However, because of the uncertainty in the exact track (as well as the location of the ship) and given the lack of inner core data for this system, it is retained as a Category 1 hurricane both over the Gulf of Mexico as well as at landfall in Northeast Florida. It definitely could have been a stronger hurricane, but without more substantial evidence this cannot be included as such. Moreover, large Category 1 hurricanes do occur as well. (Committee further notes that system could have been stronger, given time of occurrence in early September.)

1874 / Storm 6 ? NW FL has a designation of category one. The Metadata, Raw Observations and Partagas report have no supporting data elements that indicate a need to increase the single best track point in the Gulf to have 70kt winds just before landfall. This update appears to have led to the category one designation for AFL. The recommendation is that if this change was made because of the strength of Storm 6 as it entered the Atlantic, this information should be stated in the Metadata file. Otherwise eliminate the AFL1 designation.

Response (from Landsea, committee concurs): This system is designated as a hurricane in the Gulf of Mexico and for Northwest Florida based upon the Lake City sea level pressure reading of 996 mb around 06 UTC on the 28th. This value (likely not a central pressure) was taken well inland a few hours after landfall. It is likely that the central pressure at landfall was on the order of 980 to 990 mb with winds around 70 kt. Thus it is retained as a Category 1 landfall for northwest Florida. (note: this observation will be added to the metadata.)

1878 / Storm 11 ? Table 8 that is part of the documentation related to the HURDAT reanalysis project lists PA as impacted at category one strength. Now that the HURDAT website has been revised to indicate that footer data includes states impacted not specifically landfalls, it is recommended that the HURDAT footer for this storm include a PA1 designation as it appears this change should have been made based on the Metadata file.

Response : Committee agrees with Landsea that IPA1 should be added (see note above on notation, as system was well inland). It should also be noted that in addition to wide-spread wind damage, a significant storm surge, driven up Delaware Bay, was observed in Philadelphia.

1879 / Storm 2 ? The Metadata shows a reduction in wind speed as the storm approached MA. This appears to be the reason that the MA1 designation was removed from the storm footer information. However, the best track data point of 6Z on 08/19 at 41.4N 70.8W is within the MA barrier islands with 70 kt winds even after the revision in the Metadata file. The recommendation is to either restore the MA1 designation or adjust the best track data accordingly.

Response : Although storm size must be considered, it is difficult to rule out that some hurricane force winds occurred along the coast, hence designation of MA1 is accepted.

1882 / Storm 6 ? Table 8 shows a pressure reading developed from peripheral readings as the justification for the 70kt wind speeds at the designated impact location. This is referenced in the Metadata as well, however nothing in these sources, the raw data for Storm 6, or the Partagas report produce readings that appear to be useful anywhere near the time of impact. The recommendation is to provide reference to that data that was used for this designation or reduce the storm intensity from a category one impact for AFL to TS.

Response (from Landsea, committee concurs): This system is ambiguous as to its Florida landfall intensity due to lack of substantial data between the time it impacted Cuba until it struck Florida. Given the uncertainty, it is possible that it was not of hurricane force. However, as it restrengthened significantly into a hurricane soon after re-emerging into the North Atlantic, the system was likely of at least strong tropical storm or minimal hurricane force when it struck Florida. Moreover, categorization of the system as a Category 1 hurricane at landfall in Florida is consistent with Dunn and Miller's (1960) assessment of this being a "minimal" tropical cyclone, which in their terminology is a Category 1 or 2 hurricane. Thus Category 1 at landfall is retained.

1885 / Storm 2 ? Table 8 and the HURDAT data list the states affected in the order of SC, NC, GA, DFL. This listing is not consistent with the designations for other storms listed where impacted states are in sequential order of impact occurrence (for instance Rita 2005). The recommendation is to reorder the states to DFL GA SC NC.

Response (from Landsea, committee concurs): While the listing of US impacts by state is usually in sequential order, this is not necessary and is of no consequence. No change is needed.

1886 / Storm 3 ? Table 8 lists an impact of 29.7N 85.2W with 85kt winds, and then because of the wind decay model as discussed in the Metadata, the next best track point revision is 0Z on 07/01 at 30.4N 84.4W with winds at 70kts. This places the storm center within 17 nautical miles of GA. The recommendation is to list GA as a state impacted at category one levels or alter the best track and FL impact levels accordingly.

Response : Given the notation discussed above, it is agreed that IGA1 will be added for this system.

1887 / Storm 4 ? Similar concern as that of the previous storm that made impact on the panhandle of FL. The best track data which was revised as noted in the Metadata has a center location less than 6 miles from AL on 07/27 at 18Z with 65kt winds. This would suggest that AL was also impacted with category one winds. Recommendation is that AL receives a category one designation or that the best track data be revised accordingly.

Response : Given the notation discussed above, it is agreed that IAL1 will be added for this system.

1888 / Storm 6 ? The Metadata file discusses the reasoning behind the recorded 70kt winds. The best track data therefore contains a reading on 09/26 at 12Z that is positioned 3.5 nautical miles from Nantucket at 41.2N 70.2W with 70kt winds. However, no impact designation was made for MA. The recommendation is to indicate MA as a state impacted at category one strength or adjust the best track data accordingly.

Response (from Landsea, with further inquiry, committee concurs): The center of the hurricane passed very close to Massachusetts, but the peak winds Category 1 winds remained offshore. This analysis is further strengthened by obtaining the U.S. Army Signal Corp observations from Nantucket, Massachusetts, which observed a maximum of 52 kt. Because of the high bias of the anemometers utilized in that era (Ferguson and Covert, 1924) this converts to only 41 kt for a 5 min peak or 43 kt for a 1 min peak (Powell and Houston, 1996). Given the exposure and location of Nantucket relative to the track of the hurricane, Nantucket should have received close to the strongest winds in the storm anywhere in Massachusetts. Thus the 55 kt tropical storm impact for Massachusetts is retained.

1891 / Storm 3 ? As the Metadata indicates, the storm intensity at US impact is highly suspect. The normal procedure based on Table 8 seems to be a downgrade procedure on these storms. Should it really be listed as a hurricane when even the record prior to the reanalysis questioned the strength of Storm 3?

Response (from Landsea, committee concurs): This system is listed as a minimal hurricane at landfall in Southeast Florida. This portion of the coast was quite unpopulated at the time and little concrete information is available to determine the intensity one way or another. (This is why the documentation suggests that accurate records begin roughly in 1900 for determining the number and intensity of hurricanes for this region.) The process for altering HURDAT is to make changes (either for stronger or weaker) when observational data is available. In the absence of data to the contrary, the decision is generally to retain the original HURDAT estimates. Thus this system is retained as a minimal Category 1 hurricane at landfall in Southeast Florida, though it could certainly have weakened to a tropical storm.

1893 / Storm 4 ? While the NHC BTCC comments indicated that the record had been corrected per our suggestions, only the issues of the VA impact was resolved. The best track data show the storm within 4 nautical miles of the NJ coast on 08/24 with very strong category one level winds. The recommendation is to show NJ as impacted or to alter the best track data accordingly.

Response : Given the speed and asymmetries, it is unlikely that New Jersey experienced hurricane force winds, thus this recommendation is not accepted. However, upon review it seems that Connecticut should be included. CT1 to be added to HURDAT.

1893 / Storm 6 ? Best track data shows FL was impacted first by this storm but it is listed last in the footer of the HURDAT record. The recommendation is to move the DFL1 designation to the first position on the states impacted list in the footer.

Response (from Landsea, committee concurs): While the listing of US impacts by state is usually in sequential order, this is not necessary and is of no consequence. No change is needed.

1894 / Storm 5 ? Raw data wind observations from Block Island of 73kt appear to support the Metadata decision for providing category one designations for NY and RI. Given that this reading is less than 15 nautical miles from CT and the best track direction of the storm, it seems that CT should have also received a category one designation. The recommendation is for CT1 to be added to the footer information for this storm.

Response : Hurricane force winds in Connecticut cannot be ruled out, thus CT1 will be added for this system.

1908 / Storm 5 ? Best track data for 09/01 shows a track that makes landfall with TS force winds over mainland NC and continues over the outer banks with the same TS level winds of 45kt. Recommendation is to adjust the header to indicate a landfall or adjust the best track data for lower winds and/or move the storm farther out in the Atlantic.

Response (from Landsea, committee concurs): Agreed. This system should be listed as a tropical storm landfall for North Carolina.

1911 / Storm 3 ? The best track data show a storm that made impact clearly in SC before the storm move inland into GA. However the HURDAT footer for the storm list GA before SC. The recommendation is to put SC before GA in the HURDAT footer.

Response (from Landsea, committee concurs): While the listing of US impacts by state is usually in sequential order, this is not necessary and is of no consequence. No change is needed.


Dunn, G. E., and Miller, B. I., 1960: Atlantic Hurricanes, Louisiana State University Press, Baton Rouge, La., 326 pp.

Fergusson, S. P., and R. N. Covert, 1924: New standards of anemometry. Mon. Wea. Rev., 52, 216-218.

Landsea, C.W., J. L. Franklin, C. J. McAdie, J. L. Beven II, J. M.Gross, R. J. Pasch, E. N. Rappaport, J. P. Dunion, and P. P. Dodge, 2004. "A re-analysis of Hurricane Andrew's (1992) intensity." Bull. Amer. Meteor. Soc., 85, 1699-1712.

Powell, M. D., and S. H. Houston, 1996: Hurricane Andrew's landfall in South Florida. Part II: Surface wind fields and potential real-time applications. Wea. Forecasting, 11, 329-349.

Tannehill, I. R., 1938: Hurricanes, their nature and history. Princeton University Press, Princeton, N.J., 257 pp.

Return to Overview page

Stay Connected