By: ncforecaster , 5:37 AM GMT en Diciembre 01, 2011
This particular blog entry is a complete and thorough accounting of all known North Atlantic basin tropical cyclones (TC)-of at least tropical storm (TS) intensity-that developed during the period of 1851-2010, as contained in HURDAT.
What is HURDAT?
In simplest terms, HURDAT is the "official" hurricane database that contains a detailed record of all tropical storms (TS) and hurricanes (H) known to have developed somewhere within the North Atlantic basin-which includes the Atlantic Ocean, Carribean Sea, and the Gulf of Mexico-for the period of 1851 to the present. It is also the official record for all landfalling tropical storms and hurricanes known to have impacted the United States (U.S.) coastlines, as well. This database was initially created in support of the Apollo Space program in the 1960's, and contained six-hourly positions and intensities for all TC's that had been documented up until that time (Jarvinen et al. 1984). Unfortunately, there were (and still are) many systematic and random errors within the database that needed to be corrected (Neumann, 1994). Another issue that needed to be addressed were biases contained in the database resulting from a greater understanding of TC's that had developed over the years, and the enhanced analysis techniques used by NHC forecasters (Landsea, 1993).
As a result, researchers with the Hurricane Research Division (HRD)-led by Chris Landsea- undertook the work to correct these "errors" through a project called: "The Atlantic Hurricane Database Re-analysis Project (HRP)."
The stated "objective" of the HRP can be found on the main page of its website here, and specifies the following: "The Atlantic Hurricane Database Re-analysis Project is an effort led by the Hurricane Research Division to extend and revise the OAR's North Atlantic hurricane database (or HURDAT). Going back to 1851 and revisiting storms in more recent years, information on tropical cyclones is revised using an enhanced collection of historical meteorological data in the context of today's scientific understanding of hurricane and analysis techniques."
The stated "Goal" of the HRP can also be found at the above link and entails the following: "The primary goal for this project is to provide an extended and corrected Atlantic hurricane database of individual tropical cyclone tracks and intensities for both the entire Atlantic basin as well as U.S. landfalling storms. This fits in well with the goals of NOAA and HRD to better understand variability of extreme events, such as tropical storms.
As result of these efforts, HURDAT has been extended back in time to 1851 (from its original starting point of 1886), and an extensive revision has been made to the database for all known TCs through the 1930 North Atlantic basin hurricane season, as of the time of this writing.
Limitations to the historical record:
Although the HRP has greatly improved the accuracy of HURDAT (and the associated historical record contained therein), there are some significant and inherent limitations that mitigate against a complete accounting of all TC's that have traversed the warm waters of the North Atlantic basin, as well as all TC's that have impacted the U.S. coastline, throughout this period of record.
As one might expect, the historical record is considered less complete as one goes back farther in time, especially for the period that preceded the introduction of organized aircraft reconnaissance and the installation of geostationary satellites into space. With the aforementioned in mind, let's take a brief overview of the inherent limitations-in respect to an accurate accounting of all Atlantic basin TC'S-that are entailed within the aforementioned period of record.
During the second half of the 19th century-corresponding to the first fifty years of the historical record-the biggest impediment to an accurate accounting of the number (and intensity) of TC's, as well as those that made a U.S. landfall, was the lack of resources available to detect their existence (and their true intensity). Unlike today, there were no geostationary or polar orbiting satellites to locate storms out over the open waters of the Atlantic. Recon flights into TC's had yet to be initiated, so an accurate estimation of a particular TC's intensity was greatly inhibited (likely underestimated). There also weren't any radar detection capabilities during this period of time, either. As a result, weather observers of the time had to rely on the sporadic data collected by ships at sea and/or the few observation sites that were unfortunate enough to find themselves in the path of one of these menacing storms. Even then, further complications existed. Without the access of radio transmission (unavailable until 1905), weather observers had to wait until ships had returned to port in order to review their respective ship logs. Worse yet, many more ships were unable to survive their encounter with these tropical terrors and the data was lost forever, along with the men who had taken it. In short, the detection of tropical storms and hurricanes was essentially limited to those TCs that affected ships at sea and/or those that impacted a populated land area. In other words, if there wasn't someone in the area where a respective TC happened to traverse, there could be no record of its existence.
The latter case mentioned above is problematic since many areas of the U.S. coastline was still very much sparsely populated. Case in point, one of the most hurricane prone areas of the U.S. coastline-Miami, Fl.-wasn't incorporated until 1896. Making matters worse-as far as a complete documentation of all TC landfalls is concerned-a small hurricane (like hurricanes Andrew 0f 1992, Bret of 1999, and Charley of 2005) likely went undetected because of the confined area of damaging winds and the sparsity of the population along many areas of the U.S. coastline, during this period in time.
Taking all the aforementioned in to account, the researchers with the Hurricane Research Division estimate that the number of "missed" TC's (those likely unaccounted for in the historical record) ranged upwards of 6 per year, for the period of 1851-1885. With the increase in ship traffic and population increases along coastal areas, the estimation of "missed" TC's (per season) decreases to a range of up to 4 per year for the period of 1886-1900 (Landsea et al. 2004).
As stated by Landsea, et. al. (in the "Documentation for 1851-1910 Alterations and Additions to the HURDAT Database), "By no means should the tropical cyclone record over the Atlantic Ocean be considered complete for either the frequency or intensity of tropical storms and hurricanes for the years of 1851 to 1910."
The authors go on to say, "However, more accurate and complete information is available for landfalling tropical cyclones along much of the United States coastline." That being said, they also note the following: "Because of the lack of continuously populated coastal regions over this era, this record represents an incomplete listing of the frequency and intensity of tropical cyclones that have impacted the United States".
The first half of the 20th century-corresponding to the second fifty-year period of the historical record-saw significant technological advances and the advent of new resources that greatly enhanced the detection, observation, and forecasting of tropical cyclones.
The first significant milestone of this particular era began with the advent of radio transmission that originated in 1902. It allowed for direct ship to shore communications-which was impossible prior to this time-which enabled forecasters to gather the data in real-time and use it operationally. It wasn't too long before the first known observational weather report was made from a ship on December 3, 1905, and it has been an invaluable tool for forecasters and mariners alike, ever since (Neumann et al. 1999). Before this major technological advancement occurred, weather observers had no way of gathering data from, or about, any TCs that didn't have a direct impact on land. Ironically, there was a downfall to this new found ability to transmit real-time data to and/from ships at sea, in that ships now had the information necessary to intentionally avoid encountering these devastating storms. With that, there was a "modest decrease" in the number of ship observations during this particular period.
Another huge break-through in TC detection and observation took place on July 27, 1943. On that fateful day, Major Joesph Duckworth flew a "propeller-driven, single-engine North American AT-6 Texan trainer into the eye of a hurricane". When he returned to base, the stations weather officer asked Maj. Duckworth to take him along for a second penetration into the eye of the storm (53rd Weather Recon Fact Sheet). With these two historic flights, Maj. Duckworth helped pave the way for the countless reconnaissance missions that would follow. These reconnaissance flights, begun in the summer of 1944, have greatly improved the science of TC forecasting, and the importance of which, can hardly be overstated. On the other hand, regular reconnaissance (Recon) flights into TCs was still in its infancy, during the latter part of the period we are examining. In the beginning, military units were tasked to fly regular reconnaissance missions/tracks across the Atlantic between North America and Allied Western Europe-only during daylight hours-in order to detect the presence of TCs. This definitely helped improve the detection of some TCs, but many others were "missed" because they didn't track through the areas RECON fights were tasked to cover. Even when RECON did locate a storm, or were sent out to investigate the intensity of one, they were hesitant to fly into the eyewall (in order to get to the eye) of strong hurricanes. Consequently, a direct measurement of a TCs central pressure was somewhat rare during the first decade of aircraft reconnaissance flights. Regardless, the implementation of a formal program for regular aircraft reconnaissance missions to gather data on prospective TC's and determining the intensity of others, has proven to be one of the most important advances in the science of TC forecasting and observation.
In short, this period of record (1901-1950) saw an increased improvement in the detection, observation, and forecasting of TC's. The continued increase in population along coastal areas of the U.S.-during this period-also provided for a more accurate accounting of U.S. tropical cyclone landfalls. That being said, Landsea, et. al, stated the following in "A Reanalysis of the 1911-1920 Atlantic Hurricane database": "It is estimated that more than three tropical cyclone's a year were likely missed in the pre-geostationary satellite era between 1900 and 1965 (Landsea 2007)". It serves as a reminder of the limitations in TC detection, observation, and forecasting that existed throughout the era of 1901-1950.
As noted in the previous section, an organized program for aircraft reconnaissance (RECON) flights began in 1944. From that time to the present, RECON flights have greatly improved the science of TC detection, forecasting, and observation. On the other hand, these RECON flights only covered about one-half of the North Atlantic, and were generally only flown into storms that had traversed areas west of 50W longitude. Consequently, it is highly likely that TC remaining in the far eastern Atlantic basin, went undetected prior to 1966.
Arguably, the single greatest technological advancement (in TC detection) began in April of 1960-with the launch of the very first experimental weather satellite into space. This monumental achievement paved the way for the subsequent launch of the very first geostationary satellite into space, just 6 1/2 years thereafter. Unlike the previous "Polar" orbiting satellite, the geostationary weather satellites provided weather forecasters with a complete view of the entire North Atlantic basin. As a result, TC "counts" are considered most reliable for the period beginning in 1966. Other technological advancements since that time-some of which will be discussed in the next section-have led hurricane researchers (such as Chris Landsea) to suggest that it is exceedingly likely that one additional TC per season, remains unaccounted for in the historical record, for the period of 1966-2000.
This period also saw other significant technological advancements. They include, but aren't limited to, the implementation of a coastal radar network in 1955, deployment of ocean data collecting buoys in the early 1970's, use of aircraft launched dropsondes beginning in the early 1970's, and the development of highly sophisticated computer models to help determine the forecast track and intensity of TC's. This particular period also saw the development of various other tools and techniques that were derived to help better understand, forecast, and track TCs. One such example is the Dvorak technique-developed in 1973-that is used to estimate the intensity of a TC based solely on visible and Infra-red satellite imagery.
In short, this period of record (1951-2000) saw a remarkable improvement in the detection, observation, and forecasting of TC's. Moreover, it also coincided with an era of explosive population growth along coastal sections of the U.S. Consequently, there is very little, if any doubt, about the accuracy of TC counts for the U.S. mainland, during this particular period of the historical record. As noted in the previous blog entry, Landsea, et. al. stated that "a little more than three tropical cyclone's a year were likely missed in the pre-geostationary satellite era between 1900 and 1965 (Landsea 2007)". For the period of time thereafter (1966-2000), Landsea, et. al. estimates that roughly one tropical cyclone (per season) remains unaccounted for in the historical record.
Unlike the previous eras we've discussed, this particular era is considered to be the most accurate for both TC "counts" and storm intensity. In fact, technological advances during the past decade (corresponding to the first decade of the twenty-first century) have made it possible to detect all TC's that form anywhere within the entire Atlantic basin.
These new technological advances have provided hurricane forecasters with new tools and data sources that has led to continued improvement in the detection, observation, and forecasting of TC's. These "new tools and data sources" include Quikscat, the advanced microwave sounding unit, and the cyclone phase space analysis. Without these, it is estimated that one TC per season would've been "missed"-likely misidentified as an extra-tropical cyclone instead-during the past decade. Obviously, the same would be true for the entire historical record that preceded it.
In short, this period of record (2001-2010) saw a continued improvement in the detection, observation, and forecasting of TCs. Moreover, it also coincided with continued population growth along all coastal sections of the United States. Consequently, there is essentially very little doubt about the accuracy of both TC counts and storm intensity for the U.S. mainland, during this particular period of the historical record. That being said, there are still inherent limitations within the inexact science of TC forecasting that makes a precise estimation of the maximum sustained wind associated with individual landfalling hurricanes somewhat more ambiguous. Fortunately, this too continues to improve over time, as new technological tools and data sources continue to become available.
With all the aforementioned taken into consideration, let's take an extensive look at some of the most "fascinating facts" of the entire historical record for tropical cyclone activity throughout the North Atlantic basin.
1) Total number of tropical cyclones (includes Subtropical Storms): 1446 (an average of 9.04 per season).
2) Total number of hurricanes: 853 (an average of 5.31 per season).
3) Total number of "major" hurricanes: 308 (an average of 1.93 per season).
Note: One HURDAT list doesn't include H Carol of 1954 in its MH totals, while another one does. For this compilation, H Carol is accounted for in the total of MHs.
4) Total number of TC's by month:
a) September = 492
b) August = 366
c) October = 294
d) July = 110
e) June = 83
f) November = 65
g) May = 21
h) December = 10
i) April = 2
j) March = 1
k) February = 1
l) January = 1
Note: These statistics are reflective of the month in which TC's initially achieved TS or STS intensity.
5) Total number of H's by month:
a) September = 332
b) August = 225
c) October = 161
d) July = 54
e) November = 41
f) June = 32
g) May = 4
h) December = 4
i) March = 1
j) Others = 0
6) Total number of MH's by month:
a) September = 148
b) August = 87
c) October = 51
d) July = 10
e) November = 7
f) June = 4
g) May = 1
h) December = 0
i) Others = 0
7) Most intense TC recorded:
There are two different equations or meterological parameters that can be used to determine the strongest hurricane to make a U.S. landfall. One is based on a respective storms MSW, while the other focuses on its minimum central barometric pressure. That being said, a hurricanes lowest central pressure at landfall is typically used to categorize a listing of the "most intense" TCs to develop within a particular ocean basin. With that in mind, Storm #22 (hurricane Wilma) of October 19, 2005 is the "most intense" TC known to have traversed the Atlantic basin during the period of 1851-present. At peak intensity, it had a lowest central pressure reading of 882 mb.
The strongest known hurricanes to develop anywhere within the Atlantic basin-in terms of MSW-are hurricane Camille of August 17, 1969 and hurricane Allen of August 7, 1980. At peak intensity, each one of the aforementioned hurricanes generated a MSW of 190 mph.
8) Most TC's to form in one season: The 2005 H season produced 28 documented TC's of tropical storm and/or hurricane intensity.
9) Most hurricanes to develop in one season: The 2005 H season also produced the most TC's of hurricane intensity, with 15.
10) Most "major" hurricanes to develop in one season: The 1950 H season spawned the most TC's of MH intensity-during this period-with 8.
11) Total number of TC's to make a U.S. landfall: 575 (an average of 3.59 per season).
12) Total number of U.S. landfalling hurricanes: 285 (an average of 1.79 per season).
13) Total number of U.S. "major" landfalling hurricanes: 95 (an average of 0.60 per season).
14) Total number of U.S. landfalling TC's by month:
a) September = 200
b) August = 134
c) October = 109
d) June = 63
e) July = 53
f) November = 14
g) May = 8
h) December = 0
i) February = 1
j) Others = 0
Note: Storm #6 of 1885 (Sept./Oct.), Storm #2 of 1899, Storm #5 of 1933, Storm #3 of 1953, H Edna of 1954, H Elena of 1985, and H Dennis of 1999 each made a U.S. landfall and/or strike during two separate months, respectively.
15) Total number of U.S. landfalling H's by month:
a) September = 106
b) August = 78
c) October = 53
d) July = 26
e) June = 19
f) November = 3
g) May = 1
h) December = 0
i) Others = 0
16) Total number of U.S. landfalling MH's by month:
a) September = 44
b) August = 29
c) October = 16
d) July = 4
e) June = 2
f) November = 0
g) May = 0
h) December = 0
i) Others = 0
17) Most intense U.S. landfalling hurricane:
There are two different equations or meterological parameters that can be used to determine the strongest hurricane to make a U.S. landfall. One is based on a respective storms maximum sustained wind (MSW), while the other focuses on its minimum central barometric pressure. That being said, a hurricanes lowest central pressure at landfall is typically used to categorize a listing of the "most intense" TCs to strike the U.S. shoreline. With that in mind, Storm #2 (The Great Labor Day hurricane) of September 2, 1935 is the "most intense" TC known to have made a U.S. landfall. At landfall, it had a lowest central pressure reading of 892 mb.
The strongest known hurricane to make a U.S. landfall-relative to a storms MSW-is hurricane Camille of August 17, 1969. At landfall, it had a MSW of 190 mph.
18) Most TC's to strike the U.S. coastline in one season: The 1916, 2004, and 2005 H seasons each produced 9 TC's of tropical storm and/or hurricane intensity that made a direct strike on U.S. shores.
19) Most hurricanes to strike the U.S. coastline in one season: The 1886 H season produced the most U.S. landfalling TC's of hurricane intensity, with 7.
20) Most "major" hurricanes to strike the U.S. coastline in one season: The 2005 H season spawned 4 hurricanes that came ashore at MH intensity.
IMPORTANT NOTE: I will continue to provide additional "fascinating facts", relative to this particular era of hurricane history, as time permits. In the meantime, I hope each one of you have a great rest of the week!:)
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.
Comments will take a few seconds to appear.
50 ° F