SR/SSD 98-16

Technical Attachment

A Historical Study of Tropical Storms and
Hurricanes that have affected Southwest
Louisiana and Southeast Texas

David Mark Roth
National Weather Service
Lake Charles, LA

This work includes storms documented from 1886 through 1997 and is intended to add valuable knowledge about historical tropical cyclone activity in the County Warning Area (CWA) served by the Lake Charles National Weather Service Office. It should provide an excellent reference for those who live in Louisiana and Texas and those interested in the area's climatology.

Area of Study

The principal region of study includes the Lake Charles National Weather Service Office CWA; the area for which the office is responsible for issuing warnings, forecasts, advisories, and statements. In southeast Texas, this includes the counties of Jefferson, Orange, Jasper, Newton, Tyler, and Hardin. In Louisiana, this includes the parishes of Cameron, Vermilion, Iberia, St. Mary, Upper and Lower St. Martin, Calcasieu, Jefferson Davis, Acadia, Lafayette, Beauregard, Allen, Evangeline, St. Landry, Vernon, Rapides, and Avoyelles.

In addition, places outside the Lake Charles CWA were included in this study as well since tropical cyclones typically affect regions 70 miles west and 100 miles east of the center. There are exceptions, of course. For example, Hurricane Carla of 1961 made landfall near Corpus Christi and produced storm surges on the order of 5 feet as far away as Lake Pontchartrain. This storm was not included in the survey because it made landfall over 200 miles away from Lake Charles.

To account for the above, a radius of 150 nautical miles from Lake Charles was chosen. All tropical cyclones that have reached tropical storm strength and have passed through this area were included in this paper. This region encompasses the metropolitan areas of Houston and Beaumont in Texas and the cities of Lake Charles, Lafayette, Alexandria, Baton Rouge, and Shreveport in Louisiana (see Figure 1).

Rough Climatology

During the 112 seasons studied, 71 tropical cyclones passed within the area in question. Of these 71 tropical storms, 34 were hurricanes, and eight of these hurricanes were considered major (Category 3 to 5 on the Saffir-Simpson Scale). This gives a frequency of one tropical storm through the area every 1.6 years, one hurricane every 3.3 years, and a major hurricane every 14 years. The longest the area went with a break in tropical storm activity was seven years (1905-1912) and fifteen years was the longest interim without hurricane activity (1900-1915). In contrast, four of the seasons studied (1932, 1957, 1985, and 1989) had three tropical cyclones enter the area. Table 1 shows hurricanes which have entered this region.

Hurricanes that have made landfall within 150 miles of Lake Charles


Number 9 - Oct 12, 1886 Number 5 - Aug. 27, 1934 Fern - Sept. 10, 1971
Number 3 - July 05, 1888 Number 2 - Aug. 15, 1938 Edith - Sept. 16, 1971
Number 1 - July 05, 1891 Number 2 - Aug. 07, 1940 Carmen - Sept. 08, 1974
Number 8 - Sept. 07, 1893 Number 1 - Aug. 21, 1942 Babe - Sept. 05, 1977
Number 2 - Sept. 12, 1897 Number 1 - July 27, 1943 Alicia - Aug. 18, 1983
Galveston - Sept. 09, 1900 Number 4 - Sept. 19, 1947 Danny - Aug. 15, 1985
Number 2 - Aug. 17, 1915 Audrey - June 27, 1957 Juan - Oct. 28, 1985
Number 1 - Aug. 06, 1918 Debra - July 25, 1959 Bonnie - Aug. 26, 1986
Number 2 - Sept. 22, 1920 Cindy - Sept. 17, 1963 Chantal - Aug. 01, 1989
Number 3 - Oct. 16, 1923 Hilda - Oct. 03, 1964 Jerry - Oct. 16, 1989
Number 3 - Aug. 25, 1926 Betsy - Sept. 10, 1965 Andrew - Aug. 26, 1992
Number 2 - Aug. 14, 1932

Table 1. Dates indicate closest approach/initial landfall of hurricane

Origins of the Tropical Cyclones

Of the tropical cyclones that affected the designated area in Louisiana and Texas, 38 originated in the Gulf of Mexico, 16 in the Carribean Sea, 15 in the Atlantic, and 2 in the East Pacific Ocean. Since 1886 the earliest storm to arrive in the region was on May 30th in 1959 (minimal tropical storm Arlene), and the latest storm affected the area on October 28th in 1985 (hurricane Juan).

In May and June, the region of genesis was the Gulf of Mexico. When systems form in the Gulf of Mexico, half approach the Lake Charles area from the southeast and the other half from due south; either way, areas around Lake Charles may have only a couple of days to prepare for the impending storm. Even worse, many of the storms in the survey were intensifying rapidly as they made landfall, such as Audrey in 1957. This makes preparation at the start of the hurricane season not only a good idea, but a necessity. A storm forms in the Gulf and affects the region about once every 16 years in the May 30-June 30 period.

By July and August, the areas where storms form spreads into the Carribean, southwest and central Atlantic. This normally gives the region four to five days to prepare. A storm has affected the region, on average, once every fourteen years in July and once every eight years in August; most storms moved in from the southeast.

In September, development is fairly evenly split between the Atlantic and Gulf of Mexico basins, while the Carribean lagged behind. Storms moved in from the southeast and south; one storm affected the area about every four years.

In October, the favored area of genesis shifted back to the western Carribean, with the Gulf of Mexico and eastern Pacific lagging behind (both storms that formed in the eastern Pacific then moved into the Gulf of Mexico occurred during October). This would give the region three days to prepare for the storm. Storms moved in almost exclusively from due south during the month and affected the area once every eleven years.

Only four cyclones of the 71 which have affected the Lake Charles area formed in the Atlantic north of the 20th parallel, two of which formed north of the 30th parallel. One of the tropical cyclones that entered Louisiana initially made landfall in Georgia and moved westward through the southeastern United States just inland from the Gulf Coast, never fully emerging into the Gulf of Mexico. This fact means that even though a brewing storm may seem far away off the Carolina coast, there remains a remote possibility that it may yet affect the northwestern Gulf Coast. This shows why it is important to take a good look and pay close attention to systems most anywhere in the Atlantic basin; they may eventually become a real threat to the Louisiana and Texas shoreline.

Date of Storm Entrance

The period from May 21st through October 31st was divided into periods of 10 or 11 days, depending upon the length of the month. Despite May, July, and October having 31 days, it was pleasing to see no glaring effects of the 11 day periods on the tropical cyclone distribution. There appears to be one major peak in tropical storm activity during the traditional peak of the hurricane season (September 11-20), and two smaller peaks in mid- to late June and in early August.

Figure 3 shows how dangerous September can be for the area of study. In hurricane activity, there appears to be a small early season peak in late June and early July, and then the major peak occurs in August and early September, skewed just left of the tropical storm peak.

Memorable Storms in the Lake Charles CWA

The Rainmakers.

In order to see which storms had a significant rainfall contribution to the area, rainfall data from Beaumont/Port Arthur and Lake Charles were compiled from months when tropical cyclones were in the region. Months that showed rain totals greater than six inches were further investigated. Those months were checked for individual storm totals. For Lake Charles, data from 1893 to present were used. For Beaumont/Port Arthur, Jefferson County Airport (just north of Port Arthur) was used as the source of data from 1946 to present and prior to that, the city of Beaumont was used back to 1901. Southeast Texas rain events during tropical cyclones tend to be slightly greater than those that have occurred at Lake Charles during this time span. Storms which produced at least six inches of rain are listed in Table 2 in order of decreasing rainfall.

It is worth noting that the slower a tropical cyclone moves, the more rainfall it is likely to produce. Allison and Claudette, as tropical storms, produced greater than 30 and 40 inches of rain respectively in sections of southeast Texas, causing significant, and in the case of Claudette, unprecedented river flooding about a week later. Both cyclones essentially stalled over a period of two days north of Houston in order to produce these excessive rains.

Table 2a: Tropical Cyclones and Their Associated Deluges for Beaumont/Port Arthur Table 2b: Tropical Cyclones and Their Associated Deluges for Lake Charles
Storm Year Duration of Rain Rain Total Storm Year Duration of Rain Rain Total
Cindy 1963 4 days 17.52" Allison 1989 5 days 14.05"
Danielle 1980 2 days 17.17" Number 2 1940 2 days 11.65"
Allison 1989 7 days 15.96" Number 2 1915 4 days 10.73"
Number 2 1915 5 days 15.47" Delia 1973 3 1/6 days 9.99"
Claudette 1979 4 days 12.83" Juan 1985 4 2/3 days 9.92"
Number 10 1949 3 days 10.82" Claudette 1979 2 3/8 days 9.18"
Number 1 1943 1 day 10.34" Debra 1959 2 1/3 days 7.78"
Juan 1985 6 days 10.25" Audrey 1957 22 hours 6.90"
Debra 1959 3 days 7.69" Number 10 1949 3 days 6.10"
Number 2 1940 3 days 7.07"

In addition, a developing tropical low moved into extreme southwestern Louisiana in late August of 1962 and caused 22 inches of rain at the Sabine Wildlife Refuge, along with several tornadoes to the east of its point of landfall. This shows that even if a system is weak, it may become more than a mere nuisance to the area. Floods after Claudette and Allison caused headaches for as long as two weeks after the storms had moved away from the region.

Table 3: Hurricanes and Their Fatalities

Storm Year Deaths
Number 8 1893 2,000
Galveston Hurricane 1900 8,000
Number 2 1915 275
Number 1 1918 34
Number 3 1926 25
Number 2 1932 40
Number 4 1947 51
Audrey 1957 390
Hilda 1964 39
Betsy 1965 75


Table 3 lists chronologically those storms which while in the area of study resulted in losses of life of 25 or more. The table shows that as recently as 1965 exceptional storms have resulted in many deaths. The hurricanes of 1893 and 1900 are legendary. It is worth noting, however, that the last storm to cause such a loss was Betsy in 1965. This can be attributed to better preparation and to better warnings, especially considering that three of the eight major hurricanes that have struck the region (Carmen, Alicia and Andrew) occurred after Betsy.

Average Lead Time for Storms in the Area of Study

Data from the 71 storms were examined in greater detail to determine average motion prior to landfall, specifically while in the Gulf of Mexico. The most common direction for a storm to approach from was the southeast, with 13 storms entering from that region. The storms from the southwest moved fastest, as they were being picked up by the mid-latitude westerlies; average motion was 12 mph. Storms from the east moved slowest, where the average motion was about 9 mph. Figure 4 was produced with an average motion of 8.9 knots (10.2 mph) in mind; this was the average motion of all the storms in the sample.

Note that the radials are labeled in hours, not nautical miles; each radial is equal to about 214 nautical miles. This was done to give the reader some rough estimate, at a glance, of how long it would take a "typical storm" to move to Lake Charles. Note that storms in the Gulf do not give much advance notice prior to landfall. On average, they strike the Lake Charles area with 48 hours. Storms start adversely affecting the area as much as 24 hours before landfall.

It is worth pointing out that storms over the southern Gulf of Mexico (Bay of Campeche) can meander for days. This is because sometimes the Mexican Plateau seems to divert the steering winds for tropical cyclones to the north, causing some storms to stall for days; recent examples include Opal and Roxanne from the 1995 season. Most of these storms, however, either move into Mexico or eventually accelerate off to the northeast, leaving the studied area in the clear.

Possible Effects of ENSO (El Niño/Southern Oscillation)

Years of moderate to strong warm ENSO events were checked for any possible relationship with tropical cyclones affecting Louisiana and southeast Texas. These warm events seemed to have no significant correlation to an increase or decrease in tropical storms and hurricanes entering the area; it has been recently discovered though that it does make a difference further down the Texas coast.

Unexpectedly, ENSO events seemed to have a strong correlation to major hurricanes in this area. Five of the eight major hurricanes struck the region during the middle of moderate to strong El Niño events. If one expands the criteria to within one year of a moderate ENSO seven of eight become included. Further study is needed to totally resolve this issue.


The following is a list of definitions of terms used in this paper.

Tropical Cyclone - A low pressure system which develops over warm ocean surfaces and has a warm core, has a closed surface circulation, and maximum sustained one-minute winds of 34 knots (39 mph/17 m/s) or greater.

CWA - County warning area; area in which a National Weather Service Forecast Office issues forecasts and warnings. For Lake Charles, this includes 17 parishes in Louisiana and six counties in Southeast Texas.

Tropical Storm - A tropical cyclone with maximum sustained one-minute winds of less than 64 knots (74 mph or 37 m/s).

Hurricane - Tropical cyclone with maximum sustained winds of 64 knots (74 mph or 37 m/s) or greater. These storms generate life-threatening storm surges.

Major hurricane - Hurricanes with maximum sustained one-minute winds of 100 knots (115 mph or 58 m/s) or greater. In addition to high storm surges, these storms can create catastrophic damage from extremely high winds.

Saffir-Simpson Scale - Has categories of 1 through 5 to rate maximum sustained one-minute winds, central pressure, and storm surge to a hurricane's related damage. Five is the highest category with winds in excess of 130 knots (155 mph or 78 m/s). In the western Pacific Ocean, category 5 hurricanes are also known as super typhoons.

El Niño/Southern Oscillation (ENSO) - Global climatic shifts associated with, and in the case of El Niño, significant warming of the waters of the eastern Pacific Ocean west of Peru about once every 5 to 7 years. Amongst the global consequences of El Niño (warm ENSO events) include suppressed tropical cyclone activity in the Atlantic basin, floods along the West Coast of the United States, and droughts in Australia and Indonesia. During La Niñas or El Viejos (cold ENSO events), the reverse is true.

Acknowledgments. The author wishes to thank Steve Rinard at the NWS office in Lake Charles along with Mike Kosiara and Paul Trotter at the NWS office in Slidell, and James Lushine of the NWS office in Miami for their input in the paper. An examination of a possible El Niño correlation would not have been possible if not for Ricardo Correa-Torres of Florida State University providing information on years of moderate and strong El Niño/warm ENSO events. The Storm program was used to make Fig. 2.


Hebert, Paul J., Jerry D. Jarrell, Max Mayfield: The Deadliest, Costliest, and Most Intense Hurricanes of This Century (And Other Frequently Asked Hurricane Facts), National Oceanic And Atmospheric Administration, NOAA Technical Memorandum NWS NHC-31, 1995.

Kelly, Joseph T., Alice R. Kelley, Orrin J. Pilkey, Sr., Albert A. Clark: Living with the Louisiana Shore. Duke University Press: Durham, 1984.

Newmann, C.J., B.R. Jarvinen, A.C. Pike, 1987: Tropical Cyclones of the North Atlantic Ocean 1871-1992 (plus inserts to 1996), National Climatic Data Center, Historical Climatology Series 6-2, 186 pp.

Rasmusson, Eugene M., Thomas Carpenter: "Variations in Tropical Sea Surface Temperature and Surface Wind Fields Associated With the Southern Oscillation/El Niño," Monthly Weather Review, 1982. 110:354-384.