On 3 September a westward moving tropical wave showing no signs of organization emerged from the northwest African coast over the open waters of the North Atlantic. During the next several days a broad low pressure area associated with the tropical wave developed over the Atlantic midway between Africa and the Lesser Antilles with ship reports indicating the circulation extended nearly to the equator. An organized circulation center was not evident on satellite imagery until the system was approaching the Windward Islands on 8 September. On 9 September the low was classified by satellite as the 12th tropical depression of the 1988 hurricane season when it was located about 400 miles east of Barbados. See Figure 1.
The depression moved on a west northwest course around 15 knots with satellite and reconnaissance reports indicating that it had attained tropical storm strength as it moved over the Lesser Antilles the afternoon of 9 September. Tropical Storm Gilbert rapidly strengthened on 10 September and was classified as a hurricane by that evening.
Gilbert continued to strengthen as it brushed the south coast of Hispaniola, then passed directly over the island of Jamaica as a category 3 hurricane on the Saffir-Simpson Scale with sustained winds of 110 knots and a minimum pressure of 960 mb (28.35 inches). Following the passage of the center across Jamaica, Gilbert went through a remarkable intensification period with the pressure falling from 960 mb (28.35 inches) to 888 mb (26 . 22 inches) in 24 hours. The 888 mb pressure was observed by a NOAA plane near 19.5N and 83.3W at 2152 UTC on 13 September and is the lowest sea level pressure ever recorded in the Western Hemisphere. At the same time the plane measured 160 knot sustained winds at 10, 000 feet with a peak wind of 173 knots. During the period of rapid strengthening Gilbert continued on a west northwest course at 15 knots under the influence of a persistent high pressure to the north. The center passed a short distance southeast of Grand Cayman Island with a wind gust of 136 knots recorded there at 1900 UTC. See Table 2.
The center of Gilbert crossed over the northeast portion of the Yucatan Peninsula on 14 September as a category 5 hurricane, the first category 5 hurricane to first landfall in the western hemisphere since Camille in 1969. The hurricane lost strength quite rapidly as the eye moved across the Yucatan Peninsula with the minimum pressure rising to around 950 mb (28.05 inches) by the time the center emerged over the southwest Gulf of Mexico. Gilbert continued on the same west northwest course around 15 knots across the Gulf and reached the northeast Mexican coast just north of the town of La Pesca around 2200 UTC on 16 September as a category 3 hurricane.
The center of the weakening storm passed south of Monterrey, Mexico on 17 September then turned toward the north and moved across western Texas and into Oklahoma as a heavy rain storm on 18 September. It finally merged with a developing frontal low pressure system over Missouri on 19 September.
The synoptic six-hourly positions by latitude and longitude, along with lowest sea level pressure, maximum winds and classification can be found in Table 1. The track of Hurricane Gilbert with 0000 UTC and 1200 UTC positions is shown in See Figure 1.
Weather observations in the three landfall areas were difficult to obtain. The only weather office to measure the maximum winds in Gilbert was Kingston, Jamaica which reported sustained winds of 101 knots with gusts to 122 knots. An unofficial report of 105 knots with gusts to 128 knots was measured by a ham radio operator located 15 miles northeast of Kingston. There were no official reports of 118xinun winds near the center in the landfall areas of Mexico. Along the lower Texas coast winds in the Brownsville area gusted to 58 knots with gusts to 72 knots reported near Port Isabel by an observer with a truck 110unted anemometer. A gust of 53 knots was observed by the National Weather Service in Corpus Christi. See Table 2.
Likewise no minimum pressure readings in the eye of Gilbert were reported as it moved ashore in Mexico. The weather office in Kingston had a minimum pressure of 964.8 mb in the eye. The pressure was estimated to be near 900 mb when the eye of Gilbert moved over Cozumel, Mexico. See Table 1.
Torrential rains accompanied the hurricane with between 5 and 10 inches falling over the coastal sections and much greater amounts in the mountainous areas of Jamaica and Mexico. Massive flooding in the Monterrey, Mexico area caused many of the deaths attributed to Gilbert. Rains of 2 to 4 inches fell across south Texas with local amounts of more than 8 inches observed near Aransas Pass.
A 15 to 20 foot storm surge (rise of the level of the sea) likely occurred along the immediate coast near and just to the north of where the center moved inland over the northeast Yucatan peninsula. A surge of 8 to 13 feet struck the coast of eastern Mexico near and just to the north of landfall. There was a report of a 9 foot surge topped by 30 foot waves on the northeast coast of Jamaica. Tides of 3 to 5 feet above normal were reported along the Texas coast with a number of low-lying roads under water. There was considerable beach erosion on Padre Island. See Table 2 for details of meteorological statistics.
At least 29 tornadoes were observed across south Texas with most of the damage occurring in the San Antonio area. This total included 10 in the lower Rio Grande Valley, 5 around Corpus Christi and at least a half a dozen in the San Antonio area. The greatest destruction there occurred at the Air Logistics Center on Kelly Air Force base where a number of large storage hangers were destroyed with damage estimated near 22 million dollars. A tornado in Del Rio destroyed 15 homes and damaged 50 others with damage estimated to be nearly 2 million dollars.
Casualty and Damage Statistics
The death toll from Gilbert was reported to be 318 people. Several countries in the Caribbean reported deaths, with the death toll of 202 from Mexico being by far the largest. The only deaths in the United States were due to tornadoes in San Antonio, Texas with 3 people killed in that city. The latest death statistics by country: Mexico 202, Jamaica 45, Haiti 30, Guatemala 12, Honduras 12, Dominican Republic 5, Venezuela 5, United States 3, Costa Rica 2 and Nicaragua 2.
The estimated damage throughout the area was probably near five billion dollars. Jamaica had nearly two billion dollars damage and in Mexico damage was estimated between one and two billion dollars. Reports from the Mexican government indicate that more than sixty thousand homes were destroyed in that country. The most extensive damage over the United States was in San Antonio where damage caused by a number of tornadoes around to 30 million dollars. Elsewhere over south Texas damage was confined to beach erosion along Padre Island and in the Lower Rio Grande where high winds and a few tornadoes downed a number of trees, signs and power lines. Total damage in the United States was estimated to be between 40 and 50 million dollars.
Forecast and Warning Critique
Warnings generally conformed with the track of Gilbert over the Caribbean. Due to the possible strengthening of the system into a tropical storm as it approached the Lesser Antilles, a tropical storm watch could have been issued for a portion of the Windward Islands. Over the Gulf of Mexico hurricane warnings were issued from Tampico, Mexico to Port O'Connor, Texas. In retrospect, this warning area is probably excessive. However, the size and strength of the hurricane, plus the uncertainty of the forecast track, made this size Warning necessary. North of Port O'Connor a hurricane watch was issued for the remainder of the Texas coast. This was reasonable since many of the guidance products from NHC and NMC forecast a turn toward the north as the hurricane approached the Texas coast. See Table 3 for a summary of the Warnings. The time between when a hurricane warning was issued until the center crossed the coast was about 22 hours for Jamaica, 26 hours for the Yucatan peninsula and 34 hours for eastern Mexico. Hurricane watches were issued 9 to 15 hours prior to the warnings for these areas. Of course weather conditions deteriorated, well before the “center” arrived, but these lead times generally provided sufficient times for preparations. Similar warning lead times ware provided for other warned areas.
Most of the NHC objective forecast techniques handled Gilbert's track rather well during the period the hurricane was in the Caribbean. The NHC83 made an excellent 72-hour forecast when Gilbert was just east of Jamaica. The model forecast landfall over northeastern Yucatan within a few hours of the actual time the eye moved over the Yucatan peninsula. Over the Gulf of Mexico most of the techniques forecast a turn toward the northwest with landfall on the Texas coast. After the hurricane moved into the southwest Gulf of Mexico the NHC Cliper model was best at forecasting the point of landfall. On one occasion the NMC QLM hurricane model forecast did rather wall through 48 hours, but forecast a turn toward the southwest in the last 24 hour period.