Above: An aerial photograph taken and released by the Netherlands Ministry of Defence on September 6, 2017 shows the damage of Hurricane Irma in Philipsburg, on the Dutch Caribbean island of Sint Maarten. Image credit: Gerben Van Es/AFP/Getty Images.
A hurricane watch and a storm surge watch are up for Florida, from the Florida Keys to Jupiter Inlet, as Category 5 Hurricane Irma heads steadily west-northwest at 16 mph. Hurricane Irma delivered a devastating blow to the Lesser Antilles islands of Barbuda, Saint Barthelemy, Anguilla, and Saint Martin/Sint Maarten early Wednesday morning, then plowed directly though the British Virgin Islands on Wednesday afternoon, packing top winds of 185 mph during the entire rampage. Ten deaths are being blamed on the storm—six on St. Martin, two on Saint Barthelemy, one on Barbuda, and one on Anguilla. Damage photos from the islands show destruction characteristic of an EF4 tornado, and Irma’s 185 mph winds were indeed in the EF4 range of tornado wind speeds. The first island Irma hit, Barbuda (population 2,000) was judged “practically uninhabitable” by the prime minister after 90% of the buildings were damaged and 60% of the population was left homeless.
Irma slightly weaker on Thursday afternoon
Irma has been weakening slightly today. Peak winds fell from 185 mph at 1 am EDT Thursday to 180 mph at 2 am, then to 175 mph at 11 am—still solidly in the Category 5 range. The minimum pressure rose from 914 mb at midnight to 921 mb at 11 am. The most recent 11:56 am EDT eye report from the Hurricane Hunters found another bump up in pressure, to 923 mb. Infrared satellite loops show that the heavy thunderstorms that surround Irma’s eye are no longer as symmetrical or intense. This is particularly true on the storm’s west side, where some moderate wind shear near 10 knots is affecting the storm. Irma’s weakening is also likely due to an eyewall replacement cycle and interaction with the large island of Hispaniola, to the south of Irma. Total precipitable water loops from Thursday, for example, showed that Hispaniola was blocking the inflow of moisture from the south, reducing the amount of moisture in Irma’s core. Irma's circulation is also pulling air across and then down the high mountains of the island into the core of the hurricane; the downslope winds dry as they descend, injecting more dry air into Irma.
Despite this weakening, even 175 mph winds are catastrophic, and Irma will cause extreme damage in the Turks and Caicos Islands when it moves through on Thursday afternoon, and in the southeast Bahamas on Thursday night and Friday. Parts of central and eastern Cuba and northern Hispaniola may receive localized rainfall of up to 15 inches as Irma passes by. Storm surge levels could reach as high as 20 feet in the Southeastern and Central Bahamas, and as high as 10 feet on parts of Cuba's northern coast.
Irma maintained 185 mph maximum winds for an extraordinary 37 hours, the longest for any tropical cyclone on record, globally, in the satellite era (since 1966.) The previous record was Super Typhoon Haiyan in the Northwest Pacific, at 24 hours, according to Dr. Phil Klotzbach.
Figure 1. Official forecast for Irma from the National Hurricane Center as of 11 am EDT Thursday, September 7, 2017. image credit: NHC.
Forecast for Irma: Southeast U.S.
Irma is virtually certain to curve rightward this weekend and affect Florida and the southeast U.S. coast in some form or fashion, but there remains a nagging amount of uncertainty in the east-west location of Irma’s expected south-to-north track. This hinges on exactly where and when Irma will be turned northward by a trough now moving across the eastern U.S.. Where and when this turn occurs will be crucial for wind and surge effects in Southeast Florida. One consideration that is hard for computer models to incorporate is the indirect effects of the upper-level outflow associated with Hurricane Katia in the Bay of Campeche (see below). According to Brian Tang (University at Albany, SUNY). "Katia's showers and thunderstorms and the resulting upper-level outflow will likely affect the evolution of the southward end of the trough that will eventually turn Irma to the north."
The Wednesday night runs of our best track models—GFS, European, UKMET, and HWRF—covered a gamut of possibilities, from a SW Florida landfall (UKMET) to a coast-hugging Florida track starting near Miami (Euro) to a track just off Florida’s east coast (GFS). Over the long haul, the European tends to be the best performer of these three models, and it has done the best with Irma thus far, according to analyses from Brian Tang (University at Albany, SUNY). Keep in mind, though, that any of the three models can be the best for a given storm at any point in time (just as any high-quality football team might win on any given Sunday).
There was a westward shift in the model guidance at 12Z Thursday. The HWRF model now brings Irma over the highly populous corridor from Miami to Palm Beach, with the GFS only about 20-30 miles offshore, while the latest run of the Euro now brings Irma northward along the center of the Florida peninsula.
The most reliable track guidance of all is from NHC’s consensus models, which blends the various tracks in different ways to come up with a best-estimate outlook. These consensus models have been performing very well for Irma, and they have been very consistent in bringing Irma along Florida’s southeast coast (see Figure 2). NHC forecast tracks are typically close to the model consensus tracks. Remember that, on average, there is significant track error in forecasts beyond 3 days. As of 12Z Thursday, the NHC’s 5-day “cone of uncertainty” now encompasses the entire coastline of the Florida peninsula, extending north to southern North Carolina and well inland. If you are within the cone, you are at risk of impacts from Irma and need to prepare accordingly.
Figure 2. Consensus model output from 18Z (2:00 pm EDT) Wednesday, September 6, 2017, and 12Z (8:00 am EDT) Thursday, September 7. Consensus models are created by blending the top track models in various ways. These consensus tracks are based on model runs from the previous 6 or 12 hours. Image credit: Brian Tang, University at Albany/SUNY.
Figures 3 and 4 show guidance from the GFS and European ensemble runs from Wednesday night. In these ensembles, multiple parallel runs of the same model are carried out with slight variations in the starting-point data to simulate uncertainty.
Figure 3. The 20 track forecasts for Irma from the 0Z Thursday, September 7, 2017 GFS model ensemble forecast. About half of the GFS ensemble members bring Irma northward along the Florida east coast, and nearly all members produce a landfall in the southeast U.S. Image credit: CFAN.
Figure 4. The 0Z September 7, 2017, track forecast by the operational European model for Irma (red line, adjusted by CFAN using a proprietary technique that accounts for storm movement since 0Z Thursday), along with the track of the average of the 50 members of the European model ensemble (heavy black line), and the track forecasts from the “high probability cluster” (grey lines)—the five European model ensemble members that have performed best with Irma thus far. Only two of the five high-probability members move Irma northward along the Florida coast, but four of the five members produce a landfall in either Georgia or South Carolina.
It may take several more model cycles to increase the certainty in Irma’s track relative to the Florida coast, especially because it hinges on the timing and location of a right-hand turn that will not happen until Saturday. In the meantime, here are some things we do know:
—Irma is likely to be at least a Category 4 when it nears Southeast Florida on Sunday. There is nothing in the atmosphere on the large scale that would be expected to produce significant weakening before at least Saturday. As Irma moves just north of Hispaniola on Thursday, and offshore or just over the north coast of Cuba on Friday into Saturday, the interaction with land may disrupt Irma somewhat, as noted above. On the other hand, water temperatures will remain steady or slightly increase along Irma’s path, which is favorable for maintaining Irma’s strength. Irma may experience moderate wind shear by Sunday (around 15 knots), but it is safest to assume that Irma will maintain at least Category 4 intensity on its final approach to Florida, as reflected in NHC’s 11 am Thursday forecast. More significant weakening is expected from Sunday into Monday, because of increasing wind shear and land interaction. Irma is still predicted to approach the Southeast coast of GA/SC as a Category 3 hurricane.
—Regardless of its strength at final landfall, Irma has the potential to generate a catastrophic storm surge across the Southeast coast from northern Florida to South Carolina. Under almost all track scenarios for Irma, the vast amount of water being pushed northwest by the storm is expected to slam into the Southeast coast and produce a devastating storm surge. Even if Irma’s winds decrease as expected by Monday, this water will already be in motion, packed with tremendous momentum. From our Wednesday PM post on the storm surge potential: “Even areas up to a hundred miles to the north of where the center makes landfall could potentially see record storm surges. The area of most concern is the northern coast of Florida, the coast of Georgia, and the southern coast of South Carolina, due to the concave shape of the coast, which will act to funnel and concentrate the storm surge to ridiculous heights.” See our Wednesday post for more details.
Within the storm surge warning area along the Southeast Florida coast, water levels above ground (inundation) could reach 5 to 10 feet, according to official NHC storm surge guidance as of 11 am EDT Thursday. Storm surge expert Dr. Hal Needham has launched a storm surge website for Miami with details on the city's surge history and surge potential.
—Large parts of the Florida peninsula would experience damaging winds on any of the most likely tracks for Irma. Irma’s hurricane-force winds now extend up to 50 miles from its center, and that radius will increase over time. Even after Irma’s top winds weaken, the wind field will gradually expand over time. Tropical-storm-force winds extend up to 160 miles from Irma’s center, which would be enough to straddle the entire Florida peninsula. In 2005, Hurricane Wilma crossed the state at Category 2/3 strength from Cape Romano to Jupiter. Even though Wilma’s center was 60 miles north of Fort Lauderdale, the city’s downtown experienced significant damage, with many glass facades sheared off by high wind. Floridians should be prepared for massive loss of windows and glass facades with Irma, especially if it tracks near or over Florida’s East Coast. It's worth noting that Broward County—population 1.9 million—has not experienced a direct hit from a major hurricane since 1947. At that point, the county’s population was about 85,000.
Figure 5. Visible-wavelength satellite image of Hurricane Jose at 11:45 am EDT Thursday, September 7, 2017. Image credit: RAMMB / CIRA@CSU.
Hurricane watches for Northern Leeward Islands ahead of Jose
Recovery efforts in the northermost Leeward Islands, hard-hit by Irma, may be complicated by the approach of Hurricane Jose, which continues to gain strength about 700 miles east of St. Lucia. Jose’s top winds had increased to 105 mph as of 2 pm EDT Thursday, and conditions are quite favorable for continued strengthening: warm SSTs (around 29°C or 84°F), light to moderate wind shear (10 – 15 knots), and a moist atmosphere (mid-level relative humidities of 60 – 65%). On satellite, Jose is an increasingly impressive hurricane, with expansive outflow, a large, well-organized core of showers and thunderstorms (convection) and a gradually emerging eye.
Jose is predicted to become a major Category 3 hurricane by Friday, if not sooner. Unfortunately, Jose’s projected track has angled just far enough west to require Hurricane Watches for the northern Leeward Islands. The highest-probability members of the European model ensemble (0Z Thursday) track Jose just north of the islands on Saturday, but some members of the Euro and GFS ensembles do bring Jose over the northernmost Leewards.
Jose will likely slow or stall later next week over the open Northwest Atlantic south of Bermuda. There is huge model disagreement on Jose’s track beyond that point, in the 7- to 10-day time frame when there is very little accuracy in model track forecast.
Figure 6. Infrared-wavelength satellite image of Hurricane Katia at 12:32 pm EDT Thursday, September 7, 2017. Image credit: NASA/MSFC Earth Science Branch.
Katia prompts hurricane warnings for Mexican coast
Hurricane Warnings have been issued from Cabo Rojo to Laguna Verde, Mexico, for yet another Atlantic system, Hurricane Katia. On Thursday morning, Katia was in a holding pattern in the Bay of Campeche, remaining stationary about 200 miles north-northeast of Veracruz with top sustained winds steady at 80 mph. Though Katia does not appear especially well-organized on satellite, it has a compact mass of intense convection, with a fair amount of upper-level outflow, especially to its north and east.
Steering currents are expected to kick in and push Katia toward the coast of Mexico’s Veracruz state by Friday night. Katia is a small hurricane, which allows for more rapid strengthening and weakening based on the conditions in play. From Thursday up to landfall, Katia will be encountering very warm SSTs (30-31°C or 86-88°F) and light to moderate wind shear, with a dip in wind shear below 10 knots on Thursday night into Friday. Given the very moist atmosphere enveloping Katia (mid-level relative humidities of around 70%), it is quite possible Katia could become a major hurricane before landfall. The official NHC forecast as of 11 am EDT Thursday tops out at 105 mph.
Should we end up with three simultaneous major hurricanes at some point on Friday or early Saturday (a real possibility), it will be a first in Atlantic records going back to 1851, and certainly for the satellite era, going back to the 1970s. Our last set of three simultaneous hurricanes—2010’s Igor, Karl, and Julia—were all majors, but there was no overlap between their periods of Category 3 or stronger intensity. According to Phil Klotzbach, the only time on record with three simultaneous Category 2 storms in the Atlantic was in 1893. He noted: “Of course, there may have been some underestimates in storm intensity during the Grover Cleveland administration!”
Dr. Jeff Masters co-wrote this post.
Figure 7. As tweeted by Brian McNoldy on Thursday, the positioning of our current trio of Atlantic hurricanes on Thursday morning, September 7, 2017, was oddly similar to that of another trio—Karl, Igor, and Julia–on September 16, 2010. This was the last time we saw three simultaneous hurricanes in the Atlantic basin. Adding to the weirdness: both sets of names start with K, I, and J, going from west to east. Image credit: @BMcNoldy.