Matthew Slamming Hispaniola; Southeast U.S. Landfall a Growing Threat

Colossal amounts of rain are soaking Haiti and the Dominican Republic as Category 4 Hurricane Matthew heads for an encounter with the western end of Hispaniola. As of the National Hurricane Center’s update at 8:00 pm EDT Monday, Matthew was located about 200 miles south of Port-au-Prince, Haiti, moving just east of due north at 8 mph. Matthew’s top sustained winds were holding at 140 mph, solidly in the Category 4 range. Just before 8 PM EDT, a Hurricane Hunter flight found a minimum surface pressure of 934 millibars, down from the 940 mb reported in the 8 PM EDT advisory. However, surface winds measured by the SFMR instrument had not yet increased. Matthew may be completing an eyewall replacement cycle, with the original small eye decaying and a larger outer band taking over. Hurricane Hunters described the new eye as ragged, elliptical and kidney-shaped, about 17 by 30 miles across. Depending on how soon this cycle is completed, it’s possible that Matthew's winds could either increase or decrease by 5 - 10 mph prior to the storm making landfall in southwest Haiti around 8 am EDT Tuesday.


Figure 1.Hurricane Matthew’s eye was clearly evident in this enhanced infrared satellite image from 2345Z (7:45 pm EDT) Monday, October 3, 2016. Image credit: NOAA/NESDIS.

As of late Monday, Hurricane Warnings were in effect for Haiti, Cuba’s eastern provinces, and the southeastern and central Bahamas. Matthew’s center is likely to pass over or very near the western tip of Haiti. This will likely spare Jamaica from widespread winds above hurricane force, although very heavy rains can still be expected. On the other hand, Matthew’s track is close to a worst-case scenario for the beleagured nation of Haiti, as it will bring the hurricane’s more dangerous right-hand side across Hispaniola. Far southwest Haiti may experience Matthew’s small core of intense hurricane-force winds, and a much larger area of powerful south winds slamming against tall mountainsides will lead to phenomenal rains over Haiti as well as much of the Dominican Republic (DR). The rains will likely be enhanced by moisture associated with a persistent band of showers and thunderstorms that has flanked Matthew’s east side for days (see this discussion of the mysterious “blob” from Marshall Shepherd). This feature’s rapid movement toward Hispaniola has actually led to heavier rains thus far in the Dominican Republic than in Haiti. A personal weather station in Cabo Rojo, on the southern coast of the Dominican Republic near the border with Haiti, recorded 22.91” of rain in thirteen hours between 3 am and 4 pm on Monday, including a remarkable 5.33" in the hour from 6 am to 7 am. While PWS data is often suspect, these are believable rainfall amounts based on the satellite presentation of Matthew.



Figure 2. Flooding and coastal damage in Jacmel, Haiti on Monday afternoon, October 3, 2016. Jacmel is on the south coast of Haiti, about 25 miles southwest of the capital of Port-Au-Prince. Image credit: Ruth Chervil.

NHC warns that localized rainfall amounts could total 40” over southern Haiti and the southwestern DR, with widespread 15” - 25” amounts. Massive flooding and landslides are a virtual certainty, with the impacts especially severe on Haiti’s deforested landscape. As discussed in Jeff Masters’ post this morning on the hurricane history of Matthew’s targets, one of the few analogs for this northward-moving major hurricane is Hazel (1954), which killed more than 1000 people in Haiti.


Figure 3. Visible satellite image of 2145Z (5:45 pm EDT) Monday, October 3, 2016. Very intense rainbands associated with the mysterious “blob” east of Matthew can be seen at far right. Image credit: Naval Research Laboratory.

Models are in fairly close agreement in taking Matthew across the eastern tip of Cuba late Tuesday. This would keep most of the island on the hurricane’s weaker western side, but only a slight change in track could make a huge difference in the impacts on Cuba. On Wednesday into Thursday, Matthew will be charging across the central and northwest Bahamas. The intensity forecast for Matthew at this point is tough, with both high- and low-end possibilities. Matthew could be cut back to Category 2 or 3 strength by its brief passage over the high mountains of Haiti and/or Cuba. However, it will then have a chance to restrengthen over the deep, warm waters of the Bahamas. Matthew will be surrounded by a very moist atmosphere (relative humidities around 80%) and will pass over sea-surface temperatures close to record highs for early October. It's possible to envision Matthew striking the central Bahamas on Wednesday at anywhere from Category 2 to Category 4 strength. The offficial NHC forecast has Matthew at Category 3 strength through at least Friday.

A growing threat for the Southeast U.S.
A significant westward shift unfolded Monday in computer model guidance on Hurricane Matthew, and this has big implications for the hurricane’s potential impact on the U.S. East Coast. The main reason appears to be stronger ridging south of 98L and north of Matthew than earlier predicted, which may help to nudge Matthew far enough west for major impacts along the Southeast U.S. coast. Last night’s 50 ensemble runs from the 00Z Monday European model included a number of tracks making landfall along the U.S. East Coast. Most concerning was that all four members of the Euro “high-probability” cluster--the members that most closely match the operational run--depicted Matthew making landfall on Florida’s East Coast. Today’s 12Z Monday Euro ensemble continued along the same lines (see Figure 4 below).


Figure 4. Track forecasts from the four European model ensemble members [gray lines] that most closely match the operational run [red line] during the first 72 hours, starting at 12Z (8:00 am EDT) Monday, October 3, 2016. The red line is a version of the 12Z Monday operational model track that has been adjusted and calibrated using a proprietary technique to account for systemic model errors. The high-probability cluster (grey lines) perform better than other ensemble members at forecast times of five days and beyond. Image credit: Climate Forecast Applications Network (CFAN).


A track mirroring the Southeast coastline?
Other models have now moved into the Euro ensemble’s camp. The 12Z and 18Z GFS brought Matthew considerably closer to the Florida coast than earlier runs, with a projected landfall in northern South Carolina this weekend (see Figure 4 below) and a second landfall on Cape Cod less than 36 hours later. The 18Z GFS nudged Matthew’s track so that it arrives very near the central Florida coast on Thursday night, then hugs the coast all the way to Cape Hatteras, North Carolina, on Sunday morning. The 12Z and 18Z Monday GFS ensembles (GEFS) featured a majority of runs making landfall somewhere between Florida and North Carolina, a major shift west from previous GFS ensembles. The 12Z Monday run of the HWRF model tracked about 1 degree (roughly 60 miles) west of its previous two runs, now showing a potential landfall in eastern North Carolina by late Friday.

Obviously, a direct landfall from Matthew could inflict a devastating blow. The key variable in Matthew’s track later this week is how far north and west the hurricane moves before it begins the expected northeastward motion that will take it toward the Northeast U.S. and Canadian Maritimes. A gradual, arcing turn near the coast, as depicted in the 18Z GFS, would avoid a perpendicular, head-on landfall (the type that occured in New Jersey with Hurricane/Superstorm Sandy). This would act to reduce the intensity of storm surge at any given point. Such a track would also keep the mainland primarily on the less-intense western side of Matthew. On the other hand, such a track would bring hurricane-related impacts, including strong winds and heavy rains, to many millions of coastal residents. One particular concern is the risk of very heavy rain from eastern North Carolina into the Delmarva region. The last two weeks have already left totals exceeding 10” in some areas (see Figure 6 below). Any rains from Matthew would fall atop saturated soil, and even strong tropical-storm-force winds could lead to widespread tree uprooting. We can also expect major beach erosion as Matthew churns northeastward.

The governors of Florida and North Carolina have already declared states of emergency for all of FL and parts of central and eastern NC.


Figure 5. Total rainfall as observed by radar and rain gauges over the 14-day period ending at 12Z (8:00 am EDT) Monday, October 3, 2016. Image credit: NOAA/NWS Advanced Hydrologic Prediction Service.


Figure 6. The 50 forecasts from the 12Z Monday European (ECMWF) model ensemble (left) and the 20 forecasts from the 18Z Monday GFS model ensemble (right) agree that Matthew could make landfall almost anywhere from Florida to Nova Scotia late this week, over the weekend, or early next week. The GFS ensemble has been more tightly clustered than the ECMWF ensemble throughout Matthew’s life. Ensemble model runs are produced by running the same model for the same timeslice a number of times, with the starting-point conditions for each run varied randomly in order to mimic the uncertainty in our observations of the atmosphere. This produces a better sense of the future uncertainty in a given forecast. Image credit: Image credit: Climate Forecast Applications Network (CFAN).

98L: probably no steering influence on Matthew
An area of low pressure associated with a tropical wave (Invest 98L) is over the central tropical Atlantic several hundred miles north-northeast of the northern Lesser Antilles. In their 8 pm EDT Monday Tropical Weather Outlook, NHC gave this system 2-day and 5-day odds of development of 60%. High wind shear will likely keep any development of 98L rather limited, and we doubt 98L will be able to exert a significant steering influence on Matthew.

Northeast impacts: Still too soon to tell
The outlook for Matthew beyond the Southeast coast remains quite uncertain. Model solutions vary in how close Matthew’s northeast track might come to the Mid-Atlantic, Northeast, and New England, ranging from well inland to well offshore. We’ll also need to see how strong Matthew is by the time it passes the Outer Banks. A track just off the Southeast coast would keep Matthew over the warm waters of the Gulf Stream, whereas an inland track would lead to more rapid weakening.

The bottom line
Matthew poses an unusually widespread threat to the U.S. East Coast, with impacts possible from Florida to Maine. Residents should monitor the National Hurricane Center’s five-day outlooks (as shown below), as well as any local statements posted to the NHC’s webpage dedicated to Matthew.

We’ll be back with our next update by Tuesday morning at the latest. Meteorologist Steve Gregory is making regular updates on Matthew.

Bob Henson and Jeff Masters


Figure 7. WU depiction of official National Hurricane Center forecast for Matthew as of 5:00 pm EDT Monday, October 3, 2016.