Joaquin Hammers Bahamas; Future Track Still Uncertain

Category 3 Hurricane Joaquin is giving the Central Bahama Islands a ferocious pounding as the storm moves very slowly over the islands. An Air Force hurricane hunter aircraft made three penetrations of Joaquin's center on Thursday morning, and found top surface winds of 125 mph. The central pressure held steady at 942 mb between the first two passes at 7:47 and 9:23 am EDT, but dropped to 939 mb at 11:20 am, so Joaquin is still intensifying. The hurricane had a large 36-mile diameter eye that was fully closed in their second pass through. Joaquin took advantage of wind shear that had fallen to the moderate range, 10 - 20 knots, on Thursday morning. Visible and infrared satellite loops show that Joaquin is a moderate-sized hurricane with impressive organization, with a solid core of intense eyewall thunderstorms surrounding a clear eye. Upper level winds analyses from the University of Wisconsin show that the hurricane has maintained an impressive upper-level outflow channel to the southeast, which allowed the hurricane's rapid intensification over the past 24 hours. Ocean temperatures in the region remain a record-warm 30°C (86°F), but may start to cool due to Joaquin's slow motion.


Figure 1. GOES-13 image of Hurricane Joaquin over the Bahamas as seen on Thursday, October 1, 2015, at 10:30 am EDT. At the time, Joaquin had top winds of 125 mph. Image credit: NASA/GSFC.


Figure 2. Winds were rising across the Central Bahamas on Thursday morning, and were a brisk 39 mph, gusting to 58 mph, at 3:13 am EDT at a personal weather station on Exuma Island. Shewp's Webcam from Exuma Island on Thursday morning showed a darkening sky with heavy whitecapping of the waters, as Joaquin approached, but the webcam and weather station stopped reporting at 8:52 am.

Impact of Joaquin on the Bahamas
Joaquin's main threat to the Bahamas is likely to be wind damage. The 11 am Thursday Wind Probability Forecast from NHC gave the highest chances of hurricane-force winds of 69% to San Salvador Island (population 930). Hurricane-force winds are slightly less likely on Cat Island (population 1,500), to the northwest of San Salvador Island. Heavy rains of 10 - 15 inches in the Central Bahamas may also cause considerable flooding damage, as well as the large waves of the storm riding up on top of the expected 5 - 10' storm surge. Thursday morning satellite imagery showed that Joaquin had stalled out over the Central Bahamas; with only a slow motion expected for the next day, the islands will receive and extended pounding, increasing the odds of significant wind damage.


Figure 3. This Maximum Water Depth storm surge image for the Bahamas shows the worst-case inundation scenarios for a Category 3 hurricane with 120 mph winds, as predicted using dozens of runs of NOAA's SLOSH model. For example, if you are inland at an elevation of ten feet above mean sea level, and the combined storm surge and tide (the "storm tide") is fifteen feet at your location, the water depth image will show five feet of inundation. No single storm will be able to cause the level of flooding depicted in this image. The regions of the Bahamas most vulnerable to storm surge tend to lie on the southwest sides of the islands. Since Joaquin is approaching from the northeast, the storm's peak on-shore winds will be affecting the northeast sides of the islands, where deeper offshore waters tend not to allow larger storm surges to build. NHC is forecasting peak water levels (the depth of water above the high tide mark) of 5 - 10 feet from Joaquin in the Bahamas. See wunderground's storm surge pages for more storm surge info.

Outlook for Joaquin: Out to sea?
With favorable conditions for at least the next day (low wind shear and very warm sea-surface temperatures), Joaquin may yet intensify further. The 11 am EDT Thursday advisory from NHC pegs Joaquin’s top sustained winds at 125 mph, and the NHC outlook brings Joaquin to Category 4 status, with top sustained winds of 140 mph projected by Friday. Joaquin is located close to the region where Hurricane Andrew grew from Category 1 to Category 5 status during a year with very suppressed hurricane activity, 1992 (featuring an El Niño event during the first half of the year.) This serves as a reminder that the subtropics can be a worrisome breeding ground for strong hurricanes even during an El Niño year, when activity in the deep tropics tends to be suppressed. We can expect some fluctuation in strength if an eyewall replacement cycle takes hold over the next day or two, as is common after hurricanes go through a rapid intensification phase. Once Joaquin begins moving northward under the influence of stronger upper-level flow, we can expect its top sustained winds to eventually decrease while the size of its wind field increases. The waters are unusually warm across much of the Northwest Atlantic, which may help Joaquin sustain its strength longer than one would otherwise expect.

The track forecast for Joaquin remains low-confidence, although there was an important shift in the 00Z and 06Z Thursday computer-model guidance in favor of keeping Joaquin away from the U.S. East Coast. The global-scale GFS model, which had been predicting a North Carolina landfall for more than a day, shifted in its 00Z Thursday run to a track toward Long Island. The 06Z Thursday run of the GFS showed an even more dramatic shift eastward, with Joaquin hugging the Nova Scotia coastline. Members of the GFS ensemble also reflected this shift, with most but not all of the 06Z GEFS members showing an offshore track. (Ensembles are produced by running a model many times, each with slightly different initial conditions to represent uncertainty in the atmosphere’s starting point.) The UKMET also shifted significantly eastward, moving from a North Carolina landfall in its 12Z Wednesday run to a Cape Cod brushing in its 00Z Thursday run. The high-resolution HWRF and GFDL models stuck to their guns, with their 06Z Thursday runs continuing to depict landfall in North Carolina or Virginia.

If the trend toward an offshore track holds up in Thursday’s model guidance, kudos must go to the ECMWF model. Its operational run has consistently called for a track well away from the U.S. East Coast, as was the case at 00Z Thursday. Just as significant, most of the ECMWF ensemble members (about 40 out of 50) showed an offshore track in the 00Z Thursday runs, whereas a large part of the ensemble had previously shown of a U.S. landfall. Analyses of the ECMWF ensemble for 12Z Wednesday and 00Z Thursday indicate that the ensemble members who did the best in the first few hours of the forecast were consistently taking Joaquin offshore (see Figure 6).



Figure 4. The latest runs from our two top models for forecasting hurricane tracks: the 8 pm EDT Wednesday September 30, 2015 (00Z Thursday) run of the European model (left), and the 2 am EDT October 1 (06Z) run of the GFS model (right) both took Joaquin on a path out to sea that misses the U.S. coast, but were still very far apart. Image credit: wundermap with the "Model Data" layer turned on.







Figure 5. The ensemble runs of our two top models for forecasting hurricane tracks, both run at 8 pm EDT Wednesday September 30, 2015 (00Z Thursday). The 50 members of the European model ensemble (top) had only about 10 of its 50 members that showed a U.S. landfall, while about 10 of the 20 members of the GFS model ensemble (bottom) did so. Compared to the runs done 24 hours previous, the European ensembles had shifted considerably to the east, away from the U.S., with the GFS ensemble members less so. Ensemble runs take the operational version of the model and run it at lower resolution with slightly different initial conditions, to generate an "ensemble" of possible forecasts. 





Figure 6. The European model ensemble run at 8 pm EDT Wednesday September 30, 2015 (00Z Thursday, October 1) had four of its 50 members (grey lines) that tracked the movement of Joaquin exceptionally well during the period 00Z - 12Z October 1. All of these four members had tracks for Joaquin that missed the U.S., with two of them hitting Canada. The operational (high-resolution) version of the European model is shown in red. Image taken from a custom software package used by TWC.

The ECMWF model is known for its high-quality representation of atmospheric physics and its ability to smoothly incorporate data from a variety of sources. The model is not infallible; back in January, it famously and erroneously predicted that Manhattan would get walloped by several feet of snow, the result of it having placed an upper-level low about 100 miles too far to the west. However, in cases of model disagreement, the ECMWF is often the first to pick up on subtle large-scale features that turn out to be crucial in steering a hurricane or other storm. This was the case during 2012’s Hurricane Sandy, when the ECWMF was ahead of all models in depicting the rare leftward hook into New Jersey that Sandy ended up taking. Two key factors at play with Joaquin are the upper-level low cutting off over the Southeast U.S. and another upper low taking shape well northeast of Joaquin. Most models had projected that the Southeast low would pull Joaquin into its northeast side, a la Sandy, whereas the ECMWF and other models now appear to be reckoning that the upper low in the Atlantic will play a larger role in steering Joaquin.



Figure 7. This WunderMap image shows the GFS-analyzed steering flow at 200 mb (about 40,000 feet) at 06Z (2:00 am EDT) Thursday, October 1, 2015. Joaquin's future track is being shaped by an upper-level low that will cut off from a sharp trough now in the eastern United States (A) and by another upper low developing at the base of another sharp trough in the north central Atlantic (B).


Given the interplay between these two features, it is still too soon to confidently project that Joaquin will remain offshore, but it is fair to say that the ominous HWRF and GFDL tracks are now lower-probability, high-impact possibilities. The strength of Joaquin and the residual disagreement among models calls for continued keen vigilance and careful analysis. Another caveat is that the onshore and offshore forecast tracks do not diverge a great deal until after Friday, so quick action would be needed if the lower-probability onshore solution turned out to be correct.

A key experimental tool for better forecasts missing for Joaquin
One potential aid to making better hurricane track and intensity forecasts is the use of real-time radar data from NOAA's two P-3 hurricane research aircraft. Over the past two years, these aircraft have flown numerous missions into Atlantic hurricanes and tropical storms, sending back real-time radar data that was ingested into the HWRF model, one of our top models for predicting both hurricane tracks and intensities. This real-time data was shown to measurably improve the forecasts from this model. Unfortunately, both NOAA P-3 aircraft are grounded this week for maintenance issues. One aircraft has been undergoing a months-long process to have new wings put on, leaving just one P-3 for this year's hurricane season. Unfortunately, last Friday, de-lamination of that plane's lower fuselage radome, which was deep and too broad to fix at the Aircraft Operation Center's base in Tampa, was discovered. The shell has been trucked to Jacksonville for repair, and the repair will not be done until Friday at the earliest. However, NOAA's jet has been flying upper-level dropsonde missions around the clock, and data from these missions has been getting fed into the models for Joaquin.

Regardless of Joaquin's path, a potentially devastating rain/flood/surge event
Even if Joaquin does stay offshore, a very large pressure gradient between it and a surface high far to the north will keep a broad easterly fetch of wind heading into the U.S. East Coast, leading to a prolonged bout of coastal flooding and erosion over the next several days. Storm-surge expert Hal Needham emphasized the rarity of the situation in a blog post on Thursday morning. “The duration of this wind event is absolutely mind-boggling,” says Needham. Strong, sustained onshore winds (more than 20 mph) could be affecting the mid-Atlantic coast for more than 96 solid hours, regardless of Joaquin’s track. High water will be present for as many as 10 high tides over several days, increasing the risk of erosion and flooding along the coast as well as up to a few miles inland. “This developing situation is truly historic and has not been observed in the modern history of the Mid-Atlantic Coast,” says Needham.

Complicating matters even further, a potentially destructive multi-day bout of heavy rain and inland flooding is on tap, focused on the southern Appalachians and nearby coastal plains, as the cutoff Southeast low continues to pump rich tropical moisture (with at least some contribution from Joaquin’s circulation) over a preexisting frontal zone. Both the ECMWF and GFS model solutions lead to moisture inflow at the 850 mb level (about a mile above sea level) that is close to unprecedented amounts—“off the charts,” as NOAA’s Weather Prediction Center (WPC) put it in a Thursday morning discussion.



Figure 8. 5-day predicted rainfall amounts from 12Z (8 am EDT) Thursday, October 1, to Tuesday, October 6. Image credit: NWS Weather Prediction Center.

Widespread rainfall in recent days (2” – 4” in many areas) has saturated the ground in many areas, which will add to the flood risk. The focus of the heaviest rain may shift from western GA/SC/NC/VA toward the mid-Atlantic toward Sunday and Monday, depending in large part on interactions between the Southeastern upper low and Joaquin. The 5-day rainfall amounts predicted by WPC are astounding: most of the region from northeast Georgia to New Jersey is projected to receive at least 5”, with 15-20” predicted across the bulk of South Carolina. Local amounts are often substantially greater than these large-scale predictions.

The bottom line: regardless of Joaquin’s track, a large and populous part of the United States is in for what could be historic rainfall and a very serious flooding risk.

We’ll have another update later today.

Jeff Masters and Bob Henson