Hurricane Warnings for the Bahamas From Joaquin; Threat to U.S. East Coast Grows

Figure 1. Latest satellite image of Joaquin.

The U.S. outlook for Joaquin
A hurricane watch could be required for portions of the U.S. East Coast as early as Thursday night. The forecast for Joaquin is very complex, and the confidence in both the intensity and track forecast for the storm is low. Joaquin is trapped to the south of a high pressure system whose clockwise flow will push the cyclone very slowly to the southwest or west-southwest at about 3 – 6 mph. As the storm progresses to the southwest, the strong upper-level winds out of the north currently bringing high wind shear of 20 knots will gradually decrease, continuing to allow Joaquin to strengthen. The 8 am EDT Wednesday run of the SHIPS model predicted that wind shear over Joaquin would fall to the moderate range, 10 – 20 knots, on Thursday and Friday. These conditions should allow Joaquin to intensify to a Category 2 hurricane by Thursday. As Joaquin progresses to the west, the storm will also increasingly “feel” the steering influence of a strong upper-level trough of low pressure situated over the Eastern United States on Friday, and begin to turn north. These winds may also open up another upper-level outflow channel to the northwest of Joaquin on Friday, potentially allowing the storm to intensify to Category 3 strength. However, as Joaquin gets closer to this trough, its winds will bring high wind shear of 20+ knots, likely halting the intensification process and causing weakening by Sunday.

Figure 2. Our two top models for forecasting hurricane tracks, both run at 8 pm EDT Tuesday September 29, 2015 (00Z Wednesday) , came up with two very different solutions for the path of Joaquin. The GFS model showed Joaquin making landfall in Virginia, while the European model took the storm to the northeast out to sea without hitting the U.S. Image credit: wundermap with the “Model Data” layer turned on.

Figure 3. The ensemble runs of our two top models for forecasting hurricane tracks, both run at 8 pm EDT Tuesday September 29, 2015 (00Z Wednesday). The 50 members of the European model ensemble (top) and the 20 members of the GFS model ensemble (bottom) both had numerous model runs that took Joaquin into U.S. East Coast, and ones that missed the U.S. coast entirely. 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. The operation high-resolution (and presumably best-guess) forecast for the models is shown in red. The European model ensemble had four members that tracked the movement of Joaquin exceptionally well during the previous 12 hours; three of those four members had tracks for Joaquin that missed the U.S., and one that hit the coast near New York City. Image taken from a custom software package used by TWC.

The big trend from the 00Z Wednesday (8 pm EDT Tuesday) suite of computer model guidance was a marked convergence toward a landfall in the vicinity of North Carolina or Virginia. The 00Z HWRF and GFDL models were joined by the 00Z GFS in hooking Joaquin toward the northwest on Friday and accelerating the hurricane into the coast between Cape Hatteras, NC, and the Delmarva Peninsula as a significant hurricane on Saturday/Sunday. The high-resolution HWRF and GFDL output showed central pressures in the 940-950 mb range at landfall, with wind speeds on par with a Category 2 hurricane. The 00Z UKMET solution angled more north-northwestward, with Joaquin arriving near the southern end of the Delmarva and scraping up the coast into eastern New Jersey and New York. Among the major dynamical models, only the European (ECMWF) model remained adamant that Joaquin would head to sea well before reaching the southeast U.S., although its 00Z track was a bit west of previous runs. The leftward hook prominently featured in the other models is being driven by the increasingly negative tilt (NW-to-SE) to the upper trough deepening over the eastern U.S. late this week. The models are projecting that this trough would pull in Joaquin on its northeast side, in much the same way that a strong upper-level low did with Hurricane/Superstorm Sandy in 2012. However, in Joaquin’s case, the process would unfold a couple of hundred miles to the south. The ECMWF run shows a very similar upper-level pattern to the other models, but the timing of the trough’s interaction with Joaquin and with Invest 90L is such that the hurricane is shunted to sea instead of being tucked into the northeast side of the trough. 90L was centered at 8 am EDT Wednesday about 1000 miles east-northeast of Joaquin. In their 8 am EDT Wednesday Tropical Weather Outlook, NHC gave 90L 2-day and 5-day odds of development of 40% and 70%, respectively.

Figure 4. Model track guidance initialized at 12Z Wednesday (8 am EDT) shows a continued clustering of model solutions toward North Carolina and the mid-Atlantic. This early-track guidance uses 12Z data on Joaquin to update the previous model runs from 06Z. This map does not include the ECMWF model, whose 00Z operational run took Joaquin out to sea. Image credit: Levi Cowan, tropicaltidbits.com.

Back in 2012, the ECMWF model caught on to the leftward hook of Sandy’s track several days before other models. The ECMWF’s high overall skill means we cannot entirely discount its out-to-sea forecast for Joaquin just yet. At the same time, the strong consistency among other leading models in projecting a NC/mid-Atlantic landfall cannot be ignored. We can gain more perspective on this scenario by looking at the ECMWF and GFS ensemble output from 00Z Wednesday. In each ensemble, the model is run a number of times for the same situation, but with the starting conditions varied slightly to represent the uncertainty in our starting-point observations of the atmosphere. The ECMWF and GFS ensembles from 00Z Wednesday are much more similar in flavor than you might expect from looking at their single operational runs. Both models have a majority of ensemble members heading for North Carolina and the mid-Atlantic, with a few outliers heading to sea.

If Wednesday’s 12Z (8 am EDT) models continue to zero in on a NC/VA landfall, and especially if the ECMWF comes more fully around, then this solution will become a more high-confidence forecast. The NHC has been nudging its “cone of uncertainty” toward the left, still splitting the difference between the ECMWF and other solutions while acknowledging the westward trend. The entire U.S. coast from the Outer Banks of NC to southern New England was located in the 5-day cone issued at 11 am EDT Wednesday and valid at 2 am EDT Monday. Even if NHC moves more fully toward the NC/mid-Atlantic scenario, we can still expect to see a large swath of coastline remaining in the “cone” as we get closer to Joaquin’s eventual landfall.

Potential impacts from Joaquin
Apart from the remaining uncertainty about a U.S. landfall, Joaquin is now poised to bring hurricane-force conditions into or very close to the southeastern Bahamas. WIth luck, these islands will remain on the weaker left-hand side of Joaquin. If the hurricane makes a sharp turn to the north on Friday as predicted, the effects should be considerably less on the northwestern Bahamas.

It is relatively rare for a hurricane to make a Sandy-like left hook into the U.S. East Coast. Such a track was unprecedented for New Jersey in hurricane annals, and even in the NC/VA area, it is uncommon enough that the likely effects would be both unusual and high-impact. The closest analogue from recent years is 2003’s Hurricane Isabel. After a much longer life as a Cape Verde system and a Category 4 hurricane from the Central Atlantic (briefly a Category 5), Isabel angled sharply northwestward and made landfall on North Carolina’s Outer Banks as a strong Category 2 hurricane. Isabel then continued on a fairly direct track to western Pennsylvania as it weakened. Isabel’s trajectory brought huge surf to the coast from North Carolina to New Jersey, with a major storm surge pushing into the Chesapeake Bay and nearby waterways, plus widespread impacts from high wind and heavy rain. Joaquin is not as large or long-lived a storm as Isabel, but if it moved slightly to the north of Isabel’s path, its track could be even more favorable for a Chesapeake surge. Hurricane-force winds would be another factor to contend with, especially just north of Joaquin’s track during and just after landfall. Storm surge expert Dr. Hal Needham of LSU has a detailed look at the potential for storm surge from Joaquin along the U.S. East Coast in his Wednesday morning blog post, Widespread Storm Surge Event to Impact U.S. Atlantic Coast.

One very worrisome aspect of Joaquin is the torrential rains that it could bring from the Carolinas to the Northeast and perhaps even New England. Heavy rains and scattered flash flooding have already occurred in parts of these areas over the last 24 hours, as a preexisting front is overtopped by near-record amounts of water vapor streaming over the region ahead of the trough that will help steer Joaquin. The hurricane itself, arriving after several days of antecedent rainfall, has the potential to produce truly historic rainfall totals. This morning’s 7-day outlook from the NOAA Weather Prediction Center, which goes with the NC/mid-Atlantic scenario, shows widespread 5-10″ amounts from North Carolina to southern New England. Model output suggests that localized 7-day totals of 10-20″ or more are not out of the question, depending on Joaquin’s exact track. We’ll have more on the ongoing and potential flood risk in our afternoon post.

Figure 5. Projected 7-day rainfall amounts from 12Z Wednesday, September 30, to 12Z October 7. Image credit: NOAA Weather Prediction Center.

We’ll have a new post this afternoon.

Jeff Masters and Bob Henson

from:    http://www.wunderground.com/blog/JeffMasters/comment.html?entrynum=3131