Thousand-Year Rains Possible in Carolinas; Joaquin Headed North

Hurricane Joaquin continued to lash the Bahamas on Friday morning as it turned north on a course expected to keep it well away from the U.S. East Coast. However, several days of coastal flooding and beach erosion will occur from New Jersey to North Carolina, and extremely heavy rain could produce dangerous impacts in South Carolina. It was a long night of screaming winds, pounding waves, and lashing rains for residents of the Central Bahama Islands, where dangerous Hurricane Joaquin maintained Category 4 intensity with 130 mph winds. The eyewall of Joaquin affected Crooked Island/Acklins Island (population 600), and Long Island (population 3,000) for many hours, and no doubt damage is heavy to extreme on those islands. Joaquin has turned to the north, as seen on microwave satellite animations, and as the storm plows northwards at 3 - 6 mph on Friday, San Salvador Island (population 900) will likely feel eyewall winds. The Hurricane Hunters made multiple passes through the hurricane Friday morning, finding that the central pressure had gradually risen from 935 mb to 939 mb. The size of the eye has been fluctuating considerably, and the Hurricane Hunters noted a secondary maxima of winds away from the eyewall, indicating that an eyewall replacement cycle may be ready to begin. These cycles that lead to a collapse of the inner eyewall, followed by a temporary weakening as a new outer eyewall is established. Wind shear continued to be in the moderate range, 10 - 20 knots, on Friday morning, and visible and infrared satellite loops showed that Joaquin continued to maintain a formidable appearance. Upper level winds analyses from the University of Wisconsin show that the hurricane has now has two impressive upper-level outflow channels, one to the northwest, and one to the southeast. Ocean temperatures in the region remain a record-warm 30°C (86°F). These conditions should allow Joaquin to maintain at least Category 3 strength until Saturday.


Figure 1. Lightning flashes in one of Hurricane Joaquin's spiral bands in this nighttime image taken in the early morning hours of October 2, 2015 from the International Space Station. The lights of Miami are visible in the upper left. Image credit: Commander Scott Kelly, ISS.


Figure 2.  GOES-13 visible image of Hurricane Joaquin taken at 8:45 am EDT October 2, 2015. At the time, Joaquin was a Category 4 storm with 130 mph winds. Image credit: NOAA Visualization Lab.

Forecast for Joaquin
Joaquin is finally embarking on its long-awaited turn toward the north, and the Bahamas are likely the only land areas that will feel a direct impact from the storm. Microwave satellite animations on Friday morning showed the convective core of Joaquin shifting toward the north of the center, and upper-level outflow is now streaming toward the northwest, some of it becoming entrained in the frontal system off the East Coast.

The 00Z Friday (8 pm EDT Thursday) computer model runs continued to lean heavily toward an offshore track for Joaquin. The 00Z GFS and ECMWF solutions inched slightly westward from their previous tracks, bringing Joaquin a bit closer to Cape Cod through a subtle left swing in its path. The 06Z GFS run shifted back toward the east, well away from New England, and the 12Z GFS run also remained far offshore. A slight northward bend in the otherwise northeastward track remains in the GFS, ECMWF, and UKMET solutions, as noted in the 11:00 am EDT forecast discussion from NHC. The ECMWF’s 00Z Friday ensemble runs were quite closely clustered around the offshore track, with only a couple of its 50 members suggesting the potential for a New England landfall. In contrast, more than a third of the 00Z and 06Z GEFS ensemble members continue to indicate the possibility of a SC/NC landfall, although the operational GFS model has not shown such a solution for some time. Among other major models, the Canadian GEM and the U.S. NAM (including the 12Z Friday NAM ran) also point toward an East Coast landfall, but take heed: these are historically among the least-reliable track models, so we would be wise to heavily discount them in favor of the GFS and ECMWF. 


Figure 3. GFS ensemble members from the GEFS run on 06Z Friday, October 2, lean heavily toward an offshore track for Joaquin as depicted in the official NHC forecast, although a few members still bring Joaquin along a looping onshore path near the U.S. East Coast. On the right-hand side are the ensembles’ projected tracks for Invest 90L. Image credit: NOAA Earth System Research Laboratory.

The official NHC forecast track as of 11:00 am EDT Friday keeps Joaquin hundreds of miles away from the U.S. East Coast, and NHC has enough confidence in this track that the “key points” section of its latest forecast discussion does not mention any potential for a U.S. landfall. The persistence of a few model outliers should not be a particular cause for concern at this point, but it does remind us that the upper-level features that will steer Joaquin are complex and dynamic. The two main influences on Joaquin’s track remain the upper low now cutting off over the Southeast U.S. and Invest 90L, located more than 1000 miles east of Joaquin. 90L originated from an upper-level low that has incorporated remnants of former Tropical Storm Ida. The NHC is giving 90L an 80% chance of developing into a subtropical or tropical cyclone over the next 48 hours as it drifts northward. The presence of 90L is creating a pathway for Joaquin to head northeast.

It appears that the strong jet stream diving around the Southeast low will kick eastward around the base of the low over the next couple of days, pushing the eastern part of the low offshore. Together with the influence of slowly developing 90L, this should keep Joaquin moving on a north to northeast track Friday and Saturday. As Figure 3 suggests, a more northeastward motion would lend confidence in the current expectation of an offshore track, while any significant component of motion toward the west today and Saturday would keep open the door for the far-less-likely possibility of a track hooking around the Southeast upper low. We’ll be watching the 12Z Friday model guidance closely and will have more on the forecast for Joaquin in our afternoon update.


Figure 4. Projected rainfall (in inches) for the 72-hour period from 12Z (8 am EDT) Friday, October 2, 2015, to Monday, October 5. Image credit: NOAA/NWS Weather Prediction Center.

Epic rainfall likely for South Carolina
The latest 3-day Quantitative Precipitation Forecast from NOAA's Weather Prediction Center is calling for 10 - 15" inches of rain for the majority of South Carolina, including the cities of Charleston and Columbia.
 
This forecast assumes that Hurricane Joaquin will not come anywhere close to the state. The rain will be due to what meteorologists call a "Predecessor Rain Event" (PRE) (see this paper on them, h/t to Stu Ostro of TWC: http://journals.ametsoc.org/doi/pdf/10.1175/2010MWR3243.1). In a Predecessor Rain Event, tropical moisture well out ahead of a landfalling tropical cyclone interacts with a surface front and upper-level trough to produce heavy rainfall, often with significant inland flooding. The PRE can develop well to the left or right of the eventual track of the tropical cyclone. Slow-moving Hurricane Joaquin is perfectly positioned to transport a strong low-level flow of super-moist tropical air that has water vapor evaporated from record-warm ocean waters north of the Bahamas westwards into the Southeast U.S. Once this moisture hits land, it will encounter a cut-off upper low pressure system aloft, with a surface front beneath it, which will lift the moist air, cooling it, and forcing epic amounts of rainfall to fall. The air will also be moving up in elevation from the coast to the Piedmont and Appalachians, which lifts the air and facilitates even more precipitation. Satellite imagery is already hinting at development of this connection of moisture between Joaquin and the Southeast low and frontal system.


Figure 5. The maximum rainfall predicted to fall in any 24-hour period during the 5-day period from 5 am EDT October 2 to 5 am EDT October 7, according to a high-resolution Weather Research Forecast (WRF) model run done by MetStat, Inc. (http://www.metstat.com.) In some areas of North Carolina and South Carolina, 24-hour rainfall amounts one would expect to fall only once in a thousand years are predicted. MetStat computed the recurrence interval statistics based on gauge-adjusted radar precipitation and frequency estimates from NOAA Atlas 14 Volume 8, published in 2013 (http://dipper.nws.noaa.gov/hdsc/pfds/.) MetStat does not supply their precipitation recurrence interval forecasts or premium analysis products for free, but anyone can monitor the real-time analysis (observed) at: http://metstat.com/solutions/extreme-precipitation-index-analysis/ or on their Facebook page.

Using about a century of precipitation records, NOAA has constructed a Precipitation Frequency Data Server, which estimates how often we might expect to see extreme rainfall events recur.  According to NOAA's Precipitation Frequency Data Server, these could be 1-in-1000 year rains for some locations. (Hydrologists would refer to a 1-in-1000-year rain as having a typical "recurrence interval" of 1000 years. The idea is that such events are not always separated by 1000 years; the same amount of rain could conceivably occur the very next year, or might not occur until thousands of years later.) The three-day 1-in-1000 year rainfall amounts for Charleston, Greenville and Columbia are 17.1", 17.8", and 14.2", respectively. The 24-hour 1-in-1000 year rainfall amounts for Charleston, Greenville and Columbia are 14.8", 15.9", and 12.5", respectively.

The storm to beat in South Carolina is Tropical Storm Jerry of 1995, which dumped up to 18.51" of rain over a small region of Southwest SC. The storm to beat in nearby eastern North Carolina is Hurricane Floyd, which dumped prodigious amounts of rain in mid-September 1999, less than a month after Hurricane Dennis had drenched the region. Floyd produced a broad stripe of 15" - 20" rains, with a maximum total of 24.06" at a site five miles north of Southport, NC (about 30 miles east of the NC/SC border). To get such widespread multi-day totals outside of a tropical cyclone would be a monumental feat.  Averaged across the state as a whole, the wettest three calendar months in South Carolina weather history are July 1916 (14.41"), September 1924 (13.16"), and September 1928 (12.70"). All of these were related to tropical cyclones passing through or near the state. If the NWS precipitation forecasts are in the right ballpark, then the first few days of October 2015 might approach or even exceed these all-time monthly records for the entire state--without any help from a landfalling hurricane or tropical storm!

Texas and Oklahoma have already notched their wettest months on record (by far) this past May, and Illinois had its second-wettest month on record in June. Our warming climate is making intense short-term rains (such as the highest 1-day totals) even heavier in many parts of the United States and the world, although less research has been done on trends in monthly rainfall.

For more on the science of extremely heavy rainfall, see Bob Henson's May 2015 post, The Rains of May and the Science of Recurrence Intervals.


Figure 6. Projected maximum flood category for the 24-hour period from noon EDT Friday, October 2, through Saturday, October 3, 2015. The worst impacts today through Saturday are expected through the southern part of the Chesapeake Bay. Image credit: NOAA/NWS Advanced Hydrologic Prediction Service.


Figure 7. Strong on-shore winds along the mid-Atlantic coast due to the pressure gradient between Hurricane Joaquin and a strong high pressure system over the Northeast U.S. were creating storm surge heights of 2 - 3' in many locations, and over 3' on Virginia's Delmarva Peninsula. Image credit: Hal Needham.

Long-duration coastal flooding under way
The combination of Hurricane Joaquin, the Southeast U.S. low, and a strong ridge well to the north is leading to an unusually prolonged period of steady onshore flow and high surf along the U.S. East Coast from New Jersey southward to North Carolina. The highest-impact coastal flooding and beach erosion can be expected along the Virginia and Delaware coast, including Ocean City, MD, and the Hampton Roads area of VA, which includes Norfolk and Virginia Beach. The Wakefield, VA, NWS office is calling for several rounds of moderate to severe coastal flooding through the weekend. See the latest blog post from storm-surge expert Hal Needham for more details on this event.

We’ll have an update later this afternoon.

Jeff Masters and Bob Henson