Dangerous Flood Potential in Texas, Oklahoma from Invest 91L

Residents of the Southern Plains need to keep a wary eye on tropical disturbance Invest 91L as it moves into the northwest Gulf of Mexico. The storm is already bringing winds near tropical storm force over the northern Gulf of Mexico; the South Marsh 268 oil rig located about 50 miles south of the Central Louisiana coast measured sustained winds of 40 mph, gusting to 47 mph, at 8:35 am CDT Monday morning. The Hurricane Hunters, flying at an elevation of 1000 feet, found flight level winds of up to 55 mph and surface winds in excess of tropical storm force (39+ mph) on the northeast side of 91L's center near 9:15 am EDT. Although 91L has tropical storm-force winds, the plane has not yet found a well-defined surface circulation, and the system did not qualify to be named Tropical Storm Bill by the National Hurricane Center on Monday morning. Satellite images showed that 91L's heavy thunderstorms have increased in intensity, and were beginning to consolidate near the low-level center of circulation, which was becoming much more defined. Wind shear on Monday morning was high, 15 - 25 knots, but rising air from 91L's thunderstorms should gradually erode the upper-level low situated above 91L, replacing it with an upper-level high that will help ventilate the developing cyclone. Sea-surface temperatures are increasingly warm ahead of 91L, with unusually high readings for mid-June of 28 – 30°C (82° – 86°F) located near the middle and upper Texas coast. These warm SSTs and the improved upper-level outflow should give 91L a brief window of potential intensification before landfall on Tuesday morning. In their 8 am EDT Monday Tropical Weather Outlook, NHC gave 2-day and 5-day odds of development of 80%. Landfall of 91L should occur Tuesday morning along the Texas coast, according to the 00Z runs of the GFS and ECMWF models.


Figure 1. Invest 91L approaching the coast of Texas as seen by the MODIS instrument at approximately 2:30 pm EDT June 15, 2015. Image credit: NASA.

Regardless of whether or not 91L becomes Tropical Storm Bill before it makes landfall, the system will post a distinct threat of serious flooding over a broad swath from eastern Texas into Oklahoma. Both states just experienced the wettest single month in their history, and soils remain near saturation. Even without such a worrisome precondition, systems like 91L are notorious for producing enormous amount of rain, sometimes with tragic results. 91L has a large and very moist circulation, and steering currents will be weak as the system slowly moves around a strong, hot dome of high pressure over the Southeast. Slow-moving systems need not be intense prior to landfall to generate huge rainfalls once they’re inland, as demonstrated by a number of Gulf of Mexico systems during June and July, when upper flow is often listless. Tropical Storm Allison is a textbook example: in June 2001, Allison drifted into east Texas, then circled back southward and made a second landfall in Louisiana, dumping as much to 38” of rain over six days across parts of the Houston area. Catastrophic flooding from Allison killed at least 41 people and caused some $9 billion in damage. Michael Lowry, hurricane specialist with The Weather Channel, points out that three of the five wettest tropical cyclones on record for the U.S. mainland occurred in Texas, and none of them attained hurricane strength. These include Allison as well as Tropical Storm Claudette, which led to a national 24-hour rainfall record on July 25-26, 1979, near Alvin, Texas—an astounding 42 inches (possibly underreported by as much as three inches due to an overflowing gauge, according to NOAA).


Figure 2. Models generally agree on a track for 91L toward the central Texas coastline.


The projected track of 91L shifted notably westward in last night’s 0000Z model runs, which suggests a larger part of hard-hit eastern Texas and southeast Oklahoma will end up on the much wetter right-hand side of the system. The atmospheric moisture content over southeast Texas is projected to be near record levels for mid-June. The juxtaposition of a slow-moving tropical system with preexisting soil saturation over this region is very unusual and particularly worrisome. Widespread rainfall of 2” – 5” is expected along 91L’s track (see Figure 3), and localized amounts beyond 10” are quite possible with training echoes, especially if the system slows as much as some models are suggesting it might. A flash flood watch is now in effect for parts of southeast Texas from Monday night through Wednesday afternoon. Tornadoes will also be a concern, including this evening over southeast Texas, as instability will be on the high side due to abundant moisture and very warm air at low levels.


Figure 3. Projected five-day precipitation totals from the NOAA Weather Prediction Center, for the period from 8:00 am June 15 to June 20, show the expected track of 91L around the high-pressure center in the southeast U.S. Image credit: NWS/WPC

The “brown ocean effect” and how it could keep 91L going
Tropical cyclones normally dissipate soon after coming ashore, but research over the last few years has shown how it’s possible for a tropical cyclone to maintain its strength or even intensify over land. The most dramatic example is Tropical Storm Erin, which weakened to a depression after landfall on the Texas coast before unexpectedly strengthening over west central Oklahoma three days later. On the night of August 18-19, 2007. Erin’s central pressure dropped from 1007 to 995 mb, and its peak sustained surface winds jumped from less than 25 mph to around 60 mph. A 2011 study in Monthly Weather Review led by Clark Evans (now at the University of Wisconsin–Milwaukee) found that large amounts of latent heat being released from unusually wet soils appear to have helped boost the storm’s intensity, although Evans is continuing to investigate the role of other factors.


Figure 4. Data from the NEXRAD radar near Oklahoma City shows Tropical Storm Erin as it formed a small eye-like feature during intensification at 1000 GMT on August 19, 2007. Image credit: Clark Evans, Russ Schumacher, and Thomas Galarneau, “Sensitivity in the Overland Reintensification of Tropical Cyclone Erin (2007) to Near-Surface Soil Moisture Characteristics,” Monthly Weather Review, doi:10.1175/2011MWR3593.1, American Meteorological Society, from NOAA/NWS data.


At the University of Georgia, Theresa Anderson and Marshall Shepherd carried out a NASA-funded global survey, published in 2013 in the International Journal of Climatology, of 227 tropical cyclones and their behavior after landfall between 1979 and 2008. While many of the cyclones dissipated in the usual fashion, and other went through extratropical transition, 16 held or increased their strength while inland. There were 8 in northwest Australia (where the systems have their own name, agukabams), 3 in India, 3 in eastern China, and 2 in North America. Those two were 2007’s Erin and a much less impressive case, 1996’s Hurricane Fran, which deepened by 1 mb while still a tropical depression over southern Ontario. Erin’s overland intensification produced three times the wind-speed increase of any other cases in the 16 examined by Anderson and Shepherd. They dubbed the inland-strengthening process the “brown ocean effect,” although in the case of red-dirt Oklahoma, it might be the “red ocean effect”! The study found that latent surface heat flux was significant in all of the cases: "The land essentially mimics the moisture-rich environment of the ocean, where the storm originated," said Anderson in a NASA news release. In a subsequent paper, the researchers (with David Radcliffe) showed that latent surface heat flux from the land can actually be larger than from the ocean for brief periods. Another common thread was a uniformly warm, moist tropical air mass across the restrengthening region. With so much water in the Texas and Oklahoma soils, and with rich atmospheric moisture already in place and set to increase, this week could end up providing a fascinating, if troublesome, test of the brown-ocean hypothesis.

Wunderground member Levi Cowan is posting daily video updates on 91L at his Tropical Tidbits website.



Figure 5. The Daily Averaged Fractional Water Index for the topmost 2 inches of soil over Oklahoma on Sunday, June 14. The index ranges from 0 (completely dry) to 1.0 (completely saturated). Image credit: Oklahoma Mesonet.


Weaker Tropical Storm Carlos still a heavy rain threat for Mexico
Tropical storm warnings are up for the Southwest coast of Mexico, as Tropical Storm Carlos continues its slow west-northwest motion at 6 mph, parallel to and about 70 miles offshore from the coast. Fortunately, Carlos is a very small storm, with tropical storm-force winds that only extend outwards about 45 miles, and so far its heavy rains have mostly stayed offshore. Satellite loops showed that Carlos had weakened some on Monday morning, with wind shear partially exposing the surface circulation to view, and the heavy thunderstorm activity waning in areal coverage and intensity. However, the storm was making a modest comeback by late Monday morning, and ocean temperatures are warm enough and wind shear is low enough to allow intensification back to a Category 1 hurricane by Tuesday morning. This intensification may be halted when Carlos gets closer to land; the 00Z Monday run of the European model predicted that Carlos would weaken significantly on Tuesday evening in advance of a Wednesday morning landfall near Manzanillo, Mexico. Dry air and a more stable atmosphere may also lead to weakening on Tuesday. Heavy rains of 5 - 10 inches will likely affect portions of the Southwest Mexican coast through Wednesday, which will cause flash flooding and mudslides. Carlos brought 1.46" of rain to Acapulco, Mexico on Saturday and Sunday. An Air Force hurricane hunter aircraft will investigate Carlos on Monday afternoon and again on Tuesday.

Bob Henson and Jeff Masters


Figure 6. Radar image of Hurricane Carlos from the Acapulco, Mexico radar, taken just before the radar failed at 8:30 pm EDT June 14, 2015. The heaviest precipitation was just offshore.

Wunderblogging hurricane expert Steve Gregory has posted a 3 pm EDT Monday afternoon update on 91L.