Will Global Warming Make Hurricane Forecasting More Difficult?

Will Global Warming Make Hurricane Forecasting More Difficult? That’s the title and provocative premise of a new paper by MIT hurricane scientist Kerry Emanuel (early online PDF available here from the Bulletin of the American Meteorological Society.) Dr. Emanuel makes the case that the most dangerous storms—tropical cyclones that intensify rapidly just before landfall, catching forecasters and populations off guard, thereby risking large casualties—are likely to become increasingly frequent and severe as the globe warms, increasing from one such storm every 100 years to one every 5 - 10 years.




Figure 1. Since 1900, 56% of all U.S. deaths from tropical cyclones have been caused by the three storms shown above.

Tropical cyclone mortality dominated by a small number of events
Since 1971, tropical cyclones (which include all hurricanes, typhoons, tropical storms, and tropical depressions) have killed 470,000 people (about 10,000 per year) and caused $700 billion in damage, according to the international disaster database, EM-DAT. Most of these deaths were caused by just a few storms—for example, three Atlantic hurricanes (the Great Galveston Hurricane of 1900, the 1928 Lake Okeechobee hurricane, and Hurricane Katrina of 2005)—caused 56% of all U.S. hurricane deaths since 1900.

Increased vulnerability due to growing coastal populations
In recent years, better tropical cyclones forecasts have resulted in reduced death tolls and lower damages than would otherwise have occurred. However, a large increase in coastal population resulted in an almost three-fold increase in the global population exposed to tropical cyclone hazards between 1970 and 2010 (Peduzza et al., 2012.) This helped fuel an increase in tropical cyclone damages of about 6% per year between 1970 and 2015, according to EM-DAT. Thus, much improved forecasts and/or major reductions in vulnerability though better preparedness and building codes are needed to avoid increasing tropical cyclone death tolls in the coming decades.

Poor intensity forecasts make us vulnerable
While track forecasts of hurricanes have improved by more than a factor of two over the past 20 years, intensity forecasts have shown little improvement. Dr. Emanuel gives four reasons for this:

1) Very high resolution computer models are needed (1 km resolution or better), which are beyond the capability of modern computers to run economically.

2) We have poor understanding of and models of the processes in the lowest few hundred meters of the atmosphere (the boundary layer).

3) We have difficulty modeling how the top few hundred meters of the ocean responds to a storm.

4) The process of taking observations that show a dramatic variation over short distances and correctly initializing a hurricane model with these observations is difficult.

The 2016 hurricane season gave us two humbling examples of how far we still have to go with intensity forecasts. As Hurricane Matthew drifted across the southern Caribbean Sea in late September, the hurricane rocketed in strength from Category 1 to Category 5 in just 24 hours (from 80 mph sustained winds at 03Z on September 30 to 160 mph at 03Z on October 1). The official NHC forecast at the start of this day-long burst was for Matthew to take three days to top out at high-end Category 2 strength (105 mph). Less dramatic but still eye-opening was Nicole’s surge from Category 1 to Category 4 strength in the Northwest Atlantic over just 21 hours (from 90 mph sustained winds at 06Z on October 12 to 135 mph at 03Z on October 13). Like Matthew, Nicole had also been predicted at the start of its rapid strengthening to remain just below the major hurricane threshold (Category 3). Dr. Emanuel gives an additional troubling example of a rapid intensification evert that was poorly forecasted: Hurricane Patricia of October 2015, which hit a relatively unpopulated portion of the Pacific coast of Mexico as a Category 4 storm with 150 mph winds after topping out as the strongest tropical cyclone ever measured—215 mph sustained winds. During a 24-hour period from October 22 at 06 GMT to October 23 at 06 GMT, Patricia intensified by an astonishing 120 mph—from an 85 mph Category 1 storm to a 205 mph Category 5 storm. During this same period, the National Hurricane Center predicted an intensification by only 35 mph. Dr. Emanuel notes, “Had the storm made landfall at the end of this period of rapid intensification, the result could have been catastrophic given the poor anticipation of the magnitude of the event.” 


Quantifying the probability of rapid intensification just before landfall
Quantifying the probability of rapid intensification just before landfall is difficult using the existing database of global tropical cyclones, which goes back about 60 years in the Atlantic, but only 35 - 45 years in the Southern Hemisphere. Since Dr. Emanuel’s study was interested in rare storms that have a return period of about once every 100 years, a computer model was used. The model generated a set of 22,000 landfalling U.S. hurricanes during the recent climate period of 1979 - 2005, and looked at storms that rapidly intensified just before landfall. The analysis found that about once per century, we should expect to see a hurricane that intensifies by 70 mph or greater in the 24 hours just before landfall. The major metropolitan areas most at risk for surprise intensification just before landfall included Houston, New Orleans, Tampa/St. Petersburg, and Miami.

Hurricanes are heat engines that take heat energy out of the oceans and convert it to the mechanical energy of wind. Thus, hurricane scientists are in broad agreement that global warming should make the strongest hurricanes stronger. The computer modeling results of Dr. Emanuel also found that global warming under a business-as-usual scenario would result in more cases of rapidly intensifying hurricanes making landfall in the United States. The odds of a hurricane that intensified by 70 mph or greater in the 24 hours just before landfall increased from once every 100 years to once every 5 - 10 years by the year 2100 in his simulations. What’s more, 24-hour pre-landfall intensifications of 115 mph or more—which were essentially nonexistent in the late 20th Century simulations—occurred as often as once every 100 years by the year 2100. With increasing coastal populations, limited skill in intensity forecasting, and steadily increasing sea levels, this potential increase in rapidly intensifying hurricanes results in the “risk of an increased frequency of poorly anticipated high-intensity landfalls leading to higher rates of injury and death,” wrote Dr. Emanuel. He recommended that “greater emphasis be placed on improving hurricane intensity prediction and on preparing populations to respond to high intensity landfalling hurricanes at short notice.”


Links
Top Ten Tropical Cyclone Events of 2016 Potentially Influenced by Climate Change (my December 2016 blog post)
Hurricane Patricia's 215 mph Winds: A Warning Shot Across Our Bow (my 2016 blog post)
Fewer but Stronger Global Tropical Cyclones Due to Ocean Warming (my 2015 blog post)
Hurricanes and Climate Change: Huge Dangers, Huge Unknowns (my 2013 blog post)
Climatesignals.org analysis of Hurricane Matthew



Video 1. Incredible footage taken during the flight of NOAA hurricane hunter aircraft N43RF through the eye of Category 5 Hurricane Patricia on the afternoon of October 23, 2015, when the storm was near peak strength, with 205 mph sustained surface winds and a central pressure of 878 mb. Lt. Cmdr. Scott Price (the mission's Aircraft Commander) made the video using a GoPro camera. The video begins inside the eyewall: note that the intense rain and wind combination makes it impossible to see the nose of the aircraft just a few feet away. Since the aircraft is flying perpendicular to the wind in order to find the center, the rainfall is blowing from left to right in front of the pilot's vision. At 37 seconds, the crew enters the eye of the hurricane, where the violent sea-state below becomes visible. Note that due to the storm's incredibly steep pressure gradient, the aircraft is pitched downward as the aircraft descends closer to the ocean, in order to keep flying at a constant pressure altitude. At 57 seconds, the curved eyewall on the opposite side of this very small eye becomes apparent. After a couple of slight turns requested by the Flight Meteorologist to report the exact center of the storm, the crew turns right to avoid the worst of the eyewall. At ~2 minutes into the video, the aircraft reaches the opposite eyewall where the crew loses visibility once again. Posted by The NOAA Hurricane Hunters on Thursday, November 5, 2015.

Jeff Masters