As Hurricane Irene
churned northwards out of the Bahamas towards the Northeast U.S. on August 25, 2011, residents there scrambled to prepare for the arrival of what could well be the most destructive hurricane ever to hit the United States. Irene had just devastated the northern Bahamas as a Category 3 storm with 120 mph winds, and the National Hurricane Center forecast called for the hurricane remain at Category 3 strength as it plowed over the Outer Banks of North Carolina. Irene was then expected to slowly weaken to Category 1 strength at landfall on the New York/New Jersey coast three days later. Fortunately, Irene surprised forecasters by weakening unexpectedly to a Category 1 storm at landfall in North Carolina, and further weakening to a strong tropical storm with 70 mph winds when it reached New York City on August 28. Even so, Irene did $16 billion in damage, making it the 7th costliest U.S. hurricane in history,
and most expensive hurricane ever to hit the Northeast U.S. (until Hurricane Sandy in 2012.)Figure 1.
Tropical Storm Irene, with top winds of 70 mph, was centered almost directly over New York City in this image taken on August 28, 2011.Figure 2.
The European (ECMWF) model is not known for making good intensity forecasts, and is not one of the intensity models the National Hurricane Center uses to make intensity forecasts. Nevertheless, its 4-day forecast of Hurricane Irene making landfall in Delaware as a borderline Category 3/Category 4 hurricane with a central pressure of 936 mb had forecasters like me sweating a bit.Why did Irene weaken?
Like most intense hurricanes, Irene underwent an eyewall replacement cycle, a process where the inner eyewall shrinks, becomes unstable, and collapses, and is replaced by a new outer eyewall that forms from an outer spiral band. Typically, this process results in a temporary weakening of the storm of 10 - 30 mph in peak winds and 10 - 20 mb in pressure. After about a day, though, the outer eyewall usually grows more organized and contracts, and the storm re-intensifies. Unexpectedly, Irene never completed its eyewall replacement cycle after it left the Bahamas, and the size of the new outer eyewall grew in parallel with an intensification of the hurricane’s outer rain bands. This resulted in the somewhat unusual case where the minimum pressure occurred 40 hours after the winds peaked in intensity. Research accepted for publication last week in the Journal of Atmospheric Science
, led by Barry Lynn of The Hebrew University of Jerusalem, "The sensitivity of Hurricane Irene to aerosols and ocean coupling: simulations with WRF spectral bin microphysics",
gives partial credit for Irene’s weakening to dust and air pollution sucked in by the storm. These particles (collectively called aerosols) invigorated the outer spiral bands and outer eyewall, and kept the outer portions of the storm strong at the expense of the inner core. Using a high-resolution model that used a 1-km grid to simulate the storm, the researchers were able to show that dust particles from the Saharan Air Layer (SAL) and air pollution particles from the Eastern U.S. could have caused a weakening of Irene’s winds of 20 - 30 mph, accompanied by an increase of 10 - 15 mb in the central pressure, and caused the observed delay of the storm’s lowest pressure occurring 40 hours after the winds peaked. Unfortunately, none of the hurricane intensity models that NHC uses include aerosol particles, or have the fine 1 km resolution used in the Hurricane Irene study. I hope that within a few years, computers will become powerful enough to run such a model in real time for use in operational forecasting.How do aerosols weaken a hurricane?
Aerosol particles of the right size and composition (called Cloud Condensation Nuclei, CCN) provide convenient places where water vapor can condense and form cloud droplets. An increase in concentration of small aerosols increases droplet concentration and decreases droplet size. The net effect is a decrease in the collision rate to form large raindrops, and a delay in raindrop formation and rain. As a result, small droplets ascend in cloud updrafts and continue growing by condensation, leading to an increase in supercooled water content. When the moisture condenses in these updrafts, it releases extra "latent heat" (the energy it took to vaporize the water originally, which the water vapor stores). This release of energy leads to an increase in cloud updrafts and an increase in cloud top height and lightning. When this process occurs in the outer bands of a hurricane, the resulting invigoration of the thunderstorms there creates heavy rain that drags down cold air from aloft to the surface, creating pools of cold air near the surface that act to block the inflow of warm, moist air into the hurricane's core, thus weakening the storm.Figure 3.
Dry air/Saharan Air Layer (SAL) maps from the University of Wisconsin - CIMSS (http://tropic.ssec.wisc.edu) for 18 UTC 21 August 2011 (top) and 00 UTC 24 August 2011. Hurricane Irene crossed a wide band of Saharan dust during its northward movement. Figure 3 (top) shows that Saharan Air Layer dust impinged on Irene in three sectors at 18 UTC 21 August 2011. Fifty-four hours later (0000 UTC 24 August 2011) most of the dust was not observable by satellite, but it was still seen to Irene's north in a cloud-free area (Figure 3, bottom). This indicates that the dust seen at the earlier time was likely absorbed into Irene’s circulation.Figure 4.
Official NHC intensity forecasts for Irene every 6 hours from 1200 UTC 23 August to 0600 UTC 28 August. The black line is the observed intensity of Irene. NHC consistently over-predicted the strength of Irene's winds.Aerosols are also credited with weakening Hurricane Katrina
According to Khain et al. (2010), ”Aerosol Effects on Intensity of Landfalling Hurricanes as Seen from Simulations with the WRF Model with Spectral Bin Microphysics”
, ingestion of aerosols also may have been responsible for weakening Hurricane Katrina
as it approached landfall in Mississippi on August 29, 2005. Katrina peaked as a monster Category 5 storm with a central pressure of 902 mb and 175 mph winds about 24 hours before landfall, but weakened to a Category 3 storm with 125 mph winds and a 923 mb central pressure when it came ashore. Up to 35 mph of the wind decrease and 15 mb of the pressure rise could have been due to the storm pulling in aerosol pollution and dust particles from Southern U.S. in the final 24 hours before landfall, the authors deduced, using a detailed computer model of the storm.Why not use aerosols to intentionally weaken hurricanes?
Since we’re pretty sure that aerosols can help weaken hurricanes, why not intentionally introduce small particles into a hurricane to control its intensity? That was the rationale behind a $1 million study by the Department of Homeland Security between 2009 - 2011, HURRMIT (The Identification and Testing of Hurricane Mitigation Hypotheses). Scientists with the project conducted a number of computer simulations on what would happen to a hurricane by intentionally spraying small aerosol particles into the storm using aircraft. They found
that in theory this approach would work, with winds decreasing 20 - 30% for Category 4 and weaker hurricanes. However, the hurricanes had to be treated during an intensification phase to get these reductions in intensity; the effect was significantly less when the simulated storm had completed an intensification cycle and was fully mature. In addition, they found that putting too much aerosol into a hurricane’s outer rain bands thrust more water substance into the thunderstorm anvils, lowering storm precipitation efficiency and short-circuiting the reduction in surface winds.
It is quite possible that seeding a hurricane with aerosol particles can actually make the storm more intense, though. The authors of the Hurricane Irene study commented that if aerosols can manage to penetrate directly to the core of a hurricane, they can act to invigorate the inner eyewall and make the storm stronger. In addition, putting aerosols into a tropical depression in its formative phase can help it, since the storm typically needs an extra boost in cloud droplets to get going (however, the dry air that often accompanies aerosols from the Saharan Air Layer often destroys a budding tropical depression.) So, we’d better be really sure we know what we’re doing if we are going to be intentionally messing with hurricanes. I am very sure that we are not really sure, and that we should leave hurricanes alone for the foreseeable future!
Related: my 2009 blog post, Bill Gates Takes on Hurricanes