Support Provided ByLearn More
Planet Earth

Hurricane Katrina: Expert Q&A

On October 20, 2005, Marshall Shepherd, a research meteorologist at NASA's Goddard Space Flight Center, answered questions about hurricanes, including Katrina.


Recieve emails about upcoming NOVA programs and related content, as well as featured reporting about current events through a science lens.

Q: Is the temperature of the eyewall of a hurricane colder than the temperature in the eye of a hurricane and the outer edges of a hurricane? If this is so, why? Mrs. Escamilla, EP Foster 5th Grade, Ventura, California

Support Provided ByLearn More
Marshall Shepherd: The center region of hurricanes is warmer than the surrounding air due to latent heat release by the organized convection. This latent heat release is primarily in the eyewall. The eye might be only 0 to 2°C warmer near the surface, but it can be 10°C warmer at 12 km altitude. Hence, a hurricane is called a warm-core system. In addition to being one of the characteristics distinguishing a hurricane from an extratropical cyclone (which is a "cold core" system), a warm core helps to explain certain aspects of the observed wind structure of hurricanes. Interestingly, the air is sinking in the eye of the storm, which leads to warming by compression. If you put your thumb on escaping air from a bike tire, you will understand warming by compression. Eyes are typically characterized by the coldest temperatures of the storm at the surface and the warmest temperatures at upper levels.

Q: Do you think hurricane experts like yourself will ever be in a position to predict the paths and intensity of hurricanes sufficiently in advance that it would be clear that people in certain areas absolutely had to evacuate because such-and-such wind damage and such-and-such a storm surge could definitely be expected? In other words, if people in New Orleans had been told unequivocally that the chances that the levees would likely fail in certain areas had such-and-such a degree of likelihood and that such-and-such areas had such-and-such chance of flooding to the rooftops, maybe fewer people would have died. Or will there always be an element of uncertainty? Thank you. Lisa Ennis, West Chester, Pennsylvania

Shepherd: Lisa, unfortunately meteorologists and planners had warned for years that a Category 3 storm or greater would likely cause failure to the levees in New Orleans. There will always be some element of uncertainty in forecasts. Forecasts are made using computer models that try to solve equations that describe the current and future state of the atmosphere. As an analogy, forecasting is similar to placing a beach ball in the Mississippi River near St. Paul, Minnesota and then trying to guess how the "river flow" will change over a period of days such that you can pinpoint exactly where downstream the beachball will be in that period of time. A daunting task. However, we have and will continue to make progress.

The problem is that the atmosphere and its processes are not linear (A doesn't necessarily lead to B, if you will). Yet our model equations are linearized. Additionally, we still do not have adequate initial conditions for the models. In order for a weather model to produce an accurate forecast, it needs to start with the right initial state of the atmosphere, ocean, and land surface. At NASA, we are experimenting with ways of gaining more and better atmospheric and ocean data (e.g., using various satellite data) for initialization and assimilation into the models. Even with perfect input, there will likely never be a perfect and exact forecast.

Today, errors in track forecasts have narrowed, but as you know, there is still a cone of uncertainty around them. Over time, we will likely continue to narrow that cone of uncertainty and extend its usefulness out beyond two to four days. The important point for the public is not to focus so much on the track forecast but on the general cone of uncertainty that is often present with it. It is unrealistic to expect a pinpoint forecast.

As we are seeing with Hurricane Wilma right now, another difficult forecast problem is intensification. Our models still do a relatively poor job at forecasting rapid intensity changes. As we learn more about how the ocean sea surface temperatures, internal cloud latent heat release, and other factors affect the storm, we will make progress here as well, but there will always be uncertainty.

Finally, we still need more information on how these storms initially develop. Genesis of hurricanes is a challenging problem, and we are studying it very closely.

Q: I was fascinated by your work with "hot towers" (as seen in the January 2005 episode of NOVA scienceNOW) and what they can tell you about how intense a hurricane might become. What are your latest findings with that work? What big questions would you like most to answer? Thanks. Anonymous

Shepherd: The big questions in hurricane research are understanding the factors that lead to rapid intensity changes in storms and what processes lead to genesis of hurricanes. NASA, NOAA, and academia are studying these questions very closely. In the summer of 2005, there was a field campaign led by NASA to address these very questions.

Another looming question, particularly in light of the recent hurricane seasons, is how much (if any) does global warming affect hurricane intensity, frequency, or size. Many are raising this issue currently, but I believe that it is too early to make conclusive statements at this point. We need more data and the ability to look back at data from active hurricane seasons in earlier years or centuries. Sea surface temperatures this year were abnormally warm. Is there a natural or human-induced explanation? We need to find out because warm water is the fuel supply for hurricanes.

As for the hot tower research, there is growing evidence that they may be signatures of intensification. A recent paper by my colleagues at NASA discuss new compelling research on the topic. You can find out more information at:

Q: I have lived through many hurricanes when I lived in New York City, plus one in Norfolk, Virginia while on a business trip. Other than the winds blowing a lot of loose debris, there was really no damage. Do you feel that if the levees had not broken or overflowed during Hurricane Katrina that New Orleans would have survived with essentially no problems? Rob Spahitz, San Diego, California

Shepherd: Rob, it is unlikely that you experienced the full brunt of a category 4 or 5 storm at the latitudes of New York or Norfolk, so it is very difficult for you to really get a sense of how different they are from the weaker storms. At those latitudes, storms have usually weakened due to cooler waters or strong upper level wind flow. Category 4 or 5 storms and associated damages are exponentially greater than the smaller category storms. In New Orleans, the levees were basically vulnerable to any storm of category 3 or larger based on studies that I have read. Planners and forecasters had warned of the New Orleans scenario for many years, and this still was not the "worst"-case scenario. Had Katrina taken more of a SE to NW track, it could have been more catastrophic.

Q: So that a 10-year-old can understand, can you explain how and why hurricanes develop and their connections to other severe weather like tornadoes? Richard Moore, Los Angeles, California

Shepherd: Rather than spend text answering that question, I would point you to the following excellent link on how hurricanes form:

As for tornadoes, hurricanes often spawn tornadoes in the right-front quadrant of the storm at landfall. Typically, these tornadoes are not the same as the tornadic storms associated with supercell thunderstorms that we often see in the Midwest. Hurricane-spawned tornadoes are often smaller vortices that spin up from the surface due to friction-generated spin associated with the rainband motion and interactions with the land surface.

Q: Is there a upper size limit to how big hurricanes can get? Could there theoretically be a hurricane that is, say, twice or three or 10 times the size of Katrina? Anonymous

Shepherd: There is likely an upper size limit on hurricanes before they become unstable, but I can't quote you a magic number as this would be a theoretical study that would have to be done with a model. Even with category 5 storms, we see a point at which they cannot sustain such intensity for long periods of time. The storm will go through various eyewall-replacement or other weakening processes. Additionally, the upper level wind environment and ocean surface temperature/roughness processes are so dynamic that nature has imposed a self-regulating mechanism on the storms.

Q: I would like you ask you: is the increasing intensity of hurricanes that scientists have been saying is now happening the result of global warming, or have we just entered a new phase of a naturally occurring cycle? If it's global warming, what is your personal opinion on how much of that global warming is man-made and how much natural? In other words, is our profligate burning of fossil fuels contributing to more deadly hurricanes? Patricia McCurdy, Wayne, Pennsylvania

Shepherd: I believe that we must look closely at the global warming question, but I personally believe that it is far too early to conclusively link recent hurricane activity to it. There are too many other variables that could equally be factors. As you note, we likely entered (in 1995) an active phase of a multi-decadal cycle in hurricane activity. These phases can last for 20 to 30 years. Many of the studies that are linking global warming to increased sea surface temperatures and hurricane intensity are only looking at data over the last 30 to 60 years. What if there were other periods of increased hurricane activity and sea surface temperatures in the 1700s or beyond?

In other words, we need to understand what the environmental conditions and hurricane climatologies were like in those periods as well, to truly ascertain whether human influence has fundamentally changed hurricane intensities or frequency. Again, I am not suggesting that hurricane activity wouldn't change in a warmer world, I just need more evidence. Remember, the Earth has only warmed by 0.6°C in the last 150 years or so. Is that small increase enough to raise ocean temperatures by one to four degrees, and if so, how? Any why is the effect so obvious in 2004 and 2005 as opposed to previous years? More study is needed.

Q: Hurricane Katrina is an obvious sign of global warming in my opinion, but how long will these horrific hurricanes and extreme weather conditions last? Is there any stopping them? I guess the root of my question lies in whether it even is a sign of global warming. Also, the environmental impacts of the hurricane have been tremendous. How are researchers dealing with this overwhelming job of trying to understand that impact on local and more global ecosystems? Kathryn Arffa, New York, New York

Shepherd: I believe that "obvious" is too strong of a word at this point. (See my answer to the previous question for more on this.)

You are correct that hurricanes dramatically alter ecosystems. There are impacts on wetlands, beach erosion, water systems, and even public health. At NASA, we try to study and model Earth as a system to understand how larger-scale processes like hurricanes affect local and regional processes like runoff or flooding. As computing power and more data become available, we will more effectively be able to link all of these scales.

Q: I know this sounds about as pie-in-the-sky as you can get, but I thought I'd ask anyway. Is there any way to tap into the enormous energy of a hurricane to help with our energy woes? Anonymous

Shepherd: If there are, I am not aware of them. Additionally, the energy that could be extracted by coastal windmill farms would likely be negligible because hurricanes are such sporadic and short-term events.

Q: Have you ever flown into the eye of a hurricane? I've always been fascinated by that. I mean, why don't the very fierce winds in the eyewall simply knock the plane out of the sky? Anonymous

Shepherd: No, I am not that crazy... :)