NCAR hurricane work reaches new intensity
Shortly before Hurricane Rita strengthened into the second Category 5 hurricane in a month, Science published an intriguing study of tropical cyclones. By analyzing satellite data, the authors, including MMM director Greg Holland, found that the proportion of tropical cyclones that reach Category 4 or 5 status has nearly doubled in the last 35 years.
The September 16 article, whose lead author was Peter Webster of the Georgia Institute of Technology, added fuel to an already contentious debate over whether climate change is playing a role in this year's unusually active Atlantic hurricane season, or whether the series of powerful hurricanes is part of a natural cycle. Researchers don't know enough about hurricanes, typhoons, and other kinds of tropical cyclones to be certain of the answer—yet.
Views of a hurricane.
The two top images show Hurricane Katrina, slightly weakened but still at Category 5 intensity, as it moved over warm water in the Gulf of Mexico on August 28. Twelve hours later, the eyewall was disrupted by interaction with the land surface. Click on each image to enlarge it.
(Images courtesy Jeff Weber, Unidata
, based on satellite data.)
But over the next few years, Greg and other tropical cyclone experts may learn much more. NCAR, in collaboration with Georgia Tech and other partners, is launching a series of projects involving computer modeling, data assimilation, and additional techniques to look at the genesis of cyclones and their structural changes at landfall, as well as to improve cyclone forecasting. The work on tropical cyclogenesis in particular may shed light on the question of whether warming global temperatures are partly to blame for the proliferation of powerful storms.
"These projects are fascinating scientifically and can yield great benefits for society," explains UCAR president Rick Anthes, who has devoted much of his research career to studying hurricanes and other tropical cyclones. "As coastal communities become more populated, it is more important than ever to understand hurricanes, their relationship with climate change, and how to improve forecasts of these storms."
In one of the major hurricane initiatives, NCAR will tap its newest supercomputer, bluevista, to study the genesis of cyclones in the world's tropical oceans. The problem is so demanding, computationally, that it will occupy much of bluevista's available time from October through December. (For more on bluevista).
Greg Holland. (Photos by Carlye Calvin, UCAR. More images
A central issue that Greg and his colleagues hope to explore is why certain tropical waves, or areas of unsettled weather, grow into cyclones while others simply dissipate. This will involve answering such questions as whether a tropical wave is far more likely to become a cyclone if the winds to its east are easterly rather than westerly.
Among the many researchers working on the project are CGD's Julie Caron, Bill Collins, Peter Gent, Jim Hack, Jim Hurrell, Bill Large, Phil Rasch, Joe Tribbia, Mariana Vertenstein, and Steve Yaeger, and MMM's James Done, Jimy Dudhia, Tom Henderson, Joe Klemp, Bill Kuo, and John Michalakes. Additional staffers in those divisions and in SCD, as well as at several universities, are contributing time and expertise to the project. In addition, the Pacific Northwest National Laboratory piece is led by the laboratory's Ruby Leung.
To study the problem, the team is nesting the Weather Research and Forecasting model (WRF) within the atmospheric component of the Community Climate System Model (CCSM). The models complement each other: WRF has a fine-scale resolution that can capture individual storms, while the coarser-scale CCSM is designed to simulate global climate over long periods of time. The models will produce a series of five-year simulations, which will be compared to actual observations.
In addition to shedding light on cyclones, the project may pave the way for improving CCSM's handling of certain tropical processes. The climate model currently fails to capture aspects of the structure of the intertropical convergence zone, for example, and it creates El Niño events every two years instead of every four to five years.
"This is a challenge for all coupled climate models, and it's something we want to address," explains CGD's Joe Tribbia. "It's hoped that by improving the representation of convection in CCSM, we will improve our ability to simulate tropical climate variability and more accurately reproduce events like large-scale tropical waves, Madden-Julian oscillations, and ENSO [El Niño-Southern Oscillation]."
The cyclogenesis project is part of a far-reaching NCAR initiative known as Predicting the Earth System Across Scales. The initiative will be a major step forward in developing a model that can seamlessly move between studies of local weather and global climate, using the correct scale for each type of problem.
Prediction and landfall
When Hurricane Katrina bore down on New Orleans, WRF was one of several models that predicted the storm's track with great accuracy. Hurricane Rita, however, tracked slightly farther to the east than the model projected.
To improve hurricane and other weather forecasts by WRF—which is still being run on an experimental basis—MMM and RAL are collaborating on a project to assimilate radar data from storms into the model. Such data would enable models to begin a simulation already armed with considerable detail about the hurricane's structure, rainbands, and wind fields. Currently WRF and other models rely on coarser-scale information from satellite observations and dropsondes when initiating a simulation.
"What we hope will come out of this is a substantial improvement in forecasting storms as they come within 24 to 48 hours of landfall," Greg explains.
Jenny Sun (sitting) talks with colleagues who are working to assimilate radar data into WRF. The team includes (left to right): Qingnong Xiao, Bill Kuo, Andrew Crook, So-Young Ha, Dale Barker, and Soichiro Sugimoto.
Jenny Sun of RAL and MMM, a member of the assimilation team, believes the project eventually will lead to improvements in WRF as well as in other forecasting models. But she warns that the work may take several years. One of the biggest challenges in developing algorithms for the assimilation is to extract useful information from imperfect data.
"The observations have errors, the model has errors, the forecast has errors. The proper treatment of all those errors is very important," Jenny explains. "It's really a long-term project."
The team also includes Dale Barker, Chris Snyder, and Qingnong Xiao (all in MMM); Andrew Crook (RAL/MMM), So-Young Ha and Bill Kuo (both MMM/COSMIC), and Soichiro Sugimoto (a visitor in MMM).
Along with fine-tuning the forecasts, Greg hopes to develop a better understanding of the physics of a hurricane's landfall. A storm's vortex can break down into smaller vortices when it comes ashore, creating severe shifts in winds and rains. Predicting these changes at landfall would help society better protect people in exposed areas from a storm's most damaging effects. Greg is working with Chris Davis and Kristen Corbosiero in MMM to investigate these landfalling cyclone characteristics and expects that the project eventually will expand to include ISSE and RAL staff.
In time, such information may prove invaluable to the insurance industry. Already, Swiss Re, one of the world's leading reinsurance companies, is collaborating with NCAR on an experimental simulation of Hurricane Katrina's impact on the Gulf Coast. By coupling Swiss Re's computer models of properties in a given city with WRF's simulations of hurricane winds and rainbands, insurers hope to eventually better anticipate property losses.
As Greg explains, "We are moving into an era where it is becoming possible to directly predict the impacts of hurricanes, and this can only help with planning adequate responses."
Are storms getting stronger?
But no matter how well societies can forecast storms, coastal communities are going to be in trouble if, in fact, tropical cyclones keep getting stronger.
Scientists are vigorously debating the question of whether there is a link between climate change and the increasing number of Category 4 and 5 storms, the most intense on the Saffir-Simpson scale. The paper in Science suggests that rising sea surface temperatures (which have warmed by 0.6 degrees Celsius, or 1 degree Fahrenheit, since 1970) are fueling more powerful cyclones. Georgia Tech's Judith Curry and Hai-Ru Chang co-authored the paper with Peter Webster and Greg.
The team utilized the best available data on tropical cyclones from around the globe for the Science article. They focused on the satellite era to minimize errors. In contrast to the sustained rise in cyclone intensity, they found that the frequency of cyclones around the world did not appear to follow a steady pattern. The number of storms increased substantially during the first part of the study period, then dropped in the last decade.
"The balance of
probability is that this [increased hurricane intensity] is associated with the changing climate."
The study lends support to climate models that show tropical cyclones will become more powerful as the world warms during the next century. Scientists are cautious, however, because other factors can affect the intensity of a hurricane, such as atmospheric moisture and variations in winds. In addition, long-term cycles, such as the El Niño-Southern Oscillation, can influence fluctuations in hurricane intensity.
"We still don't have enough knowledge at this stage to say unequivocally. There could be problems with the data, and we cannot rule out natural variability," Greg says. But he adds that in his view, "The balance of probability is that this is associated with the changing climate."
CGD's Kevin Trenberth is pretty certain that greenhouse gases are at least partly to blame for the recent increase in hurricane intensity. After Katrina devastated New Orleans and other parts of the Gulf Coast, he began working on a paper that looked at the extent to which warming temperatures are affecting moisture in the atmosphere. His conclusion: since water vapor amounts over the ocean have increased by an estimated 4% since 1970, which has enhanced both storm winds and moisture inflow to the storms, a preliminary estimate of the net global enhancement of hurricane intensity and rainfall is about 4 to 11% since 1970.
Greg has not attributed a percentage of a cyclone's strength to climate change. But he is struck by the extent to which cyclones have increased in intensity across the tropics.
According to Greg, "There used to be a one-in-five chance that a hurricane would reach Category 4 or 5 status. Now, there's a one-in-three chance. If you live in the tropics, that's a pretty big difference."
• by David Hosansky
Also in this issue...
NCAR hurricane work reaches new intensity
A visitor from New Orleans
Taking command after Katrina
A day for girl scouts
A first-class home for conferences
New HR Web system
bluevista doubles supercomputing capacity
New climate exhibit
Weathering the media coverage
Just One Look
Staff Notes home page | News Center