What was Catarina?

Forecasters, researchers debate nature of Brazil's mystery storm


by Bob Henson

The puzzling storm that moved into the Brazilian state of Santa Catarina on 28 March 2004 provoked discussion and debate even before it made landfall. It had the warm core and the clear, sharp eye of a hurricane, but the height of its cirrus clouds fell short of hurricane standards. It approached the Brazilian shore like many other damaging cyclones, but its winds—as high as 38 meters per second, or 85 miles per hour—and the destruction it wreaked topped anything seen in modern records.

Though Brazilian meteorologists referred to this storm as the Catarina phenomenon, it made international headlines as Hurricane Catarina, the first ever reported in the South Atlantic. Now a variety of specialists is taking a closer look at this vexing cyclone, trying to decide just what manner of meteorological beast Catarina was.

A swirl of intrigue

Greg Holland. (Photo by Carlye Calvin.)

"It's definitely a fascinating case—I'd go so far as to say unique," says Greg Holland, head of NCAR's Mesoscale and Microscale Meteorology Division. A tropical cyclone specialist, Holland has studied a wide range of lower-latitude systems in his native Australia.

In early July, Holland and Jack Beven (NOAA National Hurricane Center, or NHC) were invited to participate in a special two-day workshop on Catarina sponsored by the Brazilian Meteorological Society. The goal was to look at all aspects of the phenomenon, says Holland. The meeting brought together Brazilian modelers, forecasters, university researchers, naval and civil defense personnel, and government administrators.

Catarina's story begins in mid-March 2004, when a nameless low developed well off the Brazilian coast. It was initially a baroclinic system, driven by temperature contrasts, but before long it moved into a zone of light vertical wind shear and weak temperature gradients between the subtropical and polar jet streams. Once there, Catarina slowly took on tropical characteristics: a warm core, with little temperature contrast.

Already, southern Brazil's summer had been a strange one. "January and February 2004 were the coldest in 25 years," notes modeler Pedro Leite da Silva Dias (University of Sao Paolo). Although Catarina was later tagged by some as a possible sign of climate change, the waters over which it formed were actually slightly cooler than average. However, "the air was much colder than normal," says Dias. This produced the same type of intense upward heat flux that fuels hurricanes, normally seen in warmer waters.

Before long, the heat flux and light shear gave birth to a system that bore the satellite earmarks of a hurricane. No hurricanes had ever been reported in the South Atlantic—textbooks dismiss the possibility out of hand—but the eye and other features were so well defined on satellite that Beven expressed his concern about the storm to Brazil's navy, which has responsibility for forecasts in the South Atlantic. The naval meteorologists, in turn, notified Santa Catarina's civil defense unit and regional weather center of the strange system as it neared shore with the satellite-estimated winds of a Category 2 hurricane. Given the NHC's involvement, some media began using the "hurricane" label, and the Brazilians found themselves having to alert locals of an unprecedented threat.

"It was quite clear that operational forecasters had trouble handling the situation. Moreover, numerical forecasts were quite poor," says Dias. Still, many coastal residents were successfully evacuated, and others rode out the storm in their homes. Only one coastal death was reported, even though many hundreds of homes were destroyed. The exact wind speed along the coast is uncertain; a higher-altitude station well inland notched about 85 mph.

"The buildings were not built to withstand that kind of wind because they don't usually get that kind of wind," says Holland. The situation reminds him of Cyclone Tracy, a Category 4 storm that ravaged Darwin, Australia, while Holland was stationed there on Christmas morning of 1975. Although most structures in the town of 40,000 were demolished, only 10 people died on land. "In a tropical cyclone, the wind picks up gradually. You find places to hide. After Tracy, I was staggered to see so few people had been killed."

The light mid-level winds blowing clockwise around Catarina (blue arrows) allowed the storm to blossom with hurricane-force winds as it approached the Brazilian shore at midday on 27 March 2004. (Illustration courtesy Greg Holland; satellite data courtesy NOAA/University of Wisconsin.) (Click on image to enlarge.)

Catarina's cousins

Catarina (named posthumously for the region where it came ashore) also reminds Holland of another Australian phenomenon, East Coast cyclones. These semitropical systems form each year off Australia's eastern shore, sometimes coming in with hurricane-force winds. Yet, like Catarina, they lack the high-topped convection and environmental conditions of more classic tropical cyclones.

Lance Bosart. (Photo by Roger Wakimoto.)

If Catarina is a cousin of Australia's East Coast cyclones, why haven't others been reported in Brazil? Lance Bosart (University at Albany, State University of New York) is looking into that question. Bosart is teaming with his former postdoc Ronald McTaggart-Cowan (now at the University of Quebec at Montreal), Eyad Atallah and John Gyakum (McGill University), and Christopher Davis (NCAR) on a major analysis of Catarina.

The group is focusing on the unusual upper-level features in place when Catarina blossomed. Instead of the usual west-to-east upper flow, which shuttles cyclones away from the Brazilian coast, the jet stream split around the region, leaving a surface zone of high pressure to the south and low pressure to the north. This low-over-high block (it's called a high-over-low block in the Northern Hemisphere) cradled Catarina, allowing it to drift westward without disruption at upper levels.

Such patterns often help spawn East Coast cyclones off Australia, according to Holland, but Bosart says they're much more rare in the South Atlantic. Using the scant upper-air data and satellite images available, Bosart's team has identified only six cases in the last 30 years when this kind of block lasted more than five days. One case, in March 1994, did spawn a weaker version of Catarina, but the storm stayed far offshore over cooler waters. A paper soon to be published in Geophysical Research Letters by Alexandre Pezza and Ian Simmonds (University of Sao Paolo) confirms the extreme nature of the 2005 block.

Bosart believes Catarina isn't a unique occurrence, although its landfall was undeniably historic. "You have to be extremely arrogant to assume a storm like Catarina hadn't happened in the South Atlantic before the satellite era," says Bosart. As for what to call it, Bosart and Davis have been studying what they call "tropical transitions" for years. Like Catarina, Hurricane Diana (1984) emerged from a baroclinic low caught in weak shear, eventually reaching Category 3 strength off North Carolina. Bosart and Davis have identified upper-level pockets of rotational energy, known as potential vorticity anomalies, that may have helped trigger both systems.

Catarina may also be kin to other low-topped, compact cyclones, such as the "Medicanes" observed in the Mediterranean and the intense polar lows that churn in the Arctic, unobserved except by ships and satellites.

The response on the ground

While researchers tussle over how to classify Catarina, the point is moot for the Brazilian public. "The local people now have a firm idea in their heads of what a hurricane looks like," says Holland.

According to Dias, "After Catarina, if a cyclone is forecast along the southern coast of Brazil, it certainly causes havoc."

The Brazilian weather service has decided to refer to any future systems on par with Catarina as hurricanes, although it's unlikely a formal naming system will be established. In consultation with NHC, it's setting up a warning system, and it's also exploring adoption of the Weather Research and Forecasting (WRF) model. "There's no doubt that WRF would do a good job with systems like this," says Holland. "It already has established an excellent track record with North Atlantic hurricanes." Korea has adopted WRF, and Taiwan and India are planning to do the same, while China is building its own model based on WRF.

At the United Kingdom's Hadley Centre for Climate Prediction and Research, modelers saw an interesting coincidence in Catarina. One of their projections of tropical cyclogenesis for the years 2070-2100 shows a compact spike in the region where Catarina formed (see "On the Web"). However, they hasten to point out that only a single low-resolution model run produced this result, and the cause isn't yet clear. In fact, the simulation increases rather than decreases wind shear across the South Atlantic.

As Brazil deals with its newly recognized threat, Holland sees an urgency on the research side as well. "I think it really does behoove the community to sit down and come up with a definition and improved description of this type of cyclone," he says. "You have a system that looks like a hurricane on satellite imagery and was thus called a hurricane by a large fraction of the meteorological community. But it has many differing features. The question is, what was it really, and how do we handle it in the future."



Also in this issue:

Reflective research

Observation transformation: The new EOL

Ten years of SOARS

Climate affairs

ESMF: Plug-and-play modeling

New ESSL head

Catarina up close: Brazil's bizarre storm

Science Bit: The case of the disappearing lakes

President's Corner


Catarina Hadley Center WRF