UCAR > Communications > Staff Notes > March 1999 Search


March 1999

USWRP gears up for its first science symposium

Flood damage is on the increase: this 1993 flood in the Mississippi Valley (left) was one of the worst on record. Rit Carbone (above) has been examining heavy rain and other weather hazards as lead scientist for the U.S. Weather Research Program and as a key player in the World Weather Research Program. (Photos by Curt Zukosky, left, and Carlye Calvin.)

Landfalling hurricanes and heavy rain. They're two of the thorniest problems in weather forecasting. They, along with winter storms, are at the center of the U.S. Weather Research Program. After several years of problem definition and initial science at NCAR and elsewhere, the USWRP will unveil some of its first findings in a Mesa Lab symposium on 29-31 March. A sampling of the science results is included in a sidebar to this article.

The meeting will also serve as a brainstorming session among several groups: forecasters, weather researchers, social scientists, and the end users of forecasts, such as emergency managers. About 100 of these people will be on hand. For each of the two key topics--hurricanes and quantitative precipitation forecasting--there will be oral presentations, poster sessions, and two panels. The first panel will integrate research, impacts, and policy considerations on the topic at hand. The second will examine the USWRP's contributions to the issues raised in the first panel.

"I think [the symposium structure] is unique," says Roger Pielke, Jr. (ESIG), moderator of the kickoff panel. "I hope that it's as multidisciplinary as the USWRP should be."

The meeting marks a turning point for the USWRP and its international cousin, the World Weather Research Programme (See sidebar). Both are moving from multiyear planning phases into more active research components. Senior scientist Rit Carbone, who has been lead scientist for the USWRP since its inception, is passing the baton to Bob Gall (See sidebar) and turning his attention to the WWRP. Rit has been involved with both programs for years, serving as chair of the WWRP's Interim Science Steering Committee. Now that the WWRP has become part of the World Meteorological Organization, "it requires a much larger effort to lend substance to the original notion," says Rit.

The USWRP is the latest incarnation of an effort spanning the better part of two decades. It started as the Stormscale Operational Research and Meteorology Program (STORM), whose project office was based at NCAR through much of the 1980s. Ever since, a growing contingent of atmospheric scientists has been pressing for sustained support. However, the undertaking has found itself in a funding environment that includes the National Weather Service modernization and the growing national effort to study global change.

Despite the fact that funding has been consistently lower than requested, the USWRP has already made its presence felt. A series of prospectus reports reprinted in the Bulletin of the American Meteorological Society has called attention to the science issues at the heart of the program. About $10 million of the annual NCAR budget is now allocated to USWRP-related work. Some of these funds are distributed internally through a competitive process on a two-year cycle, first carried out in 1997 and recently completed for 1999. This year, for the first time, the USWRP was included in the NOAA budget from the outset, as part of the president's proposal for the upcoming fiscal year.

When it rains . . .

On both the national and global scales, it's clear that heavy rain is a central problem, even with hurricane landfall. Out of the 500 or so U.S. hurricane deaths since 1970, more than 90% were related to freshwater flooding rather than to high winds or storm surge. (Of course, a hurricane storm surge is still capable of tremendous loss of life, notes Roger: "It doesn't mean you let your guard down.")

Although it was one of the most powerful Atlantic storms on record, Hurricane Mitch caused most of its damage through flooding rains in Honduras and Nicaragua as it lingered for days in the Caribbean and then moved slowly inland. Mitch packed winds of 180 miles per hour (290 kilometers per hour) on 26 October 1998 but was only a minimal hurricane when it reached the Honduran coast three days later. More than 9,000 people are believed to have died as a result of Mitch. (Enhanced satellite image produced by Hal Pierce, NASA.)

The catastrophic Hurricane Mitch wreaked most of its havoc last fall through heavy rains and mudslides as the storm ground itself down just offshore of Honduras. "It's a perfect example," says Rit, "a really graphic example of how winds weren't the cause of any significant damage. It was all heavy rainfall. Flooding tends to be the principal societal impact of tropical cyclones."

According to Rit, predicting how much rain or snow will fall at a given spot "is the area of lowest skill in forecasting today. It's an exceedingly tough problem." Both the USWRP and the WWRP will hit the heavy-rain problem from several angles: better mesoscale modeling, improved radar estimates of rainfall, and better understanding of the societal variables.

What lies behind the global increase in flood damages? Rainfall has certainly gone up in many cases. Last year's Yangtze River floods in China were driven by all-time rainfall records that, according to Rit, bested the previous marks by factors of two or three at some stations.

However, in China and with Hurricane Mitch, deforestation and other land-use changes appear to have played into the disastrous flooding. In the United States, flood damages have risen steadily, but rainfall increases can only explain part of the increased flood damage. Roger and colleagues in ESIG are examining factors such as growth in population and wealth to which they attribute the balance of the U.S. flood damage.

An Olympian effort in 2000

As the USWRP continues its ongoing research, major international field studies will be conducted under the WWRP. The first one takes place this summer and fall in southern Europe. The Mesoscale Alpine Program (MAP) will study the effect of the Alps on local airflow and heavy rainfall. MAP will feature substantial NCAR involvement, including the NSF/NCAR Electra and the S-Pol radar. The experiment's U.S. project office is headed by Joach Kuettner, who holds the UCAR Distinguished Chair for Atmospheric Science and International Research. For more on MAP, See the winter UCAR Quarterly. The focus tightens to a single metropolitan area next year, as some of the world's most advanced nowcasting systems are assembled in Australia for the Sydney 2000 project. Side by side, the systems will predict local weather over three months. The most intensive forecasts will coincide with the Olympic Games in late September. A U.S. "team" will be composed of the people behind various expert systems, including candidates for future NWS technology such as the NCAR/RAP Auto-Nowcaster, which projects short-term thunderstorm development. The United Kingdom and Canada will also deploy advanced forecast systems.

Using all of the technology on hand, Australia's Bureau of Meteorology will issue its routine forecasts and nowcasts along with specialized Olympic outlooks. At the same time, the forecasts that would have been issued in the experiment's absence will be archived. Later, an evaluation team of atmospheric and social scientists will compare the two sets of forecasts to determine the improvement made possible by the enhanced capabilities.

"These systems have all been tested in their home countries. We know they work," says Rit. "The question is how portable are they and how much value do they have individually and in combination."

If they prove durable outside their home countries, the systems could be of use to nations whose meteorological infrastructure is less developed. A team of 10 to 15 expert forecasters from such nations as Brazil and Russia will observe the Sydney 2000 project to help determine what technology might be transferable.

Undoubtedly, some financial and proprietary hurdles will emerge down the line. Still, Rit believes that Sydney 2000 could be a big step toward improving forecasts across the world. Because an increasing share of people in developing countries live in huge urban clusters, a forecast strategy that focuses on these areas may be more economical and feasible than one that attempts to extend a single approach to every acre of land.

In the United States, says Rit, a long-standing congressional mandate for "uniformity of service" means that folks in rural Wyoming now get forecasts similar to those received by urbanites in Manhattan, despite their different needs. If the USWRP has its way, the kinds of weather forecasts we receive may someday vary depending on where we live, even if the laws of physics don't. •BH

Keeping track of the two WRPs

As reflected by their similar acronyms, the U.S. Weather Research Program and the World Weather Research Programme are closely related. Both programs address the growing societal impacts of landfalling hurricanes and heavy rainfall, among several forecast problems. However, there are some differences between the programs.

The USWRP has four major sponsors: NOAA, NSF, NASA, and the U.S. Department of Defense (primarily the Navy). Launched early in the 1990s, the USWRP is the outgrowth of over a decade of work to foster storm-related research that has real-world applications. NCAR participates as a home for the lead scientist's office and as a recipient of ongoing NSF funds. Universities and other research labs also participate through grants. The USWRP aims to

  • mitigate the effects of weather disasters
  • lower the costs of routinely disruptive weather
  • use weather information to help increase economic competitiveness
  • create highly specific predictions for densely populated urban zones
  • provide weather information to help the military accomplish its missions.

Along with its emphases on hurricanes and heavy precipitation/flooding, the USWRP is studying the Pacific origins of winter storms and the societal and economic aspects of disruptive weather. For more on the program, check the Web.

The WWRP was approved as a program of the World Meteorological Organization last February. It serves as an international umbrella beneath which many national research programs related to weather prediction can function more effectively. Its key research topics include those of the USWRP, and it is developing a similar built-in component of social science, but the WWRP funds no research itself. Instead, it initiates, endorses, and facilitates projects that require an especially large critical mass of effort.

The WWRP's three emphases are research and development; forecast demonstration projects, such as the Sydney 2000 effort (See main article); and information sharing that helps developing nations to leverage the technology and know-how from countries whose meteorology infrastructure is better developed. The WWRP is interested in a range of high-impact weather events: tropical cyclones, heavy rain and flooding, in-flight aircraft icing, sand and dust storms, winter storms developing upstream from continents, and Mediterranean cyclones. •

New leadership at MMM: Rotunno steps in

Bob Gall. (Photo by Carlye Calvin.)

"I've been in this job for eight years," says MMM director Bob Gall. "It's a job I know very well, so the challenge of doing something different is exciting."

Bob's challenge is to serve as lead scientist for the USWRP for two years, effective 1 March. He'll be taking a sabbatical from MMM for most of this calendar year to focus on his new duties. "The idea is to spend one hundred percent of my time on USWRP for about eight months, so I can get up to speed and really think about it."

In Bob's place, the interim director of MMM through August will be Rich Rotunno, an NCAR scientist since 1980 and a senior scientist since 1989. Rich is known for his studies of supercell thunderstorms, tornadogenesis, and other mesoscale weather features.

Rich Rotunno. (Photo by Carlye Calvin.)

From September through November, a yet-to-be-chosen person will serve as acting director of MMM while Rich goes to Italy to participate in the Mesoscale Alpine Program. Then, from December 1999 into early 2001, Bob plans to return half-time as division director and maintain his USWRP duties. Rich will pitch in through the newly created slot of assistant MMM director while continuing a "significant amount of research."

Bob says he's looking forward to a chance to shape the USWRP during a key period. "It's a program that has been near and dear to my heart for a long time. I think it's important to the nation."

The USWRP also has two new deputy lead scientists. Tom Schlatter (NOAA Forecast Systems Laboratory) will spend part of his time at an FL3 office near the MMM divisional headquarters. The other deputy lead is Russ Elsberry (Naval Postgraduate School). •BH

A progress report on the USWRP

Some 50 scientific presentations are booked for the First USWRP Science Symposium in late March at the Mesa Lab. With the program's research phase just getting under way, many of these talks and posters will cover works in progress. Below are a few of the highlights. The full set of abstracts can be accessed at the USWRP's Web page. Space will be very limited at the symposium. For more details, contact Carey Bousquet, ext. 8197, bousquet@ucar.edu.

  • Hurricane evacuations: Conventional wisdom holds that each hurricane evacuation costs an average of $1 million per mile of coastline. However, Chris Adams (Colorado State University) will show that the costs may exceed $50 million per mile in key areas along the Atlantic and Gulf coasts.

    Despite these costs and despite NOAA's increased skill at hurricane track forecasts, the average length of coastline warned for a given hurricane has nearly doubled since the 1960s. Hugh Willoughby (NOAA Hurricane Research Division) is studying the factors behind this increase, which include tradeoffs between lead time and overwarning. Apparently, forecasters and forecast users have been drawn toward longer forecast lead times at lower precision rather than shorter lead times at high precision.

  • Model development: NCAR is teaming with NOAA and the University of Oklahoma to create a next-generation mesoscale forecast model that will build on current mesoscale models at each institution. Joe Klemp (MMM) and colleagues have begun testing alternatives for the architecture and coding of the Weather Research and Forecast (WRF) model. A community version for research may be ready in two to three years, with an operational version possible by 2004. The WRF will feature multiple nesting of model grids at 1 to 10 km to better depict thunderstorms and other mesoscale precipitation features.

    Meanwhile, a series of field projects is testing the use of ensemble blends that outperform any single model. Dingchen Hou and colleagues at OU's Center for Analysis and Prediction of Storms tested four mesoscale models last May across the Southeast. Each model was run with slight variations at the starting points to produce a suite of possible outcomes. The resulting set of up to 25 forecasts per period showed that each model excelled in one or more measures, but the model consensus was superior to any single run.

  • Heavy summertime rainfall: Chris Davis (MMM) will present findings on long-lived mesoscale cyclonic vortices that help trigger multiday rainfall episodes. These vortices appear with mesoscale convective systems (MCSs) that typically develop overnight and dump heavy rain for 6 to 12 hours. Chris and colleagues have found that when the vertical wind shear is fairly weak, the vortices can persist for hours after an MCS dissipates, sometimes forming the seed for a new MCS the next night. If mesoscale models can track these seed vortices, then there is hope for better forecasts of multiday heavy-rain events. According to Chris, such modeling may also lead to better forecasts of tropical cyclone formation.

    Rit Carbone and other scientists from USWRP, MMM, and NOAA have embarked on a major climatological study of MCSs using data from satellites, radars, and profilers. Condensed water vapor will be calculated in three dimensions, providing an estimate of the latent heat added through condensation and the radiative effects of dense clouds devoid of rainfall. The study will help test and improve data assimilation techniques for mesoscale models. •

    • In this issue...
    Other issues of Staff Notes Monthly

    UCAR
    NCAR
    UOP
    Edited by Bob Henson, bhenson@ucar.edu
    Prepared for the Web by Jacque Marshall