# News this Week

Science  05 Jan 2007:
Vol. 315, Issue 5808, pp. 24
1. 2007 U.S. BUDGET

# NSF Braces for Opportunities Lost

1. Jeffrey Mervis

Wayne Pfeiffer and his colleagues at the San Diego Supercomputer Center didn't mind working over the holidays on a proposal due 2 February to the National Science Foundation (NSF). They knew their counterparts at other NSF-funded supercomputing centers would be doing the same thing. And besides, the prize seemed worth the extra effort—a $200 million machine capable of performing at the petascale level (1015 operations a second) and, with it, leadership of the next generation in supercomputing. What they didn't figure on, however, is that there might be no winner at all. That outcome has suddenly become quite probable, at least for 2007, after the outgoing Republican Congress adjourned without finishing work on the federal budget and the incoming Democratic leadership announced its intention to freeze spending at current levels (Science, 15 December, p. 1666) until October. A delay—or worse—in the petascale computing competition is only one of dozens of unhappy scientific consequences of the current legislative train wreck for NSF, which had high hopes for an 8% boost this year in its$5.6 billion budget. As the 110th Congress convenes this week, some science lobbyists are hoping against hope to salvage a piece of what was supposed to have been a banner year for the agency.

NSF was one of three entities—along with the Department of Energy's Office of Science and the in-house labs of the National Institute of Standards and Technology—targeted for significant increases as part of the Bush Administration's American Competitiveness Initiative (ACI). The president's 2007 budget request to Congress last February contained the first installment of what was intended to be a 10-year doubling of federal basic research spending in the physical sciences.

NSF, which currently is able to fund barely one in five proposals, is already inundated with good suggestions about how to spend its money. Its proposed 2007 budget was chock full of fresh ideas, from a new $98 million Arctic research vessel to a$25 million pot of money to fund “frontier” research at the intersection of engineering and a host of other disciplines. Some would involve global activities, such as NSF's planned $61 million boost for polar research to take advantage of the International Polar Year (IPY) that begins in March and runs until early 2009. And all were predicated on robust annual growth in NSF's budget. The winner of the petascale computer competition, for example, was scheduled to receive a$50 million downpayment in 2007, with similar increments in each of the next 3 years. Any slip in the 2007 start date, say computer scientists, could put a crimp in the entire project. “We had developed a pretty detailed timeline with milestones” for petascale applications, pilot studies on how to run them, and potential collaborations, explains San Diego's Allan Snavely, director of the center's performance modeling and characterization lab. “Now that momentum could stall. In addition, collaborative teams can't sit around and twiddle their thumbs for a year. Some will move on other things.”

NSF officials say they are sticking to the original schedule for the competition, including site visits this spring. But if the agency's budget remains flat, they will have to make an award sometime after the 1 October start of the 2008 fiscal year. And the project's long budget tail will require sustained increases through 2011, warns Thom Dunning, director of the National Center for Supercomputing Applications in Urbana, Illinois, which is also competing for the petascale machine.

The competition within the engineering directorate for “frontier” research had already attracted 257 preproposals, and NSF was planning to invite 50 teams to submit full proposals by spring and to make 11 awards. The money represents more than half of the directorate's expected 6% increase for 2007. With a flat budget, however, the new office will be lucky to fund more than a couple of projects.

With regard to the Arctic vessel, time is money. A solicitation went out in October to build the ship, and any delay in construction will add to the final cost. Geoscientists say it will likewise slow their quest to understand the impact of global warming on this bellwether region of the planet.

Social scientists also hoped to make a splash in 2007 by ramping up a program to develop what NSF calls the science of science and innovation policy. In response to a complaint by presidential science adviser John Marburger that too little is known about what drives innovation and how to measure it, NSF's social, behavioral, and economic sciences directorate was ready to plow half of its proposed $14 million increase into three research competitions. Anticipating a flat budget, directorate head David Lightfoot says that “if we decide to put out a solicitation, it will be a lot smaller.” With so much on the line, some science lobbyists are hoping that legislators will allow some melting around the edges in the year-long budget freeze they are expected to pass before the current spending resolution expires on 15 February. But most are setting their sights on the 2008 budget cycle, which begins at the end of the month with the president's State of the Union address and, a few days later, his budget request to Congress. Toward that end, three Senate advocates of legislation to authorize ACI spending levels have asked President Bush “to continue to make this issue a top priority in your budget and for your administration.” The 21 December letter from Senators Pete Domenici and Lamar Alexander, Republicans from New Mexico and Tennessee, respectively, and Senator Jeff Bingaman, a Democrat also from New Mexico, could well have been written by the scientific community itself: “If America is going to continue to be the global economic leader, we can not afford to let this wait.” 2. ENDANGERED SPECIES # U.S. Weighs Protection for Polar Bears 1. Elizabeth Pennisi Weighing up to 800 kilograms, the polar bear is king of the Arctic. And last week, the U.S. government made it the poster child for global warming. Citing warming temperatures that are melting the sea ice that polar bears call home, the U.S. Fish and Wildlife Service announced that within the next year it will decide whether to protect these animals under the Endangered Species Act. The move represents the latest in a multidecade international effort to maintain the species. And for environmental groups, it's a belated recognition by the U.S. government that global warming is affecting biodiversity and ecosystem health. “It's a really watershed moment,” says Andrew Wetzler, an attorney at the National Resources Defense Council (NRDC) in Washington, D.C., which has been pushing for polar bear preservation. The world's 20,000 to 25,000 polar bears are divided into 19 populations distributed across the Arctic. Throughout the winter, they hunt seals from sea ice that expands southward each winter and contracts as the temperature rises. Often stranded on land in the summer, the animals fast. Arctic nations have worked for decades to control the hunting of polar bears. But the dangers to the bears'habitat are a more recent concern. In 2005, the Center for Biological Diversity, an environmental advocacy group based in California, asked the Fish and Wildlife Service to add polar bears to the list of threatened species because of fear that continued habitat loss might drive them into extinction. Last week's announcement came after the center, Greenpeace, and NRDC took the agency to court for not responding to the request. There is growing evidence that polar bear habitat—sea ice—is declining, with less of it forming every year. In 2004, the eight-nation Arctic Climate Impact Assessment concluded that the Arctic was warming twice as fast as the rest of the world, that the average annual sea-ice coverage had shrunk by 1 million km2, and that summer sea ice could disappear by 2100 (Science, 5 November 2004, p. 955). In September 2005, Julienne Stroeve of the National Snow and Ice Data Center in Boulder, Colorado, reported that satellite data showed sea-ice coverage had reached an all-time low and was shrinking at an annual rate of about 8%. A 2004 study of polar bears in Canada's Western Hudson Bay—where sea ice is breaking up 3 weeks earlier than it had 30 years ago—highlights the problem for the bears. It found fewer than 1000 bears, down from 1200 a decade ago. The bears are also thinner, and fewer cubs are surviving. “As sea ice melts, bears are having a much more difficult time,” says Andrew Derocher, a population ecologist at the University of Alberta, Edmonton. Researchers don't have firm numbers on the population of bears in the southern Beaufort Sea, which extends across Canada and the United States, but the bears there are also thinner and have fewer cubs, suggesting they, too, are in trouble. In June 2005, these findings prompted the World Conservation Union (IUCN) to list polar bears as “vulnerable” on its Red List of endangered species. The U.S. announcement provides for a 3-month comment period. Only if the agency decides to list the polar bear as threatened will it examine how best to care for the species. Wetzler says that's a necessary f irst step to controlling the greenhouse gas emissions that are the real cause of the bears' plight. “The polar bears' survival is going to depend on our ability to come to grips with global warming,” he says. 3. SCIENTIFIC MISCONDUCT # Japan's Universities Take Action 1. Dennis Normile TOKYO—A series of outstanding scientific misconduct cases ended suddenly and decisively in recent days: Two leading Japanese universities fired scientists because of questionable publications and a researcher is reported to have resigned from a third university over alleged mishandling of research funds. Although university officials say the timing was coincidental, researchers say the unprecedented actions suggest that Japanese institutions are now taking a tougher line on scientific integrity. The resolution of the cases “takes a thorn out of our hearts,” says Norihiro Okada, a molecular biologist at the Tokyo Institute of Technology who with other scientists challenged work by one of the three groups. He says that bad research tarnishes the image of the entire scientific community, adding that “the recent actions of these universities are very much welcome.” On 20 December, Osaka University announced that it had fired a professor previously found by an investigating committee to have fabricated research data. In a brief statement posted on the university's Web page, President Hideo Miyahara confirmed the decision but did not identify the professor or give any details. However, Japanese newspapers identified the scientist as chemist Akio Sugino and the questionable paper as having appeared in the Journal of Biological Chemistry. The journal's Web site lists a paper co-authored by Sugino and published in July as having been withdrawn in August by the authors. According to a local press report, one or more co-authors contacted a departmental research fairness committee, which concluded that Sugino acted alone in fabricating data for at least one paper, although critics also raised questions about other papers. One of the co-authors committed suicide last September, but university authorities declined at the time to comment on a possible connection to the research misconduct allegations. Waseda University in Tokyo concluded an investigation into an alleged misuse of research funds by posting a report on its Web site on 19 December. The report noted that a professor had allegedly drawn money from public research grants to pay students for part-time work but diverted the money into a private bank account. Someone within the school tipped off university officials earlier this year. The professor was previously identified as Kazuko Matsumoto, who has maintained the funds were used for legitimate research purposes (Science, 7 July, p. 31). In its report, the university said an investigating committee found no evidence that the professor (whom it didn't name) had embezzled research funds but noted that the professor was “aware” that payment requests had been mishandled. Two university officials, including the president, received salary cuts of unspecified amounts, and two other officials were given official warnings for lax oversight. In addition, the university will return roughly$1.8 million in research funds to the government. A local newspaper separately reported that Matsumoto resigned on 22 December. She could not be reached for comment.

Finally, on 27 December, the University of Tokyo dismissed Kazunari Taira and Hiroaki Kawasaki for “unreliable” research practices involving papers that appeared in Nature, Nature Biotechnology, the Proceedings of the National Academy of Sciences, and other journals. The RNA Society of Japan raised doubts about the work in an April 2005 letter to the university requesting an investigation. A university investigating committee found Taira's group did not have raw data or notebooks to support results for a number of experiments focusing on RNA. Taira, a chemistry professor, and Kawasaki, a research associate, claimed raw data entered directly into a computer had been lost. The committee later concluded there was no evidence the experiments could be reproduced. (Science, 23 September 2005, p. 1973; 3 February, p. 595; 17 February, p. 931) In a statement posted on its Web page on 27 December, the university said it could not prove deliberate fraud but was dismissing the pair for discrediting the university through “negligent research conduct.”

For both Tokyo and Osaka universities, it was the first time faculty members had ever been dismissed for alleged scientific misconduct. Scientists say that previously, researchers under a cloud of suspicion would have been quietly asked to seek another job. But as funding and competition for funding have grown, scientists—and taxpayers—are demanding greater transparency and accountability.

Over the past year, all three of these universities introduced codes of conduct for researchers and established offices or committees to promote good ethics and investigate allegations of fraud.

But the wider scientific community may not recognize the need for enforcement. A survey conducted over the past year by the Science Council of Japan, the country's largest association of researchers, found that just 13.3% of 1323 responding institutions had adopted a code of ethics and only 12.5% had established procedures for handling allegations of misconduct.

Makoto Asashima, a developmental biologist at the University of Tokyo who chaired the council's Committee on the Code of Conduct for Scientists, says the recent announcements by the three universities will likely spur other institutions to take action. His committee drafted a suggested code of conduct and recommendations for countering misconduct. “Each university and research institution should draw up and implement its own procedures and policies regarding scientific conduct,” he says. This would help establish public trust in research institutions.

4. BIOMEDICAL POLICY

# New Autism Law Focuses on Patients, Environment

Congress has told the National Institutes of Health (NIH) to pick up the pace of its research on autism, with an emphasis on early diagnosis, treatment, and the role of environmental factors. The Combating Autism Act, passed in the waning hours of the 109th Congress and signed into law 19 December by President George W. Bush, authorizes a major increase in spending and orders NIH to come up with a detailed research plan for making progress in understanding and treating the disorder.

“It's giving us a flashing green light to move faster on autism,” says Tom Insel, director of NIH's National Institute of Mental Health in Bethesda, Maryland. What reauthorization bills don't provide, however, is any money. And with most government agencies preparing for flat budgets in 2007 (see p. 24), Jon Retzlaff of the Federation of American Societies for Experimental Biology, says it's “inconceivable” that legislators will divert scarce NIH dollars to autism.

NIH estimates that it spent $101 million last year on autism-related research. The new law allows that figure to increase to$132 million this year and to $210 million by 2011. In addition, the Centers for Disease Control and Prevention in Atlanta, Georgia, which focuses on the epidemiology of autism, could grow its programs from the current budget of$15 million to $21 million by 2011. Advocates for autism research hope that things will speed up even without an immediate funding boost. For example, an interagency committee that coordinates autism research must now submit for the first time an annual report on progress in causes, diagnosis, and treatment. “That certainly puts more emphasis on its role,” says Manny DiCicco-Bloom of the Robert Wood Johnson Medical School in Piscataway, New Jersey, and a member of the board of directors of the nonprofit Autism Speaks. “Perhaps with more teeth, [the committee] can make real changes in policy and levels of performance.” The new law also orders the committee, which reports to the Health and Human Services secretary, to create and implement a strategic plan for autism research. Advocates say that's much better than a 2003 report from the committee, a list of short- and long-term goals with no rankings or recommendations on how to carry them out. The plan must be updated every year and include a draft budget for accomplishing research goals. Insel, who chairs the committee, says he's already convened a working group. He hopes the plan will be ready by the summer. Although the law doesn't set any specific funding levels, it directs NIH to expand, if funds are forthcoming, its work on diagnosis, treatment, and possible environmental causes of autism. That's music to the ears of Jon Shestack of the advocacy group Cure Autism Now, who says that NIH hasn't done nearly enough on this front. In November, an ad-hoc NIH review committee agreed and recommended investigating the possible role of neurotoxic compounds such as pesticides and mercury, developing new biomarkers for exposure, and studying exposure in pregnant women with autistic children. Insel and other scientists agree that those topics are important but argue that, absent more money, NIH should stick with its existing programs. “The research agenda is excellent,” says epidemiologist Eric Fombonne of McGill University in Montreal, Canada. 5. BART GORDON PROFILE # New Chair of House Science Panel Takes Extreme Route to Moderation 1. Jeffrey Mervis Representative Bart Gordon (D-TN) thinks of himself as a moderate Democrat. “On fiscal matters, I'm conservative, and on personal liberties, I'm more liberal,” says the incoming chairman of the House Science Committee, which he is already touting as a user-friendly panel “of good ideas and consensus.” But the word moderate hardly describes his fiercely competitive nature, or how the 59-year-old lawyer, born and raised in the Middle Tennessee district that he has represented for 22 years, lives and breathes politics. Those traits, along with his close ties to Representative Nancy Pelosi (D-CA), the new House Speaker, and her promise to make innovation a centerpiece of the Democratic agenda, could elevate the status of the traditionally low-profile panel and make Gordon a significant player in the 110th Congress that opened for business this week. If that happens, it won't catch Gordon by surprise. He first sketched out his political future as a high school senior in Murfreesboro, Tennessee, while working on the political campaign of a family friend, and the blueprint hasn't changed in 40 years. “I decided then and there to go into public service. And as the son of a farmer and school-teacher, I felt that Congress was probably the highest office I could achieve with just hard work and some degree of certainty. And so I spent the next 18 years preparing to do that.” His 80-year-old mother, Margaret Gordon, recalls that her son “didn't have time for hobbies” as a child. “He's so focused it's pathetic,” she jokes. How focused is Bart Gordon? He admits that he didn't marry until his 50s because of the demands of his job. When Gordon decided to compete in an annual 5k charity race that pits politicians against the Washington media who cover them, he asked the nationally recognized track coach at his alma mater, Middle Tennessee State University, for advice on his workouts. It worked: Gordon holds the unofficial title of “fastest member of Congress.” Going at less than full speed just isn't in his nature. Asked whether he ever thought of taking a more relaxed approach to life, and to his service in Congress, Gordon shakes his head. “This is a fast track. … I would want to excel in whatever I do. It's just not any fun to be in the middle of the pack.” Thanks to Democratic electoral victories in November, Gordon will have the chance to lead a panel that oversees the lion's share of the government's nonmedical civilian research activities. His to-do list comes from the mainstream of his party—strengthen U.S. competitiveness, develop greener sources of energy, improve science and math education, and keep a close eye on the Bush Administration's management of the federal research enterprise. But his legislative strategy is so straightforward that it comes across as radical. “To me, a good idea is a good idea,” he says. “Rather than taking 5 or 6 years to put together a massive piece of legislation like a telecommunications or an energy bill, I think we should try to develop a consensus on the good idea and move ahead with it.” For Gordon, moving ahead on a good idea meant making a bid for Al Gore's House seat in 1984 when Gore decided to run for the Senate. Since then, Gordon has been reelected 11 times, usually by comfortable margins, despite an increasingly suburban district that tends to vote Republican. “He's pretty well convinced his potential opponents, and the Republicans, that they should do their mining somewhere else,” says longtime friend and political confidant Andy Womack, a State Farm insurance agent in Murfreesboro. Although Gordon says he has no ambitions for higher (read governor or senator) office, that doesn't mean he lacks a global vision. However, ask him whether the United States can hold its own against the growing technological prowess of China and India, and his answer couldn't be closer to home. “I've got a 5-year-old daughter who I really believe could be part of the first generation of Americans who could inherit a standard of living lower than their parents,” he says, as his throat catches and a tear forms in the corner of his eye. “That's a complete reversal of the American dream.” In the midst of moving both his personal and science committee staffs on Capitol Hill last month, Gordon spoke with Science about his political philosophy, science, and his plans for the committee. Here are excerpts from that interview. ## On passing an innovation bill “I realized that we have some jurisdictional problems over here in the House. So what I told Senator [Lamar] Alexander (R-TN) is that they should get their bill out as quickly as possible. And then rather than have a parallel bill, we'll come out with a bill that falls within our [narrower] jurisdiction. Then in conference we can put the two bills together. Yes, Humpty Dumpty can be put back together again. I'd rather do that than to slow this thing down by 2 or 3 years by trying to pass exactly parallel bills in each house.” ## On working with appropriators “I don't think there has been adequate communication between authorizers and appropriators. After 22 years, I think I have pretty good relations with both Democrats and Republicans. The appropriators have the dilemma of unlimited wants and limited amounts of money. But I think we can sit down and talk about priorities. In fact, I think it would be interesting to have some joint hearings. It needs to be a collaborative effort. Now, that doesn't mean you get everything you want. But it does mean that you agree to make the best out of limited resources. “You have to do more than just say, ‘We need more money,’ or that ‘the National Science Foundation needs to be doubled'—I'd like 5 years, but 7 years is probably more realistic. We have to sit down and do some give and take. Within the NASA budget, I suspect that whatever we do, there won't be adequate funds to do everything that NASA has been charged with doing.” ## On legislative oversight “I think that the science committee, and Congress as a whole, has acquiesced in its oversight responsibilities. And I think that if somebody is not looking over your shoulder, you become cavalier. I saw it happen to the Democrats [when they controlled Congress prior to 1995]. If you recall, the science committee, under the Republicans, did away with the oversight committee, which was our only vehicle for those investigations. At the same time, I sincerely think that the Republicans were stifling some scientific conclusions and looking to staff committees with people who would go along with those conclusions and discourage any opposition. “I think accountability is important, to save taxpayer dollars and get the most out of government programs. I also think, quite frankly, that we need to do a better job of reviewing whether or not the Administration is cooking the books with science, and prejudging its findings. I think that will be less likely to occur if somebody is looking over their shoulder. My purpose is not to embarrass someone about their prior activity, but rather to it make it clear that from now on we will be providing oversight, so don't do it anymore.” ## On a new agency for energy research “I think we should follow the DARPA [Defense Advanced Research Projects Agency] model, so there would be fewer strings. Plus, the idea of an ARPA-E [Advanced Research Projects Agency, Energy] would be to have a better focus. You can't do a hundred things. We're hoping to find the best seven or eight approaches to renewable energy, and then focus on them. Bring together the national labs, the public and private sectors, to focus on the problem. And I think that the Department of Energy needs a little encouragement to get that done. … I've talked with [Energy] Secretary [Samuel] Bodman. And his reaction is status quo. But status quo isn't getting the job done.” ## On sources of advice “The science committee has a long-tenured, well-credentialed staff, and I feel very comfortable with their advice. … I'm also frequently meeting with associations and university presidents. Of course, as my grandfather used to say, ‘The most important road in the county is the one in front of your house.’ So you need to apply a little bit of a filter to what they say. But I find that in the scientif ic community, there really aren't monetary drivers as much as passions that people have.” ## On global competition “There are seven billion people in the world, and half of them make less than$2 a day. We can't compete against $2-a-day labor, and we wouldn't want to. But now India and China and other countries are also investing in R&D and starting to combine their cheap labor with innovation. So in order to maintain our standard of living, we have to increase our productivity even more. “We want to develop the technology to be first to market, time and again. But we also need a workforce that can work at a high skill level, and not just based on recruiting the one in a hundred students who wants to be a scientist.” ## On the Administration's scientific team “I think NASA Administrator [Michael] Griffin is certainly one [of the most impressive]. Partly because who he followed [Sean O'Keefe], and partly because he is both knowledgeable and candid. We don't always agree, but you know you'll get a honest, from-the-gut assessment. We don't always get that. “Having said that, however, I think that this is a top-down Administration, and there's a lot of pressure from the top down to make the conclusions match the preconceived notions of the Administration. … I think that [presidential science adviser John] Marburger would say that he hasn't been constrained. But I think we need to look into that more. I think he's an honorable and capable man. But he's under a lot of pressure, too.” 6. EPIDEMIOLOGY # Indonesia Taps Village Wisdom to Fight Bird Flu 1. Dennis Normile Participatory epidemiology is Indonesia's first step on a long road to controlling avian influenza SUKASARI, INDONESIA—Sobandi and Rahmet Hidayat stroll down a dusty lane in this village in the highlands of central Java, stopping to chat with a group of women escaping the midday sun on the porch of a modest stucco house. After doffing their shoes and offering the traditional two-handed greeting all around, the two animal health officers sit on the tiled porch floor and ask about chickens. As children gawk at the visitors and a lone hen scratches in the front yard dirt, the women describe a big die-off that occurred last January. The vets already knew about that one; it had wiped out virtually all poultry in the entire district. It's the dry season, so instead of cultivating rice, most men are off working as day laborers or tending roadside produce stalls. Soon an elderly man joins the conversation and mentions a second, smaller die-off this past July. This is new information. The villagers hadn't thought to report it and weren't sure to whom to report such problems. As the chicken struts across the porch and the children return to their play, Sobandi and Rahmet Hidayat, who like many Indonesians use only given names, ask about symptoms to see whether the villagers can distinguish the highly pathogenic H5N1 avian influenza from other diseases. The old man gives what could be a textbook description of the symptoms—swollen and bluish discoloration of combs and wattles, lethargy, and then sudden death—that characterize H5N1. After another round of handshakes, the animal health officers continue down the lane to a second group of women peeling cassava on a porch shaded by banana and coconut trees. As the call to noon prayers comes from the village mosque, the vets squat on their haunches and ask the women about their chickens. This scene has been repeated innumerable times across Java over the past year as Indonesia has struggled to gauge the extent of its avian influenza problem, by all accounts the worst in the world. The virus is endemic among poultry throughout much of the country. “It is very serious; 30 of 33 provinces in Indonesia are infected by avian influenza,” says Bayu Krisnamurthi, who heads the Indonesia National Committee for Avian Influenza Control and Pandemic Influenza Preparedness. “You simply couldn't get more virus in the environment,” says Jeffrey Mariner, a veterinarian at the International Livestock Research Institute in Nairobi, who is helping train surveillance teams like this one under the auspices of the Food and Agriculture Organization (FAO) of the United Nations. The close interaction of infected poultry and humans inevitably leads to human cases, and Indonesia has the worrisome distinction of having the highest number of H5N1 human fatalities of any country, curiously clustered among blood relatives (see sidebar on p. 32). Experts fear that Indonesia provides the perfect setting for H5N1 to evolve into a form easily passed among humans, touching off a global pandemic. Says World Health Organization virologist Keiji Fukuda: “Reducing infections in poultry is a critical aspect of reducing the risk to people.” But, as yet, adds Peter Roeder, an FAO animal health officer, “there is still no systematic control program.” The first step in such a program is to track where and when the outbreaks are occurring, especially among the chickens kept in backyards by 60% of all Indonesian households—an estimated 300 million birds. That's the challenge for a new approach, called participatory epidemiology, that Sobandi and Rahmet Hidayat are trying. Pioneered by Mariner, formerly a professor at Tufts University's International Veterinary Medicine Program in North Grafton, Massachusetts, and Christine Jost, a Tufts assistant professor, it sounds simple enough: Train teams of vets to tap into local knowledge of where and when outbreaks are occurring, and then enlist villagers' cooperation in control efforts. The basic fieldwork provides epidemiological data on how the disease is spreading and kept in circulation, which in turn leads to higher-level strategies for control. But participatory epidemiology has never been tried for avian influenza before—and never on this scale for any disease. Even so, international and Indonesian animal health officials believe it will be a key component of bringing the H5N1 crisis under control, both here and throughout the developing world. ## Starting from scratch “When we started, we had no idea where the disease was and how much of it there was,” says FAO's Roeder. One reason is that Indonesia's animal health infrastructure had deteriorated badly over the past decade, in part, a victim of its own success. John Weaver, a senior FAO adviser in Jakarta, says that after Indonesia brought scourges such as foot-and-mouth disease under control in the early 1990s, vet services became “an easy target” for cost-cutting during the Asian financial crisis at the end of the decade. To reduce national expenditures, animal health services were turned over to local control. The fragmentation and disruption hit the country's poorer regions particularly hard. As an example, Asjachrena Lubis, a Ministry of Agriculture official, points to Maluku Province, a group of islands in eastern Indonesia. It had five government vets under the national system. But with localization, they all left for better paying jobs on Java. And now, “as far as we know, Maluku Province doesn't have a single [government] vet,” she says. Even where local governments have managed to continue animal health services, limited staffing means vets only respond to reports of problems. And even that support never extended to backyard flocks. “For those who keep a few chickens, it's not worth it to call someone who is a 2-hour motorcycle ride away. And what will they do when they get there?” asks Alison Turnbull, a former Tufts student who is helping train the new surveillance teams. Mariner had faced a similarly underdeveloped vet infrastructure when he was a Tufts veterinary graduate student working for FAO's Global Rinderpest Eradication Program in Africa in the early 1990s. “I realized that the farmers knew a lot more about where rinderpest was than the veterinary officials,” Mariner recalls. He adapted older community-based schemes to have vets ask herdsmen about animal deaths and illnesses. This participatory surveillance enabled authorities in Sudan to target vaccination programs that eradicated rinderpest from the country. The technique was later used for rinderpest in Pakistan and for hog cholera in South America. As secretary of FAO's rinderpest program, Roeder was intimately familiar with participatory surveillance and suggested it when he came to Jakarta in the summer of 2005 to help Indonesia craft a response to bird flu. Indonesian health officials agreed it would be a good fit. Krisnamurthi, of the national avian influenza committee, says, “Indonesia is not just vast geographically but also socially and culturally.” The country's 220 million people are spread over 3000 inhabited islands and represent some 350 ethnic populations. Many of these groups are wary of government programs, he says. Roeder adds that Indonesia lacks the strong central government and established veterinary capabilities that enabled top-down bird flu control programs to work in Thailand, which relied on aggressive culling, and Vietnam, which introduced massive vaccination. FAO once again turned to Mariner and Jost to help set up the program. Early in 2006, with$1.5 million in funding from the U.S. Agency for International Development, they established a pilot program in 12 districts in Java that still had some publicly funded vets, forming teams of two specializing in either participatory disease surveillance or participatory disease response.

From early on, it was clear that villagers were well aware they were dealing with an unusual—and unusually lethal—disease, says Jost. They were calling it “new” Newcastle or “strong” Newcastle disease or just “plok”—the sound of a dead chicken falling from a perch. And each week, the teams began reporting one or two previously undetected outbreaks. “It turned up much more avian influenza than anyone expected,” Mariner says. “Poultry populations were fully saturated.”

Those alarming results persuaded Indonesian authorities and international experts to push for a rapid expansion of participatory epidemiology. Even though coverage of the country is still spotty, the data being accumulated are providing clues to what keeps the virus in circulation.

For example, says Roeder, epidemiologists noted a curious pattern of outbreaks occurring after vaccination teams visited villages. They concluded that the teams were likely carrying the virus on contaminated clothing and vehicles and infecting the birds they vaccinated, which died before they developed immunity. Some villagers had recognized this pattern early on and started resisting vaccination, much to the puzzlement of authorities, who didn't make the connection. To cut such risks, the response teams are now training individuals within each community in vaccination.

Participatory response is an equally important part of the program. Mariner says that until recently the standard response was for government vets to cull all poultry in a broad swath around the villages where infected birds were found and then vaccinate widely. Local officials often feel that a show of force is politically necessary when a human case has turned up. This causes resentment among small holders, who may correctly believe that their birds have not been exposed to the virus. Telling small holders to cull their chickens “is like telling Americans to kill their dogs,” Mariner says. Delays in compensation exacerbate the ill feelings.

Instead, the participatory approach is to involve villagers in decisions—ideally, to cull all poultry directly exposed to infected birds, with immediate compensation, and then vaccinate other birds in the vicinity. Mariner says that even small holders can be convinced of the need to cull birds that have been directly exposed to H5N1-infected chickens.

Extending the reach of participatory response will require greater efforts to gain the understanding of local authorities. It will also require dependable funding for compensation and reliable supplies of vaccine. Mariner says some districts used up their yearly allotment of vaccine in a few months.

A bigger challenge is in proving that participatory epidemiology can begin to reduce the number of outbreaks. “I don't think we've had any impact on incidence [of outbreaks] so far,” Mariner admits. Mariner and Jost are planning a new pilot program to measure which control and response measures have the most effect on outbreaks.

## Growing a system

At the same time that Indonesia is verifying the effectiveness of participatory epidemiology, the country, with FAO support and financing from the United States, Australia, and Japan, is planning to extend the program to all of Java and Bali and parts of Sumatra by next May. That will require additional funding as well as more vets, or vet substitutes.

Roeder says he is confident that participatory epidemiology will have an impact. But the surveillance and response teams are just one link in an animal health infrastructure that should stretch from basic labs capable of tracking changes in the virus and verifying the efficacy of vaccines to better oversight of commercial poultry operations to a consistent response coordinated among all levels of government. The Ministry of Agriculture outlined these goals in a strategic plan a year ago. But financial support has been slow in coming.

Until that help comes, no one is willing to bet on when Indonesia will bring avian influenza under control. “This has gone past being an emergency program; we're in for the long haul,” says FAO's Weaver. Which means that Sobandi and Rahmet Hidayat and their colleagues are just at the beginning of a long journey down the dusty lanes of Indonesia's villages.

7. EPIDEMIOLOGY

# Human Cases Create Challenges and Puzzles

1. Dennis Normile

One day in mid-September, a 23-year-old Indonesian man bought four steeply discounted dead chickens at a poultry market in the central Java town of Bandung, carried them home in a plastic bag, and together with his 20-year-old brother butchered them and fed them to the family's dogs. By the end of the month, both men were dead. The older brother was buried before tissue samples were collected, but the younger one was confirmed as Indonesia's 68th human H5N1 case and 52nd fatality.

Indonesia has the greatest number of human H5N1 fatalities, 57 as of 17 December, and the highest H5N1 fatality rate in the world. Each human case increases the risk the virus will adapt to human hosts, sparking the dreaded pandemic.

Most Indonesians who have contracted the disease are not commercial poultry farmers but, like the Bandung brothers, were exposed while doing routine chores. “The key is how to make this type of person understand the danger of carrying dead chickens around,” says Bayu Krisnamurthi, chief executive of the Indonesia National Committee for Avian Influenza Control and Pandemic Influenza Preparedness, which has launched a massive public education campaign.

James McGrane, team leader for the U.N. Food and Agriculture Organization's Avian Influenza Control Program in Indonesia, says it would be culturally and logistically impossible to suddenly eliminate backyard poultry or shift from live markets to a centralized slaughter system. The country has an estimated 300 million backyard poultry and 13,000 live poultry markets.

Participatory surveillance teams are advising small holders on how to handle carcasses and urging them to put their chickens in coops at night, two small ways to improve biosecurity (see main text). The national avian influenza committee is also considering requiring live poultry markets to close periodically for cleaning and disinfecting. But passing the needed legislation will take time.

Until such measures are in place, the best hope of averting a global pandemic is rapid response. This involves monitoring to detect when a deadly virus has started spreading among humans, followed by quarantines, widespread administration of antiviral drugs, and other measures.

In early October, Krisnamurthi's committee held the first of a planned series of simulations, a “tabletop” exercise to establish government responsibilities and lines of communication for pandemic response. Eventually, the government will hold a full-scale drill involving the army and other supporting personnel. So far, scenarios for containing a budding pandemic presume that the deadly virus would emerge in a rural area. But “what if it's Jakarta?” Krisnamurthi asks, shaking his head. “It is definitely fair to say that … a very densely populated urban area … is a more difficult situation than a sparsely populated rural village,” says World Health Organization virologist Keiji Fukuda.

Meanwhile, Indonesian and international epidemiologists are trying to understand why the fatality rate there is so high—75% as compared with about 67% in China and Thailand and 45% in Vietnam. According to Triono Soendoro, director general of the National Institute of Health Research and Development in Jakarta, epidemiologists are searching hospital records for retrospective bird flu cases. There are also plans to screen poultry farmers and cullers for antibodies indicating previous exposure.

Another puzzle is why one-third of Indonesia's human cases have come in clusters of blood relatives, like the brothers from Bandung. In some clusters, relatives by marriage had similar exposures but did not contract the disease. Genetic susceptibility could be involved, says Fukuda, as could different cultural patterns or a change in the virus. Unfortunately, Indonesia is likely to be a laboratory for human H5N1 infections for some time to come.

8. BIOMEDICINE

# Puzzling Out the Pains in the Gut

1. Jean Marx

Newly identified mutations and immune cells are clearing up the mysteries of inflammatory bowel diseases and suggesting novel drug targets

In an era when people often seem obsessed with maintaining a germ-free environment, it may come as a shock to many that the human intestinal tract is home to an estimated 100 trillion bacteria—microbes that help keep us healthy by producing certain nutrients such as vitamin K and fending off pathogenic bacteria that might otherwise colonize the gut. But even though we need our intestinal inhabitants, they present a conundrum, says immunologist Casey Weaver of the University of Alabama, Birmingham: “How can the innate immune system peacefully coexist with all these organisms?”

The innate system is the body's first line of defense against microbes, but there's increasing evidence, much of it genetic, that the coexistence Weaver finds so remarkable does fail on occasion. When that happens, the painful and debilitating conditions known as inflammatory bowel diseases (IBDs) may result. Together, the IBDs, which include Crohn's disease and ulcerative colitis, afflict 1 million people in the United States alone.

These intestinal problems do not arise solely in the innate immune system, however. The more microbe-specific adaptive immune response, which is activated by the innate system and works through T and B cells, can also accidentally prompt an attack on a person's guts. Apparently, the regulatory mechanisms that would otherwise keep both innate and adaptive immune responses in check are somehow disrupted. “The fundamental idea that many people have is that IBD is caused by an abnormal response to the normal gut flora,” says Warren Strober of the National Institute of Allergy and Infectious Diseases in Bethesda, Maryland.

In one line of research that has received particular attention lately, immunologists, working mostly with animal models, have linked a newly discovered type of T cell known as the TH17 cell to the inflammation underlying IBD, as well as to other autoimmune conditions. “The story's emerging that this [T-cell] lineage is a key factor in the progression of the disease,” says immunologist Charles Elson of the University of Alabama, Birmingham.

The growing understanding of how the innate and adaptive immune systems can ravage the gut could lead to new treatments for IBDs. For example, researchers are turning up immune system molecules, including a cytokine that regulates TH17-cell activity, that may be targets for more specific drugs. Current IBD treatments, such as corticosteroids, often rely on suppressing the whole immune system and therefore leave patients extremely susceptible to infection. Immunologists hope that drugs that block just the specific immune activity underlying the gut's abnormal inflammation will have fewer such side effects.

## Trouble on the frontline

Studies of mice that spontaneously develop IBD provide one line of evidence that abnormal immune responses to microbial gut flora are a source of trouble. The mice, which develop persistent colitis in a normal environment, do not do so if reared in a germ-free environment, presumably because they have lost the ability to regulate immune responses to their intestinal partners. The bacteria of the gut, Strober says, “are in a sense like self-antigens: always there” to keep an inflammatory response going.

Five years ago, two groups, one including Judy Cho of Yale University School of Medicine, Gabriel Nuñez of the University of Michigan Medical School in Ann Arbor, and Richard Duerr of the University of Pittsburgh School of Medicine in Pennsylvania and the other led by Jean-Pierre Hugot and Gilles Thomas of the Foundation Jean Dausset in Paris, provided the first direct evidence that an abnormal innate immune response to gut bacteria is at the center of some IBD cases. The researchers linked mutations in a gene then called NOD2 (and since renamed CARD15) to increased susceptibility to Crohn's disease. Estimates are that mutations in this gene account for up to 15% of Crohn's cases.

The gene's protein, still usually called NOD2, is a sensor in the innate immune system that detects a common component of bacterial cell walls. On binding that component, it triggers inflammation and other responses. The mutations detected in the CARD15 gene should result in production of a protein deficient in the ability to recognize and interact with its bacterial targets. Researchers have proposed several ways in which this apparent loss of the protein's function would result in disregulation of immune responses to intestinal bacteria and the persistent inflammation of Crohn's, but the issue remains unresolved.

Last year, a team including Cho, Duerr, and Andrew Gewirtz of Emory University in Atlanta, Georgia, linked a variant of another innate immune response gene to decreased, rather than increased, susceptibility to Crohn's disease, although only in Jewish patients. The gene in question encodes Toll-like receptor 5 (TLR5), another bacterial sensor protein. The discovery ties in with another recent development in IBD research: identification of what appears to be an important bacterial trigger for the intestinal inflammation.

About 2 years ago, Elson's team, in collaboration with that of Robert Hershberg at Corixa Corp. in Seattle, Washington, screened intestinal bacteria from mice with IBD for protein antigens that could contribute to development of intestinal inflammation. They found that about one-quarter of the 60 antigens detected were flagellins, proteins present in the whiplike tails of microbes. Flagellins are known to trigger TLR5 activity, and the TLR5 gene variation linked to decreased susceptibility for Crohn's disease should inactivate the receptor, thereby decreasing innate responses to the proteins. “By activating TLR5, the flagellins can drive inflammation directly through this immune pathway,” Gewirtz says.

## T H 17 cells debut

Illustrating the link between the innate and adaptive immune response, work by Hershberg, Elson, and their colleagues has revealed that several types of mice that develop IBD—and also some human Crohn's patients—have high levels of antibodies to the flagellins. Direct evidence that the bacterial proteins are involved in intestinal inflammation came when the researchers induced colitis in healthy mice by injecting them with flagellin-specific T cells from colitic animals. The T cells presumably contained so-called helper cells that could spark the production of antiflagellin antibodies by B cells. The flagellins probably aren't the only bacterial proteins able to promote colitis, however.

Over the years, immunologists have developed several lines of evidence linking T cells to autoimmune inflammation, including IBD. The early work may have led researchers astray, however. At the time, immunologists had recognized two distinct lineages of T helper (TH) cells: TH1 cells with functions including the destruction of virus-infected cells and TH2 cells that cooperate with B cells to make antibodies.

The evidence, obtained with a variety of mouse models, pointed to TH1 cells as the primary culprits in IBD and other autoimmune conditions. Researchers found large numbers of the cells in various inflamed tissues, for example. In addition, differentiation of TH1 cells depends on a signaling protein called interleukin-12 (IL-12), and antibodies that take IL-12 out of action prevent inflammation in the mouse models. Even so, there were some problems with the TH1 hypothesis.

Knocking out genes needed for TH1 cell activity, including the gene for interferonγ, a key inflammation-stimulating protein made by the cells, should have prevented autoimmune inflammation from developing in the various mouse models. But in some models it didn't. If anything, the animals got even worse inflammation. “People were trying to put square pegs in the round hole of the TH1-TH2 paradigm,” is how Weaver describes the situation.

The confusion began to clear about 6 years ago when Robert Kastelein and colleagues at Schering-Plough Biopharma in Palo Alto, California, discovered a new relative of IL-12, called IL-23. Each of these cytokines consists of two protein subunits. One, designated p40, is common to both, whereas the other—p35 in IL-12 and p19 in IL-23—differs. As it turns out, the antibody used in the experiments implicating TH1 cells in autoimmunity reacts with the common subunit p40—and thus would block the activity of both IL-12 and IL-23. “That literature had to be completely reinterpreted,” Kastelein says.

Kastelein, Daniel Cua, also at Schering-Plough Biopharma, and their colleagues went on to confirm that IL-23, not IL-12, is key to autoimmunity in mice. Working with two models of autoimmunity, one a brain inflammation similar to human multiple sclerosis and the other arthritis induced by collagen injections, the researchers compared the effects of knocking out the gene for IL-12, the one for IL-23, or both. Animals without IL-12 still developed inflammation, whereas those without IL-23 did not.

In publications during the past year, the Schering-Plough team and another group led by Fiona Powrie of the University of Oxford in the United Kingdom have both shown that IL-23 is also needed for IBD development in mice. “IL-12-lacking animals go on to develop colitis,” Powrie says, “but not those lacking IL-23.”

Identification of this new interleukin led in turn to the discovery of TH17 cells. About 2 years ago, Kastelein, Cua, and their colleagues reported that the cytokine promotes the development of a population of T cells very different from TH1 cells. For example, these IL-23-responsive cells produce a different suite of cytokines, including the previously identified pro-inflammatory cytokine IL-17—hence the name TH17 cells.

Weaver's team and also that of Chen Dong at M. D. Anderson Cancer Center in Houston, Texas, have further shown that the development of TH17 cells is independent from that of TH1 cells, requiring a completely different set of cytokines and other regulatory molecules. Researchers also confirmed that this T-cell lineage can promote autoimmune inflammation by showing that injection of the cells into mice induces brain inflammation.

The IL-23-TH17 connection doesn't tell a complete story, however. In the mouse IBD model studied by Powrie and her colleagues, IL-23 can apparently induce intestinal inflammation independently of TH17 cells: The animals in question lacked all T cells. The Oxford team's results suggest that the cytokine is instead working through the innate immune system, particularly via dendritic cells, which serve on the system's frontline as antigen detectors. “That's not to say that [IL-23] doesn't contribute to adaptive T-cell responses,” Powrie says, noting that there are several modes of tissue inflammation that the cytokine could drive.

## A link to human IBD

Although virtually all the work so far on IL-23 and TH17 cells has been in mice, there is some evidence that the cytokine is involved in a human IBD. As reported online in Science on 26 October, a multi-institutional team led by Duerr and Cho performed a genomewide survey looking for gene variations called single-nucleotide polymorphisms (SNPs) associated with the development of Crohn's disease (Science, 1 December, pp. 1403 and 1461). The strongest association uncovered by the survey was a SNP located in a gene that encodes a receptor for a familiar cytokine—none other than IL-23. “The human genetics provides a very strong confirmation of what the immunologists have found,” Cho notes.

That particular SNP, which causes a substitution of the amino acid glutamine for arginine at one location in the receptor protein, is apparently protective, decreasing the chances that people carrying it will come down with Crohn's. But the researchers also found a couple of weaker associations with SNPs in the receptor gene that increase risk. Cho and her colleagues are currently trying to understand how the gene variations influence development of Crohn's. “We think [the protective SNP] might impair IL-23 function, but that's a guess,” she says.

Results of an early clinical trial also lend support—although it's not unequivocal—to the idea that IL-23 plays a causative role in human IBD. The trial, involving a total of 79 people, was designed when TH1 cells were the leading IBD culprits, and its aim was to test the safety and efficacy of an antibody directed against IL-12 in patients with Crohn's disease—the expectation being that the antibody would thwart stimulation of TH1 cells. But that antibody, like those used in the animal studies, blocks the p40 subunit and so should also inhibit the activity of IL-23.

The study showed the antibody to have few side effects, mostly soreness at the antibody injection sites. And although not all dosing regimens showed signs of efficacy, 12 of the 16 patients who received the higher antibody dose given once per week for seven straight weeks had decreased symptoms compared to only two of the eight controls. (The results appeared in the 11 November 2004 issue of the New England Journal of Medicine.) “The antibody produced a very good response in the patients,” says Strober, who was a member of the clinical research group.

Because the antibody reacts with both IL-12 and IL-23, it's not possible to say that only IL-23 is at fault in Crohn's. Indeed, Strober and his colleagues have found that the concentrations of both cytokines are increased in human Crohn's patients and that treatment with the anti-p40 antibody causes the concentrations of each to go down. “The question [of which cytokine is more important] will not be answered until treatment is tried with an anti-p19 antibody directed at IL-23,” Strober says. Such antibodies are reportedly under development at Schering-Plough.

Given the complexity of the IBDs, researchers expect that additional molecules and their genes are going to influence susceptibility to the painful conditions. The discoveries so far “are just the tip of the iceberg,” Duerr says. Other candidates include two genes called OCTN1 and OCTN2, which encode ion transporters. Geneticists are also looking at additional chromosomal sites linked to the diseases.

“Over the next months to a year, many of the low-hanging fruit will be harvested,” Duerr predicts. “Then we can start doing analysis for gene-gene and gene-environment interactions and start putting all the pieces together.”

9. FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION

# Could Mother Nature Give the Warming Arctic a Reprieve?

1. Richard A. Kerr

FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION, 11–15 DECEMBER 2006, SAN FRANCISCO, CALIFORNIA

The high Arctic appears to be on a slippery slope headed for a total meltdown. Year after year of shrinking ice and hungry polar bears seem to foretell immediate greenhouse oblivion. Now, though, some climate scientists say the region—the poster child of global-warming activists—may get a temporary reprieve.

“The [recent] warming has to have a natural component,” explains Arctic researcher James Overland of the U.S. National Oceanic and Atmospheric Administration's Pacific Marine Environmental Laboratory (PMEL) in Seattle, Washington. When that natural trend inevitably reverses, he says, “it's very likely we'll see a slowing of the warming rate in the Arctic for a while.” The region could even cool, until the next natural swing to the warm side again reinforces the greenhouse. A pause in warming could take the wind out of the sails of global-warming activists and bolster climate contrarians who say it's all just Mother Nature fiddling with climate.

At the meeting, Overland and his University of Washington colleague Muyin Wang presented climate-modeling results that point to a major role for the Arctic's natural climate swings. First, they screened 22 climate models for how well they mimicked the region's ups and downs when human-produced “anthropogenic” changes to atmospheric gases were still small. Temperatures in the Arctic swung sharply from decade to decade in the 20th century; 10 of the models couldn't produce that much natural variability and were dropped.

Then Overland and Wang added the human factor. They evaluated the remaining 12 models run under the slowly rising greenhouse gas levels of the early and mid-20th century, the more sharply rising greenhouse forcing of the late 20th century, and then a best guess of how greenhouse gases will build in the 21st century. The models produced midcentury warmings reasonably similar to an actual Arctic warming that spanned the years from 1930 to 1950, Overland and Wang reported, but the timing of the model warmings varied by many years. If the still-weak greenhouse were driving them, they should have all occurred at the same time, so Overland and Wang take the real-world warming to have been natural variability.

Toward the end of the 20th century, however, all the models produced a substantial warming at the same time. That, they believe, is the greenhouse kicking in. But “the air temperature and loss of sea ice [in the Arctic] is occurring faster than global climate models would predict,” says Overland. “I'm interpreting that as a fairly strong natural variability signal on top of long-term anthropogenic change.” If he's right, the strong warming and accelerating ice loss of the past 5 or 10 years are the product of combined natural and humanmade warmings.

In the models' futures, natural climate variability still looms large. “It will be a bumpy road,” says Overland. “It's very likely we could have a 5-year period of colder temperatures, and people could say, ‘Aha, we don't have global warming.'” But the next natural swing to the warm side would once again add to greenhouse warming and would likely “send us to a new place,” he says—a place with far less ice and even hungrier polar bears. “Jim has a very valid message there,” says Arctic researcher John Walsh of the University of Alaska, Fairbanks, but a tricky one to get across. The average citizen may have trouble grasping greenhouse warming that—even only regionally and for a few years—simply goes away.

10. FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION

# Weather Forecasting Way Out There

1. Richard A. Kerr

FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION, 11–15 DECEMBER 2006, SAN FRANCISCO, CALIFORNIA

These days, forecasting rain or shine tomorrow means running huge number-crunching simulations on some of the biggest computers around. Meanwhile, “space weather” forecasters are still not much beyond the “Red sky at night, sailor's delight” stage of their science. Such rule-of-thumb methods have proved handy for predicting the geomagnetic storms that roil Earth's magnetosphere, fire up the aurora, and endanger satellites. At the meeting, however, researchers reported new progress but warned that empirical forecasting is approaching its practical limits. Scientists will have to answer some basic questions to make further progress in the field.

Space physicist Patricia Reiff of Rice University in Houston, Texas, and colleagues have been forecasting geomagnetic storms by using low-flying satellites to measure the speed of charged particles wafting through Earth's upper ionosphere. “We haven't missed a major storm for 2 years,” Reiff boasts. The method works, she says, because the measurements reveal how the next region out—the teardrop-shaped, plasma-filled magnetosphere—responds to the solar wind, which ultimately drives magnetic storms. A few hours' worth of measurements can predict magnetospheric conditions during the next few hours. The Rice scheme also learns from past storm behavior so that the final forecast is not thrown off by the commonplace “quiet before the storm,” Reiff says.

So far, the Rice forecast has been 98% accurate, Reiff reported at the meeting, with very few false alarms. It reliably gives a few hours'warning to satellite operators concerned about damage to sensitive electronics and to managers of vulnerable power grids on the ground. However, “we're reaching the limits of empirical forecasting,” says Reiff. The next step must be full-blown simulations of space weather like those for earthly weather, preferably starting back at the sun, but “there's still some physics we have to learn,” says Reiff.

For example, researchers still lack a clear picture of how the solar wind drives space weather. At the meeting, space physicist Patrick Newell of Johns Hopkins University's Applied Physics Laboratory in Laurel, Maryland, described how he and colleagues tried to nail down which aspects of the wind are most important for determining magnetospheric conditions such as the power of the aurora. After testing scores of formulas, they found that the best one involved just three properties of the solar wind. “The amazing thing is he fit 10 different data sets” with a single formula, says space physicist George Siscoe of Boston University.

Newell thinks the correlation shows that the key to space weather is the rate at which the solar wind's magnetic field lines couple to Earth's magnetic field lines—an area that Siscoe and others pioneered. But Siscoe disagrees with that interpretation. Newell's formula is “telling us something about the coupling process that we don't understand,” he says.

11. FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION

# Snapshots From the Meeting

1. Richard A. Kerr

FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION, 11–15 DECEMBER 2006, SAN FRANCISCO, CALIFORNIA

Taking flight. After a couple of decades of development, crewless airplanes have arrived in atmospheric science. A trio of autonomous unmanned aerial vehicles (AUAVs) served as sensor platforms over the Indian Ocean last March during the Maldives Air Campaign, reported atmospheric scientist V. Ramanathan of Scripps Institution of Oceanography in San Diego, California. Under computer control, one of the 20-kilogram AUAVs would fly its 5-kilogram payload of miniaturized instruments above a cloud to measure incoming sunlight. Another would fly through the same cloud directly beneath the first to measure the properties of cloud particles and sunlight's interaction with them. And a third would fly beneath the cloud to characterize the pollutant particles from India and the Arabian Peninsula rising into the cloud. Such simultaneous in situ observations should help researchers solve a knotty problem of global warming: Are pollutant hazes masking some greenhouse warming by altering clouds?

A nastier early Mars. When the Opportunity rover sent back signs of water early in martian history, the usual descriptor was “shallow salty seas.” Sounded nice and cozy for any early martian life. But at a press conference at the meeting, rover science team leader Steven Squyres of Cornell University made a point of spelling out the team's best current understanding of early Mars, which is much less encouraging. “At the surface, this was primarily an arid environment,” he said. Only occasionally, here and there, would puddles of salty, acidic groundwater form between dunes of salt sand. As the team's latest paper puts it, “dominantly arid, acidic, and oxidizing” environmental conditions would have posed “significant challenges to the origin of life.”

12. FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION

# The Earthquake That Will Eat Tokyo

1. Richard A. Kerr

FALL MEETING OF THE AMERICAN GEOPHYSICAL UNION, 11–15 DECEMBER 2006, SAN FRANCISCO, CALIFORNIA

Denizens of the megalopolis of Tokyo are finally emerging from the threat of their own Big One. The megaquake that last struck offshore in 1923, killing 105,000 people, is not likely to return for many, many decades, researchers in a joint U.S.-Japan study reported at the meeting. But the same study finds that a far more immediate threat—including possible losses totaling $1 trillion—lies right beneath Tokyo and surrounding cities. One-quarter of Japan's 127 million people live in and around Tokyo on the Kanto Plain. Unfortunately for them, not one but two tectonic plates converge on Japan from the east and dive beneath the edge of the Eurasian Plate and Tokyo. Sliding plates sticking and then snapping free produced quakes of about magnitude 8 in 1923 as well as in 1703. How frequently do such quakes strike offshore Japan? A 20-member group asked that question at the meeting. The group was headed by seismologists Ross Stein of the U.S. Geological Survey in Menlo Park, California, and Shinji Toda of the Active Fault Research Center in Tsukuba, Japan, and funded in large part by the insurance giant Swiss Re. Each of those great earthquakes lifted the shoreline by a meter or more. That rise created wave-cut terraces perched above present-day beaches, preserving 7000 years of quake history in the terraces. By dating them, “Team Tokyo” researchers found that the last 17 quakes struck about every 400 years on average with surprising regularity. The probability of the next great quake striking in the next 30 years is then just 0.5%, the group reported. Tokyo didn't get off so easy when Team Tokyo tackled the frequency of smaller quakes beneath the Kanto Plain. To judge by the frequency of earthquakes striking right beneath greater Tokyo, large quakes like the magnitude 7.3 shock of 1855 have about a 20% chance of occurring in an average 30-year period. Combining the two results, the chances of severe shaking in and around Tokyo are about 30% for the next 30 years, the group found, due almost entirely to the threat from beneath the city. Using seismic records of 300,000 earthquakes in the area, the group believes it has pinned down the source of most of Tokyo's moderate but close-in quakes: a 25-kilometerthick chunk of the Pacific Plate broken off and stuck between the three plates beneath Tokyo. Until that jam clears in the geologic future, residents of greater Tokyo will live under the threat of a trillion-dollar catastrophe rising from beneath their feet. At the meeting, seismologist David Jackson of the University of California, Los Angeles, raised the possibility that the threat is even larger than that. The great offshore quakes may not be as periodic as Team Tokyo would have them, he warned. The next one might misbehave and come sooner than expected. 13. # Stalking Discovery From the Infinitesimal to the Infinite 1. Adrian Cho Particle physicists are moving into astrophysics, astronomy, and cosmology; their skills and big-hammer approach could help solve some of the universe's deepest mysteries Stretching for hundreds of kilometers and covered with scrub and prairie grass, the Pampa Amarilla in western Argentina would be an ideal place to graze cattle or film a Western or, on a clear night, gaze at the stars and contemplate one's place in the cosmos. But James Cronin, a particle physicist at the University of Chicago in Illinois, has chosen this unlikely venue to try to solve an enduring mystery of astrophysics. Cronin and 300 colleagues have come to the foot of the Andes mountains to snare particles from deep space that zing along with energies millions of times higher than particle accelerators have achieved on Earth. If all works as hoped, in a few years researchers will spot the sources from which such cosmic rays emanate. “That's never been done, and that would be a huge breakthrough,” says Cronin, who shared the Nobel Prize in physics in 1980 for the discovery of a slight asymmetry between matter and antimatter known as CP violation. The experiment is no small undertaking. Researchers are carpeting the plain with 1600 detectors spaced 1.5 kilometers apart to sense the avalanche of particles created when a ray crashes into the atmosphere. When it is completed, the Pierre Auger Observatory will cover 3000 square kilometers—five times the area of Chicago. True to his particle physicist's training, Cronin embraces a simple credo: “Just think big.” Cronin is only one of many particle physicists who are turning away from Earth-bound accelerators and toward the heavens. In recent years, researchers have begun explorations at the boundaries between particle physics, astrophysics, and astronomy. They are lurking in caves trying to detect particles of the dark matter that holds the galaxies together; sinking detectors into the ice at the South Pole and the waters of the Mediterranean Sea to sense particles called neutrinos from outer space; building gamma ray telescopes to open new eyes on the cosmos; and tracking stellar explosions known as supernovae to decipher the space-stretching dark energy that is accelerating the expansion of the universe. All these endeavors fall under the nebulous rubric of particle astrophysics, or astroparticle physics. “It's likely that in the next 10 years, one of these efforts will lead to a major discovery,” says Gerard van der Steenhoven, a particle physicist at the National Institute for Nuclear and High Energy Physics in Amsterdam, the Netherlands, who works on a neutrino experiment in the Mediterranean. “That makes it very exciting.” The growth of particle astrophysics is not only rejuvenating particle physics but also changing astrophysics and astronomy. Accustomed to working on immense experiments in huge collaborations, particle physicists bring their skills and strategies to fields in which the experiments are already growing rapidly in size and complexity. “You're bringing in a new culture and a new way of operating at a time when the field [of astronomy and astrophysics] needs it,” says Bruce Winstein, a particle physicist at the University of Chicago who now studies the afterglow of the big bang, the cosmic microwave background radiation. But whether particle astrophysics continues to flourish may depend on whether experiments currently in the works deliver any of the hoped-for discoveries. In fact, some say, the future of the field could depend in part on what researchers find at the next great particle collider, currently under construction in Europe. ## Cosmic connections In turning toward astrophysics, particle physics is, in a sense, returning to its roots. Physicists spotted the first bit of antimatter—the antielectron, or “positron”—while studying cosmic rays in 1932. In the same way, they discovered the first particle beyond those that make up the everyday matter around us, the muon, a few years later. But particle astrophysics stretches beyond the study of particles from space. It represents a broad movement of particle physicists into fields such as cosmology and astronomy, where they are pursuing the grandest mysteries in the universe, sometimes without a particle in sight. Most physicists trace the growth of the field to conceptual links between cosmology and particle physics forged in the 1970s and 1980s. For example, theorists realized that the abundance of helium in the universe puts a limit on the number of possible types of neutrinos, wispy particles produced in certain kinds of radioactive decay that interact feebly with everyday matter. (Physicists now know that there are three types of neutrinos.) Others noted that, when mixed into the big bang theory, CP violation might explain why the universe contains so much matter and so little antimatter. Still others realized that a particle theory might help explain the nature of dark matter, the unidentified stuff whose gravity holds the galaxies together. The standard model of particle physics says that matter is made of particles called quarks and leptons that exchange force particles called bosons. A theory called supersymmetry extends this scheme by positing that every known fundamental particle has a more massive doppelgänger that has yet to be discovered. Some of those particles might just fit the bill for dark matter. Such connections have blurred the distinction between particle physics and cosmology, says Jonathan Ellis, a theorist at the European particle physics laboratory CERN near Geneva, Switzerland. “I often find it difficult to tell when I'm writing a paper on particle physics and when I'm writing a paper on cosmology, because in my mind the two are inextricably intertwined,” he says. More recently, experimenters have joined the movement to particle astrophysics, inspired by key discoveries made in recent years. Closest to home, the biggest advance in particle physics in the past 2 decades came from researchers studying neutrinos from space with the Super-Kamiokande particle detector in a mine in Japan. In 1998, physicists found that one type of neutrino could transform into another, a phenomenon known as mixing that can occur only if neutrinos have mass. The standard model assumes that neutrinos are massless, so the observation gives researchers their first peek at physics beyond the standard model. Further a field, scientists studied distant stellar explosions known as type Ia supernovae to trace the history of the expansion of the universe. In 1998, two groups independently reported that the most distant supernovae were even farther away than expected, indicating that the expansion of the universe is accelerating. That stunning observation suggested that some mysterious “dark energy” is stretching the fabric of space. That revolutionary notion was bolstered in 2003 when researchers working with NASA's Wilkinson Microwave Anisotropy Probe satellite mapped the cosmic microwave background in exquisite detail. Analyzing the tiny temperature differences in the radiation across the sky, they found that the universe consists of roughly 71% dark energy, 24% dark matter, and just 5% ordinary matter. The very notions of dark energy and dark matter fire the imaginations of researchers who have devoted themselves to asking, “What's it made of?” says Natalie Roe, a particle physicist at Lawrence Berkeley National Laboratory (LBNL) in California. “Having realized that quarks and leptons are only 5% of the universe, I think it's only natural to ask what the other 95% is,” she says. “So dark energy and dark matter are natural targets for particle physicists.” ## Making a move When explaining their switch into particle astrophysics, researchers cite motives as varied as the particles in the standard model. Most say they were drawn by the intellectual excitement of a young field. “Particle physics was most exciting before the standard model was put in final form and verified,” says Steven Weinberg, a theorist at the University of Texas, Austin, who shared the Nobel Prize in physics in 1979 for his work on the standard model and now pursues cosmology. “In cosmology, the questions are more wide open.” LBNL's Roe, who spent a decade studying the properties of quarks to high precision, says she finds it refreshing to work in a field in which researchers generally don't know what to expect from an experiment. “I wanted to look into something that we really didn't understand, where we don't have a standard model,” says Roe, who is working on a satellite, the Supernova/Acceleration Probe, that would examine dark energy by measuring thousands of supernovae. Many researchers say they switched to particle astrophysics in search of a more congenial work environment. Daniel Akerib, a particle physicist at Case Western Reserve University in Cleveland, Ohio, says he moved away from collider experiments, which typically involve hundreds of collaborators, so he could take a more hands-on approach to his work. “I just felt like I was going to spend all my time in meetings and not have any fun,” he says. Akerib now works with the Cryogenic Dark Matter Search (CDMS), a small group that runs an extremely sensitive detector in a mine in Minnesota and hopes to spot passing dark-matter particles. Some physicists have set out in new directions because opportunities in particle physics have dwindled. David Cinabro of Wayne State University in Detroit, Michigan, had been working on an experiment called BTeV that would have run at the Tevatron collider at Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois. But in 2005, the U.S. Department of Energy suddenly axed the project. “I was faced with the prospect of starting over no matter what I did,” Cinabro says. Cinabro could have joined one of the experiments at the next great accelerator, the Large Hadron Collider (LHC) at CERN, which is scheduled to turn on late this year. Instead, he joined the Sloan Digital Sky Survey, a novel astronomy effort that uses a 2.5-meter optical telescope on Apache Point, New Mexico, to map everything in one quadrant of the sky. Making the shift wasn't easy, says Cinabro, who is studying supernovae and dark energy. “It's like going back to graduate school, because I'm as ignorant as a first-year graduate student,” he says. Still, he says he's happy with his decision. A few researchers say they have pursued particle astrophysics for the sheer adventure of it. “To me it was an opportunity to see Antarctica through the back door and not have to pay for it,” quips David Besson of the University of Kansas, Lawrence, who is working on a prototype neutrino detector at the South Pole. In a phone interview from McMurdo Station, Besson says there is something romantic about searching for radio signals produced by cosmic neutrinos crashing into the ice. “It takes you back to that sense of wonder when you were 5 years old and you'd look up and see the stars,” he says. “Not that you could do that where I grew up in New Jersey.” ## Rearranging the furniture As interest in particle astrophysics has grown, so has funding for such research. For example, in 2000, the U.S. National Science Foundation (NSF) instituted a program in particle and nuclear astrophysics, which now has a$16 million annual budget. And since 1994, the United Kingdom has funded particle physics, particle astrophysics, and astronomy jointly out of its Particle Physics and Astronomy Research Council. But even as the growth of particle astrophysics is expanding the boundaries of particle physics, it is also changing the practice of astronomy and astrophysics.

Most obviously, particle physicists bring with them technologies that are opening new avenues of inquiry. For example, NASA's Gamma-Ray Large Area Space Telescope (GLAST), which is scheduled for launch this October, will provide astronomers with an unparalleled view of the universe as seen in very-high-energy photons. But the “camera” that will detect the gamma rays is a particle detector built at the Stanford Linear Accelerator Center (SLAC) in Menlo Park, California. “In the end, what you need is the best equipment you can get to solve the problem. And if it comes from some other field, why not?” says SLAC's Eduardo do Couto e Silva.

Particle physicists have also introduced a different style of collaboration to astronomy and astrophysics, as exemplified by the Sloan survey. When sharing a telescope, astronomers traditionally allot observers time to use the instrument in turn. In contrast, Sloan researchers pull together to crank out a steady stream of data in a general format, so that collaborators can analyze the data any way they please, just as in a collider experiment. In essence, the Sloan telescope produces astronomical data just as a factory might produce brake pads.

Perhaps most important, particle physicists have appetites for huge projects that push the limits of technology, organization, and funding. “These are not people who are afraid to ask for big things, and they're used to people saying yes,” says Michael Turner, a cosmologist at the University of Chicago who served as assistant director of NSF's mathematics and physical sciences directorate from October 2003 until April 2006. As particle physicists enter astrophysics and astronomy, their habit of “thinking big” is accelerating the natural growth of the size of projects, Turner says.

But even as particle astrophysics blossoms, some researchers worry about its future. Steven Ritz, a particle physicist at NASA's Goddard Space Flight Center in Greenbelt, Maryland, and project scientist for GLAST, fears that the rise of particle astrophysics could undermine accelerator-based research. “Sometimes the movement is interpreted to mean that there's no need to build accelerators anymore, that you can do it all from space,” he says, “and that's just not right.” Even so, the number of colliders is falling, especially in the United States. SLAC will shutter its PEP-II collider in 2008, and a year later Fermilab will unplug the Tevatron, leaving the United States with no colliders for particle physics.

Others say the growth of particle astrophysics will likely slow as the size and expense of projects balloons. “It will soon hit a wall that particle physics hit some time ago, and that is the $1 billion experiment,” says Francis Halzen, a particle theorist-turned-experimenter at the University of Wisconsin, Madison. Halzen's own experiment, Ice Cube, exemplifies the growth of projects in particle astrophysics. A mammoth array of photodetectors being embedded between 1.5 and 2.5 kilometers deep in the South Pole ice, Ice Cube will detect light produced when ultrahigh-energy neutrinos crash into the ice. Scheduled for completion in 2011, the experiment will cost$271 million and involve 400 researchers.

Most of all, the future of particle astrophysics depends on what experiments currently in the works might find. Roger Blandford, a theoretical astrophysicist at Stanford University in Palo Alto, California, says the first big test will come in the search for dark matter. “Our working hypothesis is that dark matter comprises supersymmetric particles,” he says. “We could be terribly wrong.” Given that hypothesis, the prospects for the searches would brighten if the LHC discovers supersymmetric particles—and dim if it doesn't.

For the moment, researchers working in particle astrophysics are happy just to participate in such a young and dynamic field. Promises of momentous discoveries abound. Expectations are sky high.