More than 200 miles above the Arctic Circle, Inuit hunters were perplexed by the unpredictability of the spring weather. Temperatures and conditions were changing quickly and without warning, they observed. Their traditional forecasting methods, based on cloud patterns, wind, or the twinkling of stars, weren't working.
For 10 years, University of Colorado (UC), Boulder, scientist Shari Gearheard documented the Inuit's observations. But scientific weather data didn't reveal the rapid shifts the Inuit hunters reported. Then a UC Boulder colleague took a different look at the data: Instead of focusing on maximum highs and lows, statistician Elizabeth Weatherhead analyzed temperature swings. She discovered day-to-day changes in temperature and weather that linked scientific data with Inuit reports.
As a result, Inuit observations can now help scientists refine their climate models. It's an example of how some scientists are managing, with crucial input from the public, to collect and utilize information they otherwise wouldn't be able to collect. "It's not always a straight-forward process to bring those two different kinds of knowledge together. But that's where the most interesting stuff comes out. If we can't find the link, it's not necessarily because one is right or wrong, it may be that we haven't found the common denominator, or we're not asking the right questions," Gearheard says.
Gearheard's interest in Inuit knowledge began while she was an undergraduate at the University of Waterloo in Canada studying climatology. While participating in a field course in the Arctic, she met a group of Inuit hunters in Igloolik, a small community in Nunavut. "I asked them questions about climate change and weather patterns and found that they had a lot to say."
What started as an undergraduate thesis "turned into a master's degree, then a Ph.D., and a postdoc, and my career," Gearheard says. Over the past 15 years, she has collected oral histories from dozens of hunters and elders, tapping into a wealth of knowledge about environmental changes gathered through observation and from stories handed down through generations. "That kind of data is so valuable," she says. "It's like the data set every scientist wishes they had, to have observations year-round over 30, 40, or 50 years or more."
Now a research scientist at the University of Colorado's National Snow and Ice Data Center, Gearheard herself makes year-round observations in Clyde River, a community of 900 people in the territory of Nunavut. "It's sort of a reverse model for most Arctic researchers, who are based at their institutions and spend a few weeks or months in the Arctic each year. I spend all of my time here, and a few weeks out of a year I'll go south to work on campus and meet with colleagues."
Over the years, Gearheard has built a network of connections by working with indigenous organizations to identify people in Greenland and Alaska who are interested in working with scientists. She uses these connections to link local people with scientists studying snow or sea ice or weather. "We develop projects so that Inuit are members of the research team at the same level as scientists," she says. "Just like the scientists, the Inuit members are paid, and they contribute to the research design, fieldwork, analysis, and presentation of the results."
One of the best ways for the Inuit and scientists to work together is to share common experiences, she says. "It doesn't matter if you're a glaciologist or if you're a hunter; if you're standing on the ice, you will have something to say about it, based on your experience. And at that level, people can really connect."
Her work is not all about connecting the locals' qualitative observations with scientists' quantitative ones. She also gets locals involved in scientific data collection. Recently, she linked local sea-ice experts with scientists to develop a way to monitor sea ice quantitatively. Sea ice–measurement stations were installed in Clyde River; Qaanaaq, Greenland; and Barrow, Alaska. Community members monitor the stations and collect data about the ice and snow thickness and temperature, and combine it with their own observations. "Working with them just makes sense," Gearheard says. "They can take observations in January and describe what's going on in the environment any time of the year, not just during the science fieldwork season."
Just as Inuit hunters are helping researchers document the effects of climate change in the Arctic, thousands of nonscientists worldwide are contributing observations of a different sort from their own backyards. Rick Bonney, director of program development and evaluation at the Cornell Laboratory of Ornithology, oversees a dozen programs that rely on information from amateur birders to track thousands of species in North America. The studies have allowed scientists, and the public, to see which populations are growing, shrinking, or staying the same.
In eBird , the lab's most extensive project, participants submit checklists of bird sightings online. Researchers use this information to figure out where the birds are over time. Just last month, nearly 2 million records were submitted. "We can draw a map of the bird distribution and tell you something about the numbers of birds in any particular location for anywhere on the continent," Bonney says. "That's something we couldn't do without the help of volunteers." And in early June, eBird went global. The project now tracks 13,000 species in real time.
This article is part of a two-story package on scientists collaborating with the public. In the companion piece , Lucas Laursen writes about how cliff climbers, bird watchers, and Web-based star gazers are contributing to our understanding of the world and universe.
Bonney says the tracking efforts require a team of scientists working on the back end to design and manage databases, oversee collection procedures, and "handle large, messy data sets."
With thousands of people collecting information, data reliability is very much an issue. The statistical methods used to analyze the data must account for the reliability of the data, piece by piece. If someone enters too many birds, or a species of bird that is out of its range, the record is flagged and sent to one of 300 regional "editors," skilled volunteers with intimate knowledge of the birds in their region.
An editor will contact the birder to discuss the sighting, or seek a photo. "We always want to strike a balance. We want to be able to talk to the people [who] submitted data that looked erroneous, but we don't want to question everything. We try to keep those filters set to reflect the right kind of compromise in weight." Fortunately, with so many volunteers contributing data, the sheer volume washes the information, Bonney says. "We're not looking for exact numbers; we're looking for trends. And we can see those."
The data are used for publications that appear in top scientific journals; the ornithology lab includes a long list of publications . Furthermore, the data are available on the Web site for anybody, including the volunteer participants, to access and use. "One of the worst things you can do is collect a lot of data and then not do anything with it, because that kind of betrays the confidence of the participants," Bonney says.
Maintaining trust between scientific and nonscientific data collectors is especially important in cases for which research motives might be questioned, says Shelly Tallack, a research scientists at the Gulf of Maine Research Institute in Portland, who has spent the past 12 years collaborating with fishermen.
"One concern you hear time and time again from fishermen when there is a research project going on is that they worry that the data will come back to haunt them or make things even worse for them," Tallack says. "I think a lot of that concern comes from years earlier when there was much less transparency about what the data was showing and who had access to it."
Tallack relies on commercial fishermen to supply boats and expertise on finding the fish. The ideal collaboration, she says, involves nonscientists at the outset. "It might even be their research idea. In many cases, they just need to know how to go about collecting the information in a scientific manner."
Seven years ago, Tallack came to Maine to manage the Northeast Regional Cod Tagging Program, the largest such program on the eastern seaboard, which aimed to track the movements of Atlantic cod in the Gulf of Maine and beyond. The effort involved 106 vessels and included help from 250 fishermen who tagged and released 114,473 cod. More than 1000 fishermen recaptured and reported information about the tagged fish. Tallack coordinated the groups and helped train fishermen for the tagging operation.
The study was designed to last 2 years but was extended when it exceeded expectations, Tallack says. "People didn't expect the data to be as good as it was, and they didn't expect as much cooperation from fishermen. But once we were a year into the study and we had 50,000 tagged cod out there, people started to really pay attention."
As demand for the data grew, so did the need to make them fit standard fisheries models. Initially designed to track cod movements in three different cod management zones in the United States and Canada, the study evolved into an effort to help determine whether the zones contained just one fish stock or several. "When we started the project, we had no expectation that it would be used for stock assessment," Tallack says. "It meant that we then had to be creative in figuring out how to use the data for different purposes than it was initially intended for."
Tracking the movements of the fish meant relying on the compliance of fishermen to report the tags. "Not everyone wants to do this, or maybe they don't see the tag or forget to write down the information," Tallack says. Tallack spent 18 months running the fisherman-collected data through sophisticated models to see how they lined up with traditional data. She consulted stock-assessment scientists at the National Oceanic and Atmospheric Administration Fisheries Service. Tallack recounts that the consultants said, "The data [are] interesting, but we don't know how to use [them] in current models."
"I think people are trying to improve the data by including less-traditional data sets, but the challenge is in how to bring that data in to the models that have been running for 15, 20 years. This is an area where the modeling itself may need some improvement." Still, she expects progress. "Had we known from the outset that there would be a lot of quantification, which there ended up being, then I think the study itself would have been designed differently."
Despite the as-yet-unmet challenge of integrating this unconventional data into more conventional studies, the results have proven useful in assessing cod growth and movement in three large cod-management zones -- the Gulf of Maine, Georges Bank, and the Bay of Fundy. Last fall, Tallack published  an article in the journal Fisheries Research that used her collaborators' findings on cod growth. References to the data were made at stock-assessment workshops both within the United States and in Canada, and the work has appeared in technical papers.
Photo (top): Scientists from the USDA’s Agricultural Research Service often consult with farmers and agricultural extension agents on research questions affecting crop yields. (Scott Bauer. Courtesy: Agricultural Research Service/USDA)
Susan Gaidos is a freelance writer based near Portland, Maine.