A new online science project from British climate researchers is tapping into the power of crowdsourcing to enrich the historic climate record. In the next six months alone, Old Weather’s citizen scientists may process data that would take a single researcher 28 years to tackle.
Through OldWeather.org, which launched in late October, volunteers log weather observations from more than 200 World War I–era British naval ship logs. Scientists will use that data to paint a clearer picture of past weather and make predictions about the weather of the future—including potential impacts of climate change—and to test and refine climate models.
“We built models, but we want to know if they’re any good,” said Philip Brohan of the Met Office: Hadley Centre, one of the UK’s leading climate change research centers, and one of the climate scientists behind Old Weather. “We need big databases and long records.”
An ocean of meticulously recorded weather observations is locked up in ships’ logs from countries around the world and could feed such databases. The sheer volume of data is too large for scientists to process on their own. That’s why the climate scientists turned to Zooniverse, a popular network of interactive citizen science projects investigating galaxies, the sun and the moon developed by a team including Chris Lintott, an astronomer from the University of Oxford, that has lead to a series of scientific discoveries. Old Weather is its first nonastronomy project.
Citizen science is nothing new, Lintott said, but while traditionally laypeople have supplied observational data, now scientists need their help processing it.
How It Works
Old Weather volunteers don't need any climate or historical expertise—just the ability to decipher the sometimes-spidery handwriting of early 20th-century sailors. People are more adept than computers at reading and interpreting handwritten logs.
After selecting a ship, a user sees a scanned image of an actual page from the the ship’s logbook. Zooming in with a built-in magnifying tool, the user finds the weather data and types it into a color-coded pop-up box. Volunteers start out as “cadets” but can be “promoted” all the way to ship captain as they digitize more pages.
For the initial project, volunteers are recording information from the logs for 238 ships, including wind speed and direction, air and water temperature and observations about events on the ship. To ensure accuracy, multiple volunteers log each page.
The information gathered through the project will be processed by Hadley Centre’s scientists and entered into international historical weather databases, where it can be accessed and used by other researchers.
“There’s a huge need” for quality historical weather data, especially from the world’s oceans, since most climate records are from land-based data, said Brenda Ekwurzel, assistant director of climate research and analysis for the Union of Concerned Scientists, an environmental advocacy organization. “We don’t have the type of information we’d like.”
One group that likely will employ Old Weather’s data is climateprediction.net, a project of University of Oxford and other UK institutions that bills itself as "the world's largest climate forecasting experiment." It's a distributed computing project that uses the collective power of a network of volunteers' computers to test the accuracy of climate models.
"Climate models have uncertainty in them,” said Hiro Yamazaki, a research scientist for climateprediction.net. The observational data from Old Weather can compared to simulated climate models and ultimately fine-tune them.
While understanding how weather has behaved in the past offers scientists a better understanding of how it may behave in the future, the Old Weather project is not specifically focused on understanding climate change impacts.
But the data will feed into scientists’ predictions of future changes, which includes long-term climate change effects, Brohan said.
The aim of the project is not to inform policy or spur action on climate change, but to improve climate prediction models, he added. “I’m expecting influence on big policy changes to come through mathematics. We use data to improve models, and models to inform policy.”
Demystifying the Science
Beyond transcribing massive amounts of data, the scientists behind Old Weather said they hope the project will also help demystify climate science for the public.
One problem with public misconceptions about climate change is that people fail to understand what scientists are doing when they use climate models, Lintott said. “One of the things Old Weather does is to help make transparent that process….Getting involved in the process makes it less remote.”
That’s one reason Brian Elliott volunteered to transcribe logs for Old Weather. A research contracts officer at Aston University in Birmingham, England, Elliott said he was drawn by the historical aspects of the project but also liked the idea of nonscientists getting involved in a scientific project.
“I hope that we have all learned something new so far, and hopefully the weather data will be useful in due course,” Elliott told SolveClimate News in an e-mail. “If it opens up the world of science to a wider audience then it will have been a success.”
So far, Old Weather’s hundreds of volunteers have made good headway. In the projects’ first two weeks they’ve logged data from more than 126,000 pages, completing about 10 percent of the initial project, which is expected to be finished within a year. But that could be just the beginning. There are hundreds of thousands of ships’ logs in existence from different historical periods, each waiting to be captured and added to historical weather databases.
“We’re hoping this will expand backwards [in time] and go international,” Lintott said.