In an age of mechanized, digitized labor there are still some jobs where humans get it done better or more cheaply than just about any machinery—particularly when that human labor is crowdsourced by volunteers.
University at Buffalo Geologist Chris Lowry figured that out when trying to collect basic information on the water level of streams across a large watershed in western New York, an endeavor that would eat up cash using machinery or time using labor from the lab. After reading an article about a researcher who used crowdsourcing to get the public to help monitor roadkill, "I was like 'If these people can get people to help out with their research, why can’t I get people to help out with water level measurements?'" explains Lowry.
He started simple, printing out a half sheet of paper that said "'Please text me the water level,' and it had a phone number. "And then I bought a giant ruler, I brought this into the stream, I put this sign on top, and then I just waited for someone to send me a text message," he explains. "And sure enough, a couple people sent me text messages."
That basic idea turned into the pilot project CrowdHydrology at nine New York freshwater sites, starting in 2011. Now, with support from the U.S. Geological Survey, the project will expand to more than 50 new sites across New York, Michigan, Wisconsin, and Iowa. The data can help fill in gaps in data collection as budget cutbacks mean the USGS is discontinuing monitoring of certain streams.
When citizens are contributing data, quality is always a concern, but Lowry says control tests—where a pressure transducer measures water levels in the same sites where people are measuring by hand—show that "people who send us text messages do a really really good job." The level of error turned out to be to be as small as one tick mark on a ruler.
Lowry has found that engaging local communities is the key to getting a high volume of texts from any given site—more so than just foot traffic. And locations where passersby are more likely to take an interest in science—like a nature center—have worked best.
"I really think that as scientists we may just be on the cusp of crowdsourcing scientific data. I think there’s going to be a big boom in the future for using these kinds of methods," Lowry says.
More than just expanding CrowdHydrology’s physical territory of the project, the next phase will enhance its technical components as well, with a smartphone app that could geolocate the crowdsourced texts. (Currently, a program that Lowry designed with Michael Fienen of the USGS processes the texts, reverses certain damage that the iPhone’s auto-correct does—like changing 'NY’ for New York into 'my’—and feeds the data to the project’s website.)
Lowry sees the project as being useful for getting people to think about their local water systems and the way that weather and events can change the water supply. "On one side we’re using [the project] for cutting edge research. On the other, we’re using it as this outreach tool to foster the next generation of scientist."