Andrew Beers arrived on the Brown University campus his freshman year with the intention of majoring in English until he changed course and headed into public health. 

But then, he landed in Caroline Karp’s environmental law and policy class, where he wound up mapping flood zones and using sea level rise projections to estimate where the water would be in 50 years in order to imagine how land use should be regulated.

“I always thought of myself as mostly an essay writer,” noted Beers, of Glen Echo, Maryland, who graduated May 24 magna cum laude with a degree in environmental studies, a penchant for mapping and an interest in arctic research.

Next, he heads to the University of Minnesota Cedar Creek Ecosystem Science Reserve, where he will work for the biological field station, investigating how climate change threatens biodiversity.

Karp, her class and the mapping experience not only changed the course of Beers’ academic and professional journey, but also led him to explore the realm of iceberg data and develop a visualization tool, currently going by the name Iceberg Mapper. The project earned Beers the Brown University Library Innovation Prize for 2015.

Beers’ efforts and the accolades for his work reflect both on the powerful impact of an educator and what becomes possible with the collaborative relationships of Rhode Island NSF EPSCoR, nine partner institutions joined together in a collective quest to investigate the impact of climate change on marine life and ecosystems.

Data in everyday findings

Karp, a senior lecturer in Brown’s Center for Environmental Studies, played a pivotal role in Beers’ direction, both as an educator and as his advisor. She introduced him to the Coast Guard Ice Patrol logbooks, which she obtained from a professor at the Coast Guard Academy while they were trying to locate seafarer logs, and formal notebooks kept by anyone — sealers, whalers, fishermen, newspapermen — who spent time recording sea and weather condition and ecologically-relevant information in southern New England waters.

In theory logbooks can offer a sort of a reconstruction of the ecological history of an area. Someone says, ‘I saw this iceberg or animal at this place’ — that’s a data point. Some of the logbooks are very old, from before the turn of the century; as long as people here have been keeping records of navigation hazards and fishing in the North Atlantic.

Karp and her collaborators hoped that fishermen’s logbooks, some of which date back to the 18^th and 19^th centuries, would contain a wealth of ecological data, tracking animals and weather they encountered. While this was often the case for naval vessels, scientific and whaling expeditions, and commercial ventures such as the Hudson Bay Company, it was not true typically of local fishermen.

In tandem with the logs, Karp and Beers dug into the records of the International Ice Patrol, a multi-country effort run by the U.S. Coast Guard to map icebergs after the April 1912 sinking of the Titanic. In the early years, boats would hunt for icebergs and mark the size and location of any found.

At year’s end, ice patrol members pulled together annual reports, about 100 pages long, including occasional narratives about what they observed, such as whales, birds, fish, and the recurring theme of dense fog. They also kept huge charts of iceberg locations. As technology evolved, planes took over the observation and today satellite tracking does the work.

Karp collaborated with another Brown student, Kara Woo, to conduct a text-based search of the Coast Guard logs for biological data in 2010 when they encountered ‘rough weather’ trying to differentiate unique iceberg sightings in the database. Five years later, Beers confronted the same dataset from the perspective of an applied mathematician.

“In theory logbooks can offer a sort of a reconstruction of the ecological history of an area,” he explained. “Someone says, ‘I saw this iceberg or animal at this place’ — that’s a data point. Some of the logbooks are very old, from before the turn of the century; as long as people here have been keeping records of navigation hazards and fishing in the North Atlantic.”

The historical seafarer logs and ice patrol reports provide limited ecological data, but potentially important observations of the effects of industrialization on the Greenland ice sheet between the 1880s and 1940s, before airplane overflights, radar and satellites collected iceberg data, according to Karp:

“The data set is from 1880s to the present, almost our entire industrial period and the human impact on climate. The underlying questions are, to what extent does the number and distribution of icebergs in the North Atlantic correspond with measureable effects of global warming and other processes that affect sea surface temperature?”

Thousands of points of data

Hired by Karp to map the iceberg data from the early seafaring logs and iceberg patrol, Beers said he originally started transcribing the data points into an Excel spreadsheet. He then discovered that Canadian scientist Brian Hill had paid students to read and transcribe every iceberg sighting in the ice patrol records, amassing an incredible 100,000 rows or entries.

Beers also used archival records of the now-defunct U.S. Hydrographic Office, and notices published in maritime newspapers such as The New York Maritime Journal and Lloyd’s List.

All told, the data represent more than 300,000 individual sightings in a 134-year time period. But the challenge remained — what did all those points mean?

What could we learn from the iceberg locations and tracking their movement, albeit very slowly, starting at Greenland and traveling along the Labrador current and into the Gulf of St. Lawrence, where they hit the warm water of the Gulf Stream and melted?

Was there a relationship between iceberg locations and climate change? What, if anything, could these many thousands of points of data tell us?

At first, Beers said he could not see the proverbial forest through the trees; there were so many points of data, it was difficult to frame the reference and make sense of trends. But, he had heard about the data visualization work at Rhode Island School of Design (RISD), where the Edna Lawrence Nature Lab serves as the campus portal for RI NSF EPSCoR.

Last fall, Beers met with the Nature Lab’s Neal Overstrom, director and RI NSF EPSCoR partner liaison, and Jennifer Bissonnette, biological programs designer, seeking guidance on how to visualize the data.

“I said we have these logs — what is the best way to do this, what should I be visualizing?” Beers said. “They were able to give me points on where to go next and how to work with the data I had.”

Beers coded the tool in D3 and Javascript and completed the visualization — http://andrew-beers.github.io/International-Ice-Patrol-Mapper/ — by the end of January. He then set to work on his capstone paper, where he summarized the actual findings and answering the questions of whether climate change had altered the distribution of icebergs and where were the icebergs going: “Icebergs can be used as a proxy for climate change in a way they haven’t before.”

Through Iceberg Mapper, Beers explained, users can overlay multiple years using the right toolbar, and change the time-window using the bottom toolbar. The area represented is the Northeast coast of Canada, including Newfoundland, the Grand Banks, Maine, Greenland, and Iceland.

A work of art & science

Beers’ efforts and the culmination in Iceberg Mapper points to an elemental truth of research, said Overstrom: “The most effective collaborations between artists and scientists around data visualization occur when a visualization not only represents the data, but also critically informs our interpretation, leading to new insights that we may not have seen without it.”

This same confluence arose in the RI NSF EPSCoR studios conducted at RISD, Overstrom added, citing, for example, the visualization tool (http://www.gso.uri.edu/blog/exploring-the-oyster-genome/) that graduate students David Zacker and Jack Lovell developed with Associate Professor Marta Gomez-Chiarri, department chair, Fisheries, Animal and Veterinary Sciences, University of Rhode Island.

The phrase, ‘a picture is a thousand words’ is true in science as well as advertising, especially now given the vast data sets that scientists are working with. Visualization techniques can help deal with complexity to more readily reveal what the underlying data are telling us.

Overstrom also pointed to the work of Ralf Kaehler at Stanford’s Kavli Institute for Particle Astrophysics and Cosmology (http://www.slac.stanford.edu/~kaehler/) and his visualization methods for astrophysical simulation data.

“His videos are fascinating and beautiful, but more importantly they are a tool in data analysis helping his colleagues hone their computer models,” Overstrom said.

Of Beers’ project, Overstrom said the tedious work of inputting the load of information paid off tremendously. Now, individual years can be quickly juxtaposed to look for patterns that otherwise might not readily surface.

“The phrase, ‘a picture is a thousand words’ is true in science as well as advertising, especially now given the vast data sets that scientists are working with,” Overstrom said. “Visualization techniques can help deal with complexity to more readily reveal what the underlying data are telling us.”

Climate change science

For Karp, the visualization informs a compelling story: “It’s spectacular to see all of the iceberg observations, gathered from the 1800s to the present along the entrance to the St. Lawrence Seaway, compiled in one web-based format. Andrew’s Iceberg Mapper allows users to compare observations collected by individual fishermen and mariners and the insurance industry from the 1880s through the early 20^th century with the CG ice patrol data and a variety of remotely-sensed instrumentation since the 1940s.”

You go through your classes and a lot of it is you do what you’re told to do or you work on someone else’s project. This was mine. Now, I understand how research is done and discovery is made.

She said she was intrigued to see whether the data recordings of the early seafaring logbooks lined up with climate change and what ecological anomalies might be divined from pooling together the observations across time.

“It’s one thing to have the observations and another to interpret what they mean,” she noted. “How do you take the data and tell the story, and understand change over time.”

The iceberg data points to a declining Greenland ice sheet, less ice being formed on an annual basis, which means receding glaciers and rising level along with increase of fresh water flowing into seawater and changing the ocean’s composition, according to Karp and Beers.

Beers also reflected on the value added to his educational journey through Karp’s mentoring and the process of collecting the data and putting it into context. He credited his environmental studies degree to her teaching and advising, and Iceberg Mapper to her launching him on the project and continued involvement with his progress.

“You go through your classes and a lot of it is you do what you’re told to do or you work on someone else’s project,” he said. “This was mine. Now, I understand how research is done and discovery is made.”

Beers reflected on the totality of the experience, saying it will guide where he goes and serve as a reminder not to just accept what is seen at first glance, but to look further at what is actually there.

And, he laughed, “It’s also taught me that there is a lot of data out there.”

Story and photos by Amy Dunkle