When Cooperative Institute for Research in Environmental Sciences (CIRES) researchers digitized thousands of marigram charts, historic coastal water-level records dating back to the mid-1800s, they hoped the records could help to improve tsunami forecasting. These analog records often provide the only instrument-based evidence of past tsunami events, making them invaluable for scientific modeling.

What is a marigram?

A marigram is a paper chart that records sea-level changes over time, often at coastal stations or lighthouses. Before digital sensors existed, these analog records were the primary way to track tides and detect tsunami events. Some marigrams date back to the mid-1800s and remain the only instrument-based evidence of historic tsunamis. Today, digitized marigrams help scientists validate tsunami models and improve forecasting.

_{A marigram digitized by CIRES}

“Damaging tsunamis occur relatively infrequently, but as seen with recent events in 2004 and 2011, they can have a terrible cost,” said Aaron Sweeney, CIRES researcher. “These records help modelers validate tsunami propagation and inundation models. Those models, in turn, make forecasts more accurate.”

But digitization alone wasn’t enough. The real challenge was converting thousands of scanned charts into structured, standardized data that could be shared through the National Centers for Environmental Information (NCEI) database. That’s where the University Libraries was able to provide help through a collaborative research grant and the expertise of librarians.

What the Libraries did – Librarians as research partners

“My role was about bridging two worlds: library science and oceanography,” said Erik Radio, the metadata librarian assisting the project. “We brought best practices for metadata and data stewardship, but adapted to scientific standards so the data could live in national repositories.” Radio assisted by:

• Aligning metadata practices with NCEI’s scientific schema.
• Advising on strategies for handling handwritten marginalia—notes like “full moon tonight”—that add context but defy standard fields.
• Mentoring a master’s student in data science, hired through the Libraries grant, who completed the technical extraction and analysis ahead of schedule.

Sweeney added: “We had over 100 applicants for the student job supported by this grant. Erik was great throughout the process, and our student’s work exceeded expectations.”

Outcomes and impact

The collaboration produced:

• A cleaned, structured dataset ready for integration into a federal open-access repository.
• A detailed report laying the groundwork for future publications on tsunami event detection.
• A model for interdisciplinary partnerships that combine domain expertise with library-driven data strategy.

Why it matters

By combining Radio’s expertise in metadata with Sweeny’s oceanographic knowledge, this collaboration turned thousands of analog charts into structured, shareable data. That dataset will feed into national repositories and help scientists validate tsunami models. These models improve forecasting and can save lives and protect coastal communities.

Interested in learning more about collaborative research projects with librarians? Attend the Libraries Research Grant Showcase on Thursday, Jan. 29, from 11:30 a.m. to 1:30 p.m. in Norlin Library. Registration is appreciated but not required.

Register now