buoy in the ocean at sunset

There is a particular spot in Oregon’s ocean that has been under intense surveillance for decades. The electronic eyes of sensors and cameras and swimming robots have been trained on it, collecting data on temperature, salinity, currents and more. The pulse of this spot in the ocean, known to many as NH-10, serves as a stand-in for the health and status of the northern California Current ecosystem, which extends from roughly Cape Mendocino to British Columbia. Multiple monitoring programs over the course of decades have collected data at NH-10, located 10 nautical miles from shore, using technologies old and new. These programs sample at differing intervals, lengths of time, and depths, for different parameters.

What happens to these disparate data sets collected at NH-10? Who uses them? What can be learned from them? The answer, until recently, has been that it was actually a bit tough to access and interpret these data: Each individual data set was available, if you tried hard enough to get it, but no one had ever stitched them all together to enable telling long-term stories.

Recently, a team of scientists at CEOAS got out their data management knitting needles and did just that: With funding from NOAA and NSF, they combined a half-dozen NH-10 data sets into one much bigger set that describes the basic attributes of that spot in the ocean over the course of about 25 years. The mega data set is now usable and freely available online.

“With a year or five years of data, you can ask certain questions; you can observe and describe an event, for example. But with 20 or 25 years of data, you can ask different questions, and start to look at longer-term variability at this location, asking, ‘have we seen anything like this before?’” explains Craig Risien of the Ocean Observatories Initiative and lead author on the paper describing the data stitching project.

buoy being launched
An Ocean Observatories Initiative team deploys a monitoring buoy at NH-10. Photo: Kim Kenny

The data sets the team worked with originate from a half-dozen sampling programs. The oldest of the data were collected in 1997 – 2004 by the GLOBEC program, funded by NSF and NOAA, which examined the impacts of climate change on coastal marine organisms throughout the world’s oceans. After that came an alphabet soup of other monitoring programs, some of which overlapped: OSU’s National Oceanographic Partnership Program, OrCOOS (Oregon Coastal Ocean Observing System), which was absorbed into NANOOS (Northwest Association of Networked Ocean Observing Systems), deployed surface moorings (instruments floated at or near the ocean’s surface on a buoy anchored to the bottom) among other technologies; and the NSF-funded OOI Endurance Array, which is operated and maintained by CEOAS as part of a broader national initiative that deploys monitoring instruments on stationary buoys and other platforms in the Atlantic and Pacific Oceans.

These programs have collected months and years of data. Even with the team’s choice to narrow the data set down to three variables – temperature, salinity and current velocity — it’s a lot of data. The task of wrangling the nearly four-gigabyte data set fell mainly to CEOAS Research Associate Brandy Cervantes.

There were some minor challenges, Cervantes says, to combining the data sets – some gaps, some data that she knew existed but could not access – but it was ultimately fairly straightforward to pull it all together. It had just never been done. Cervantes is excited to have it all in one place.

Now that the data set is knitted together, it can be used to ask endless questions about trends and health of Oregon’s ocean. Some trends occur on such long time scales that they don’t emerge until decades of data are examined. Some phenomena are anomalous, but it would be hard to know how unusual they are without long-term information.

Among the first researchers to use the combined data set to seek answers has been Cervantes herself, and she has focused on just such anomalies – the marine heat waves that have plagued the North Pacific between 2014 and now. “These events are typically studied using satellite data,” she says, “But that doesn’t provide any subsurface information,” which is needed to fully characterize marine heat waves.

Her analyses thus far have revealed that the two marine heatwaves of 2014-2016 and 2019-2020 represent a novel increase in heatwave frequency and duration. The 2014-2016 incident was particularly severe because the signal of warm water was observed all the way to the bottom at NH-10, rather than just at the surface. “Heat waves in the Northeast Pacific are not necessarily getting hotter throughout the water column, but they are getting longer, and increasing in frequency,” she says.

The data have been made available on a data sharing site, and Cervantes and Risien feel the stitched dataset will be a valuable asset for researchers studying the northern California current system. For example, the salinity data could be used to look for trends in freshwater inputs from the Columbia River. Low-oxygen events, which seem to be on the rise in Oregon’s coastal ocean, could be correlated with currents. And the team also hopes the data set will be able to be used by educators, from high school to graduate school, to teach about data analysis and ocean science.

With the completion of this data stitching project, the team is looking north for their next endeavor. They are undertaking a similar project with data collected in the Olympic Coast National Marine Sanctuary in Washington, and look forward to comparing the two data sets. “The assumption has been that NH-10 can be used as a proxy for what's going on regionally in the Northern California current system. And we're going to ask, is that true?” explains Risien.

Another future step might be to incorporate more data sets, like those collected by autonomous swimming gliders, providing more spatial coverage of the region, or more parameters, like pH and oxygen concentrations. Over time, data sets collected in Washington and California could be added, until a blanket of data is created, describing the entire California Current system. The pattern in that blanket will become more and more clear, the team hopes, providing insight into how this critical region of the ocean is changing.

Top left:  a regional map showing the location of NH-10 (the white square). The lower three panels show data availability in time and depth for the temperature, salinity and current velocity data taken at NH-10. Color indicates the research program responsible for collecting the data. Figure: Risien, C.R, B.T. Cervantes, M. R. Fewings, J. A. Barth and P. M. Kosro. 2023. A stitch in time: Combining more than two decades of mooring data from the central Oregon shelf. Data in Brief 48 (2023). DOI: https://doi.org/10.1016/j.dib.2023.109041.