It was quite a surprise: Two reports on Chesapeake Bay dissolved oxygen levels in 2017 came to starkly different conclusions. The Maryland Department of Natural Resources reported improvements and a vastly reduced dead zone. The Virginia Institute of Marine Science found oxygen conditions at their worst since 2014.Climate change raises water temperatures which, in turn, increase the likeliness of algae blooms that block the sunlight needed by the Bay’s grasses to grow. (Dave Harp)

Scientists at both organizations were caught off guard last fall when the seemingly contradictory findings were released within a few weeks of each other and sent them scrambling to analyze the cause. 

There were significant differences in the two approaches. 

The DNR results were based on water quality samples collected during an annual monitoring program from ships at selected points in Maryland’s portion of the Bay’s mainstem during specific times of the summer. Samples are normally collected during seven cruises — two each in June, July and August, and one in September. Because of high winds, some 2017 measurements were delayed or skipped.

The VIMS analysis stemmed from a computer model rather than real-world sampling of dissolved oxygen concentrations. The model uses decades of historical data to predict oxygen conditions throughout the entire Chesapeake mainstem and its tidal tributaries based on a variety of real-time factors including temperature, salinity, water movement and other data, such as wind.

The model, funded by the National Oceanic and Atmospheric Administration, was created several years ago by scientists from VIMS and the environmental consulting firm Anchor QEA to predict oxygen conditions throughout the Bay to provide estimates of conditions for anglers and other water users.

After a review of the data, scientists agreed that both pictures were, surprisingly, correct.

“The main story isn’t that they [DNR] are ignoring the tributaries or they are ignoring Virginia. Really the difference is in the timing,” said Marjorie Friedrichs, an associate professor at VIMS. 

Low oxygen conditions are caused by several factors. Nutrients washed into the Bay spur algae blooms. When they die, the algae sink to the bottom and are decomposed by bacteria, which consume oxygen in the deepwater areas. 

Because of differences in salinities and temperature, a barrier typically forms between surface and bottom waters during the summer. As those layers stratify, oxygen in deepwater areas is gradually used up, creating so-called dead zones. But other conditions, such as strong winds, stir the Bay’s water and mix the layers, which replenishes oxygen in deep areas.

Last summer, storms that involved high winds — one included a tornado that hit Kent Island — hampered several of the DNR’s pre-scheduled monitoring trips.

“A lot of our samples were collected right after a wind event,” said Bruce Michael, director of the DNR’s Resource Assessment Service. “As a matter of fact, we had to move our cruise dates around…the wind conditions kept us from going out.”

Because wind reduces stratification, the samples collected during cruises that followed those storms showed improved oxygen conditions.

The VIMS model predicted very similar conditions to what the DNR saw on its monitoring cruises. But it presented a different overall result because it also predicted that the Bay restratified shortly after the wind events abated, and oxygen conditions worsened in deep waters. 

“It’s really a timing thing,” Friedrichs said.

As a result, the snapshots-in-time observed by the DNR monitoring found just 13.6 percent of Maryland’s portion of the mainstem of the Bay suffered from low oxygen conditions, compared with the long-term average of 18.9 percent.

The VIMS estimate found that on average, 14.7 percent of the entire Chesapeake Bay — Maryland, Virginia and their tidal tributaries — suffered from low oxygen conditions during the summer. That was the worst of the four years, going back to 2014, for which the VIMS scientists have made estimates. It also found that low oxygen conditions were about 10 percent more widespread than in 2016.

“What we bring to the table is what is happening between the cruises, and putting the cruise data in context,” Friedrichs said.

Michael said the different results don’t mean that the trends reflected in monitoring data over time are failing to reflect what’s happening in the Bay. “We have mixing events during most summers,” he said. “This summer, they just seemed to be more frequent.”

To provide a better picture in the future, Michael said he would like to place a couple of monitoring devices in deepwater areas to provide continuous water quality data if funding is available. That would help provide new information about what is happening between cruises, as well as information that could help VIMS scientists refine their model and improve its estimates.

Bay water quality monitoring historically has been limited because of its expense — it typically requires both labor, boat time and expensive equipment. 

But it is also the monitoring conducted by Maryland — and counterparts in Virginia — that provide the basis for determining whether the Bay is attaining its cleanup goals. Results show that the amount of deepwater areas meeting water quality goals has increased steadily over the last 30 years.

Additional monitoring would not likely show changes in that overall trend, but it would help provide a better understanding of how normal or abnormal are events such as last year’s extreme winds in affecting Bay water quality, said Peter Tango, a scientist with the U.S. Geological Survey who coordinates monitoring for the Chesapeake Bay Program. It might also provide a clearer picture of what was happening in more local areas, he said.

“The monitoring program could benefit from having weekly measurements or subweekly measurements to ascertain that type of dynamic in more detail,” Tango said. 

Meanwhile, to end confusion over the findings, scientists from the DNR, VIMS and the Bay Program say that this year they plan to review each others’ data to better understand the full picture when releasing results.