Less Ice, More Blooms?
ECCO-Darwin helps us better understand ties between Arctic ice & life
Our climate is changing. And nowhere it is more apparent than the Arctic Ocean. Why is it unique... and susceptible to change?
It's the smallest
With an area of about 6.1 million square miles, the Arctic Ocean is about 1.5 times as big as the United States. The Pacific is about 10 times larger, covering more than 60 million square miles.
It's the shallowest
The average depth of the Arctic Ocean is about 1,200 meters (4,000 feet). The Pacific is our deepest ocean, with an average depth of approximately 4,000 meters (13,000 feet).
It's the freshest
It has the lowest salinity – a measure of seawater's concentration of dissolved salts – of any of our ocean basins due to low evaporation, heavy fresh water inflow from rivers and streams, and limited connection to surrounding oceans with higher salinity.
How Low Can Ice Go?
Every summer the Arctic sea ice melts down to what scientists call its minimum before colder weather begins to cause ice cover to increase. The sharp decrease of Arctic sea ice over the past decades is captured in a graph of sea ice minimum area over time.
sea ice – frozen ocean water, which forms, grows, and melts entirely in the ocean; in contrast, icebergs, glaciers, and ice shelves that can be found in the ocean but originate on land
Diminished Arctic sea ice cover impacts the ocean. Why? The open ocean absorbs sunlight more effectively than bright, reflective sea ice. Melting sea ice adds fresh water to the ocean, lowering salinity. (Conversely, the formation of sea ice increases salinity since the salt is "left behind" in the seawater.) Open water is also more susceptible to being pushed by winds than ice-covered areas.
All of these affect the physical state of the ocean. But what about its biology? We use a high-resolution computer model that captures the physical and biological state of the Arctic Ocean, known as the ECCO-Darwin model.
In this study, we examine month-to-month changes from 2006 to 2013, investigating the relationship between sea ice melt and blooms of tiny algae that are suspended in our ocean, known as phytoplankton.
When sea ice cover recedes, sunlight can reach the surface of the ocean and the phytoplankton that float on it, allowing them to photosynthesize and thrive after a long period of being covered. This produces fuel for other species. Polar species from clams and krill all the way up to walruses and whales rely on these timely blooms for their food sources. – NASA Phytoplankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission
What do phytoplankton blooms look like in ECCO-Darwin data? They are mapped as high concentrations of chlorophyll, a green pigment found in phytoplankton. Below is the first year of our study: 2006.
Like this movie? Click here to check out an ECCO visualization of sea ice! (It's very cool.)
What's Up Around the Arctic?
Although the Arctic Ocean is relatively small, it has an outsized impact on our climate. In our modeling study, the Central Arctic Ocean is defined as the area north of 75° latitude. Touching five countries with complex shorelines, the remaining area is parsed into seven seas (Nordic, Barents, Kara, Laptev, East Siberian, Beaufort), Baffin Bay, and the Canadian Archipelago.
Below, we include graphs of sea surface temperature, open water area, sunlight available for phytoplankton growth, and chlorophyll at the sea surface. Color-coded data represent monthly averages in the years 2006 to 2013.
Central Arctic
Mostly far away from shorelines, the central Arctic Ocean is very cold. Even at its warmest – September peak in 2007 (red dashed line in top graph) – it still less than 1°C (34°F). In winter months, sea surface temperature is well below 0°C. Why is it still liquid? Seawater's salinity keeps it from freezing, even at -2°C.
Timing of the peaks in open water area (second graph from top) coincide very well with the warmest months.
Sunlight (third graph from top) is a rare commodity at the top of the world. The lack of sunlight can severely limit phytoplankton growth in the Central Arctic.
Nonetheless, our study indicates weak phytoplankton blooms (bottom graph) occur in the fall, well after the summer peak in sunlight.
Nordic Seas
Let's go southward, to the area located between 65°N and 75°N, lying between Greenland and Norway known as the "Nordic Seas."
In 2007 (red dashed line in top graph), our study shows the sea surface warming up to 10°C (50°F).
This corresponds to an increase in open water area (second graph from top), which means a corresponding decrease in sea-ice cover.
During 2007, the concentration of chlorophyll (bottom graph) indicates a spring phytoplankton bloom. Note that this phytoplankton bloom occurs well before the peak in sunlight used by phytoplankton (third graph from top).
Even though phytoplankton growth can be limited by a lack of sunlight, our simulation suggests that seawater warming in the Nordic Seas is the main factor boosting phytoplankton growth.
Months of darkness and persistent cloud cover make it challenging to capture satellite images of the Arctic Ocean. We include an image for each of our remaining areas, some of which correspond with our study years (2006-2013) while others were acquired later.
Barents Sea
Moving eastward, the Barents Sea region extends to the eyebrow-shaped Severny Island.
Let's focus on the concentration of chlorophyll (bottom graph). While there is a consistent spring bloom, many years have a second phytoplankton bloom in the fall season.
What might be happening to trigger these unprecedented fall blooms? The high amount of open water area (second graph from top) in the fall makes this area prone to wind-driven mixing. Such action can stir up nutrients from depth, fueling phytoplankton growth.
Kara Sea
Moving east to the Kara sea, our model shows that, for some years, there is a second peak in the concentration of chlorophyll during the fall.
The fall bloom may be caused by the delay of sea-ice formation and the persistence of open water conditions.
How might this work? The timing of the spring and fall blooms is highly correlated. Phytoplankton growth in spring can deplete available nutrients in the sunlit upper ocean. A potential fall bloom can form if there is sufficient time for wind-driven mixing to deliver nutrients to the upper ocean.
Laptev Sea
Our next stop is the Laptev Sea where we'll focus on the year 2007 (red dashed line in all graphs). As shown earlier , that year's sea ice minimum had been the lowest recorded since 1979.
In 2007, sea surface temperature reached 4°C (39°F) in July. However, in all other years studied, the Laptev Sea did not reach 4°C until August.
2007's severe Arctic sea-ice loss also shifted the timing of the open water area compared to the other years we studied. In response, the concentration of chlorophyll peaked much earlier in 2007 than other years.
East Siberian Sea
In 2007 (red dashed line in all graphs), warming at the sea surface culminated with severe sea-ice loss – and high open water conditions – in September. In fact, the prolonged period of increased open water conditions extended though November and into early December.
What was happening with the concentration of chlorophyll at the sea surface in 2007? It had peaked earlier and was on the decline during the fall.
In 2007, with plenty of open water and sunlight available, it suggests that phytoplankton growth growth was limited by a lack of nutrients. The ECCO-Darwin model – which also includes nitrate and phosphate – demonstrates that these nutrients were depleted earlier than usual, due to an earlier start of the phytoplankton bloom.
Furthermore, our analysis confirms that phytoplankton growth in the East Siberian Sea is mainly limited by the scarcity of nitrate.
Chukchi Sea
To understand in more detail the mechanisms and factors that control phytoplankton blooms in the Arctic Ocean, we focused on the Chukchi Sea during two years: 2008 and 2009.
Why these years? Because of the distinct shapes of their concentration of chlorophyll graphs. In 2008, our model simulates an increase in concentration of chlorophyll from August to September. In 2009, it shows a single bloom that peaks in July.
During the fall of both years, the model shows similar values of open water area but there is a difference in the surface concentration of nitrate, as shown below.
Comparing these two years, we concluded that the main factor causing phytoplankton blooms in the Chukchi Sea is the effect of wind mixing and the higher availability of nitrate in 2008 than in 2009.
Beaufort Sea
Located north of Alaska, Yukon, and Northwest Territories, the Beaufort Sea has been the subject of many field campaigns, including Salinity and Stratification at the Sea Ice Edge (SASSIE) during 2022.
The ECCO-Darwin model simulation shows changes in timing of phytoplankton blooms during the 2007–2012 period, particularly when compared to the first (2006) and last (2013) years of our study period.
From 2007 to 2012, our model shows an earlier increase in open water area. This corresponds to an earlier shift of concentration of chlorophyll at the surface. What was the underlying cause of these shifts in timing? Variations in the Arctic climate.
Want to learn more about river outflow on the Beaufort Sea? Check out this StoryMap:
Canadian Archipelago
With dozens of islands, the Canadian Archipelago had unusually high sea surface temperatures in 2007 (red dashed line in top graph). Note, however, that the range in overall temperatures in this region is consistently low: less than 1°C (34°F).
This 2007 warming corresponded with severe ice loss, seen as an earlier-than-normal peak in open water area.
Baffin Bay
Located between Canada and Greenland, Baffin Bay is known for its floating icebergs.
The ECCO-Darwin model results show that, in addition to the spring phytoplankton bloom, many years have also have a distinct bloom later in the year.
Want to find ocean color images like the ones featured in the previous section? Check out the PACE Ocean Color Images: Interactive Map.
That Bloomin' Time Again?
In some of the areas we studied, the ECCO-Darwin model simulates a second phytoplankton bloom during the fall. To evaluate the accuracy of our model, we calculated the percentage of fall plankton blooms (FPBs) from our study (right in slider below) against those estimated by Lewis et al. (2020) based on satellite-derived data (left in slider below).
In general, the largest values of FPB occurrence (∼80%) are located in the same regions of the Arctic Ocean. However, ECCO-Darwin tends to overestimate FPB occurrence (40%–60%) when compared to satellite-derived estimates (20%–40%).
We find general satellite data-model agreement in terms of distribution of the FPB occurrence during our study period. Most of the FPBs occurred in the Eurasian sectors of the Arctic Ocean (i.e., Nordic Seas, Barents Sea, Kara Sea, Laptev Sea). Both ECCO-Darwin and satellite observations agree on the lack of FPBs in the East Siberian Sea.
What Did we Learn?
Climate change caused observed modifications of sea-ice extent and its seasonal cycle in the Arctic Ocean during our study period.
Our simulations show that seasonal changes in sea-ice formation directly impact the timing of phytoplankton blooms in the Arctic Ocean.
The earlier break-up of Arctic Ocean sea ice triggered earlier spring phytoplankton blooms. Also, a delayed formation of sea ice during fall caused the occurrence of a second phytoplankton bloom in the fall.
We captured the observed seasonal changes because: (1) our sea-ice model can correctly simulate the climate-related changes in sea-ice extent; and (b) our plankton ecosystem model can realistically simulate the response to changes in light and nutrient availability caused by sea-ice alterations. Some of these changes might be a window into the future for Arctic Ocean phytoplankton dynamics responding to climate changes if sea ice continues to decrease. – Manfredi Manizza, Scripps Institution of Oceanography, Study Lead
Other Resources
Below is an ECCO visualization of effective sea ice from 13-Sep-11 to 15-Nov-12. Click here to download or see other versions.
Want to run your own ECCO simulation? Check out our GitHub repository.
Manizza, M., Carroll, D., Menemenlis, D. et al. (2023) Modeling the Recent Changes of Phytoplankton Blooms Dynamics in the Arctic Ocean , J. Geophys. Res. Oceans, 127, e2022JC019152, doi: 10.1029/2022JC019152