Using Peat Oxygen Isotopes to Elucidate Patterns of Sea Ice Extent over the Last 5,500 Years in the Bering Sea, Alaska

Release Date:

This article is part of the Fall 2020 issue of the Earth Science Matters Newsletter.

The Bering Sea supports some of the most productive ecosystems in the Arctic. These ecosystems exist in part because winter sea ice expands and covers much of the Bering Sea in the winter months and then retreats completely in the summer. Satellite records since 1979 reveal variation in average winter sea ice extent that is driven by the dominant wind direction, with northerly winds corresponding to greater winter sea-ice extent and vice versa. This multi-decadal timeseries did not, however, reveal any long-term trends in sea-ice extent until the winter of 2018, when sea ice was nearly absent by the end of February, two to three months earlier than average. Because of the short length of the satellite record, it was unknown if the minimal sea ice that year had precedent.

A peat core collected in 2012 from St. Matthew Island, a small uninhabited island in the central Bering Sea, held a key to answer the question on whether the early sea-ice loss in 2018 was anomalous. The core originally was collected to understand patterns of climate variability in the mid to late Holocene (~5,500 years ago to present). USGS researchers extracted cellulose from peatland plants preserved in the core and sent the samples to collaborators at the University of Alaska Fairbanks for oxygen isotope analysis. Bryophytes (mosses) comprise a large component of northern peatlands, and lacking vascular structure, they passively absorb water from their surroundings. Past research on water isotopes in peat suggests that peatland plants record precipitation conditions during their lives. This means that as layers of bryophytes die and get replaced with new growth each year in peatlands, the buildup of dead plant material can be analyzed to provide a record of precipitation through time.

With help of a model that tracks the isotopic fingerprint of precipitation, this study determined that lighter isotope anomalies were associated with years where winds dominated from the north, expanding sea ice into the Bering Sea. Whereas heavier isotope anomalies were associated with years where winds dominated from the south, coming off the North Pacific Ocean.

The authors found high variability in the oxygen isotopes over the last 5,500 years, but a general trend toward heavier oxygen isotopes suggests a long-term decline in the sea-ice extent. They found that this long-term decline and variability within the record corresponded to both increasing winter insolation (the amount of solar radiation reaching an area) and changes in concentrations of atmospheric carbon dioxide (CO2).

graph of sea ice extent and atmospheric carbon dioxide concentrations

Sea ice extent inferred from peat cellulose isotopes (light blue) versus atmospheric CO2 concentrations (purple). Sea ice declines as the level of atmospheric CO2 increases. From Figure 5 in (Jones et al., 2020).

(Public domain.)

Although the initial peat core record extended only to 2012, the researchers worked with scientists at the Alaska Maritime National Wildlife Refuge, who visit the island twice each decade to evaluate bird populations. Refuge scientists collected modern bryophyte samples during their visit in 2018, the year with the anomalously low sea ice extent. By comparing oxygen isotopes from those mosses with the peat-core record, USGS researchers confirmed that the sea ice extent in 2018 was the lowest in at least the last 5,500 years.

Comparison of reconstructed sea-ice extent with past atmospheric CO2 concentrations showed a strong correlation between atmospheric CO2 and sea ice extent. This suggests that, without mitigation of current trends in carbon emissions, the Bering Sea may be ice free year-round within the next century. The correlation analysis also revealed the potential for a lagged relationship, in which the loss of sea ice in the Bering Sea lags CO2 concentrations by several decades, suggesting that in a worst-case scenario, a complete loss of sea ice is already locked in, even with a complete stop in greenhouse gas emissions.

Loss of sea ice in the Bering Sea has implications for the future health of the Bering Sea ecosystem, which supports a billion-dollar fishing industry and indigenous communities that rely on the sea ice for their subsistence harvest. The lack of sea ice would accelerate coastal erosion and storm surge impacts from large Bering Sea storms, which disproportionately impacts indigenous communities living along the Bering Sea coastline. Sea-ice loss also would increase regional heat absorption, leading to an acceleration of high latitude warming that can thaw permafrost and impact terrestrial ecosystems. These changes in the Bering Sea would also accelerate the loss of Artic Ocean sea ice extent, which has been rapidly declining over the last several decades.

The paper “High sensitivity of Bering Sea winter sea ice to winter insolation and carbon dioxide over the last 5500 years” is published in Science Advances and is available online here: DOI: 10.1126/sciadv.aaz9588

<< Back to Fall 2020 Newsletter

Related Content