In an article published in Science Advances on 26 January, a group of researchers can document that the results of a much discussed study on the content of mercury in meltwater from the Greenland Ice Sheet cannot be reproduced, and that they cannot detect mercury in amounts above the expected level in meltwater from the ice sheet in Greenland.
In 2021, a sensational research article spread concern in Greenland and Denmark. The article very surprisingly concluded that the meltwater from the Greenland Ice Sheet contained enormous – and until then overlooked – amounts of mercury. The article led to alarming headlines in Danish, Greenlandic and international media about the possible consequences for the population, animals and the environment.
“The original study caused a great stir when it reported alarmingly high concentrations of mercury. Measurements showed that there were 100 to 1000 times the concentrations we would normally expect to find in freshwater systems in Greenland,” says Christian Nyrop Albers, Senior Research Scientist in Environmental Geochemistry at The Geological Survey of Denmark and Greenland (GEUS).
The cause of the high mercury concentrations was not determined with certainty, but the study assumed that it must originate from the bedrock under the Greenland Ice Sheet.
Quick action
The results quickly became a hot topic in the scientific community and were initially met with skepticism. Could it really be that this huge mercury release had gone unnoticed? Immediately after publication, the researchers from Aarhus University (AU) and GEUS chose to collaborate on taking meltwater samples from the same area as the researchers in the original study, to check whether they could measure mercury concentrations at the same level. However, the results were quite different. All the samples showed a result of around or below 1 nanogram of mercury per litre of water, which is by no means alarmingly high.
“1 nanogram per litre is well within the expected concentration level in natural systems. There is a natural occurrence of mercury in the environment from e.g. volcanic eruptions, and there is a global spread of mercury from e.g. the burning of fossil fuels, which causes this level of background concentrations,” says Christian Nyrop Albers from GEUS.
The researchers from AU and GEUS then contacted the journal Nature Geoscience about the lack of reproducibility of the results and the significance this might have for the credibility of the original article. The information was rejected by Nature Geoscience, as the new results did not have enough samples and/or did not bring enough relevant new information to the readers.
With support from the Danish Environmental Protection Agency’s ‘Environmental support for the Arctic’, the researchers from AU and GEUS therefore expanded the field investigations through the project ‘KvikSAND? Verification of claims about extremely elevated mercury concentrations in meltwater from the Greenland Ice Sheet’.
Project KvikSAND involved seasonal sampling near Kangerlussuaq in West Greenland as well as sampling at 21 meltwater outlets along 500 km of the western edge of the Greenland Ice Sheet. The water samples were taken in parallel sets of samples, each of which was analysed at three independent laboratories, including two leading mercury laboratories at recognised universities in Canada and France.
Unequivocal result
The results of these combined analyses are very clear: the mercury concentrations from the extended study were at the same low level as the initial control measurements.
“The conclusion of our study is that meltwater from under the Greenland Ice Sheet has a very low mercury content. That means that there is a very limited risk for both local communities and the natural environment in Greenland,” says Christian Juncher Jørgensen, Senior Researcher at AU.
The authors behind the article say that they cannot immediately explain the elevated concentrations of mercury that were shared in the original study. The researchers have compared the mercury concentrations in blank samples from the new study and the original study, respectively. Blank samples are samples of (in this case) mercury-free water brought along and treated and analysed in the same way as the field samples. This is done to ensure that you do not accidentally contaminate your field samples, e.g. during transport.
“When we made comparisons, we could see that the concentrations in our blank samples were below the analytical detection limit, while the blank samples in the original study contained up to 100 nanograms of mercury per litre. Without knowing exactly what happened in the original study, there are clear indications that the original study’s mercury content may stem from a subsequent contamination of the samples, before they were analysed several years after sampling,” says Christian Juncher Jørgensen, Senior Researcher at AU.
When a research article is published, it goes through a so-called peer-review process, where other researchers assess the quality of the study before publication. Sources of error are often caught in this process.
“In this case, it looks like an error may have gone undetected in the peer review. We hope that our new study can contribute with a certainty that the latest documentation shows that the meltwater from the Greenland Ice Sheet only has a very low content of mercury”, concludes Christian Juncher Jørgensen from AU.