Skip navigation

Featured Science Paper

The Carrington event not observed in most ice core nitrate records

This paper has an interesting history. Space weather is high on the agenda at the moment, because of potential effects on satellite, communications and power systems. Of particular interest is to assess the frequency at which events above a particular magnitude might occur. Because the observational record is short, this requires some way of obtaining a historical frequency-amplitude plot. Almost 30 years ago, it was proposed from the solar-terrestrial (S-T) community that spikes in nitrate seen in the Greenland and Antarctic ice core record might be recording the occurrence of solar energetic particle (SEP) events associated with solar flares. Almost immediately the ice core and atmospheric chemistry community reacted by stating that this was a false interpretation, and that SEPs could not leave such a strong, sharp and immediate signal in ice. However, the idea lived on in the S-T community, and was amplified a decade ago by a paper (McCracken et al) that used the nitrate spikes in a Greenland core to assess the statistics of such events. When Eric Wolff was invited to a workshop last year about the statistics of such solar events (using solar, stellar, and palaeo-observations), it became obvious that the ice core community needed to react with definite evidence about the origin of the nitrate spikes.

A group of scientists, representing many of the leading ice core groups, who between them held a lot of high-resolution ice core records suitable for diagnosing the nitrate spikes, was assembled. We concentrated on the period around the most famous of all such events, the Carrington Event of 1859 (the first recorded observation of a white light flare). We showed that, apart from the original record used by McCracken, none of our cores had a spike dated to 1859. Greenland cores do indeed have numerous nitrate spikes such as those observed earlier, but they have the characteristic chemical signature of biomass burning events. Their statistics therefore represent the stats of biomass burning coinciding with transport to Greenland. Given that it is not possible to see a signal for an event as prominent as the Carrington Event, it seems that nitrate spikes (even after filtering) cannot be used to look at the statistics of solar events. Although this is a negative result, it does an important service in resolving a longstanding clash between fields, and moving the effort onto more defendible measures such as 10Be and 14C records.

Link to the full paper in the NERC Open Research Archive


Wolff, E.W., Bigler, M., Curran, M.A.J., Dibb, J.E., Frey, M.M., Legrand, M., McConnell, J.R., 2012


Geophysical Research Letters, 30, L08503, 5pp