Volcanic eruptions of the past represent pulses of impacts to the climate system, from the ejection of aerosols into the atmosphere, and therefore provide a natural experiment to study climate response to perturbation. However, discrepancies exist between paleoclimate data and model simulations, which obstructs our ability to gain a better understanding of climate from these natural experiments.
To overcome these challenges, researchers supported by CPO’s Climate Observations and Monitoring (COM) Program investigated the causes of differences between the paleoclimate record and model simulations in a study recently published in Geophysical Research Letters. The authors examined four major sources of uncertainty: spatial coverage, seasonality, biological memory, and proxy noise. Results suggest that by accounting for these sources of uncertainty, the discrepancies are largely reconciled for moderate eruptions in the last 400 years. However, for larger eruptions, even when taking these sources of uncertainty into account, discrepancies still exist. The authors suggest these persistent differences could possibly be improved by: 1) increased spatial coverage of paleoclimate proxy data, 2) improved aerosol microphysics in models, and 3) improved estimates and representation of volcanic forcing from eruptions.
This study builds on prior NOAA-funded efforts, including the development of the Last Millenium Reanalysis, which was recently archived at NOAA’s World Data Service for Paleoclimatology this past Fall. The authors used the Last Millenium Reanalysis framework in this study to further our climate understanding and ability to use volcanic eruptions as natural experiments.