Although large, explosive volcanic eruptions are well known to be a cause of natural climate variability, several sources of uncertainty pose problems for estimating the history of volcanic forcing and its climatic response over the last millennium. Requirements for the best estimates of volcanic forcing include knowledge of eruption date, latitude, duration, and explosive power. The major source of climate perturbation is sulfate aerosol, which is formed in the stratosphere from the oxidation of volcanic SO2 gas. Thus, it is necessary to estimate volcanic aerosol mass, size distribution, and spatial extent. In this presentation, we will first briefly review how past volcanic forcing is estimated from ice cores and how it can be prescribed as aerosol mass in a global climate system model. Second, we will discuss new results from the NCAR community climate system model that compare and contrast the Arctic’s climate response to high-latitude versus low-latitude volcanic eruptions.

The late 13th Century AD is a particularly interesting case study, as several of the recent estimates of the temporal evolution of volcanic forcing suggest that it was among the most volcanically perturbed half-century of the last 2000 years. Although sulfate spikes exist in ice cores from both the Arctic and the Antarctic, none of the 13th century eruptions have been clearly attributed to specific volcanoes. We performed two sets of ensemble simulations, one assuming a sequence of tropical eruptions as the source of the volcanic sulfate spikes in the polar ice records, and one assuming multiple high-latitude eruptions.

Following large tropical eruptions, the strongest temperature response occurs in summer over Northern Hemisphere continents, where temperatures decline by 2-4 degrees C. The Arctic response is comparable. Response to high-latitude eruptions is similar for the Northern Hemisphere summer, because the local radiative balance dominates other climatic drivers in that season.

In contrast, winter climate anomalies associated with large tropical eruptions are complex. As in observations of recent volcanic eruptions, we see winter warming in some areas (northern Europe and Siberia), associated with a dynamical atmospheric response – the positive phase of the Northern Annular Mode/North Atlantic Oscillation. However, our results also show a strong effect on sea-ice extent, which increases in the North Atlantic/Labrador Sea region and is associated with regional surface temperature drops of up to 5 degrees C. High-latitude eruptions show less significant climatic response in the winter.

Closely spaced large high-latitude eruptions do not appear to be a significant cause of long-term (multiple decades to century-scale) climate change. However, with either high-latitude or tropical eruptions, there may be strong local feedbacks associated with changes in sea-ice and snow cover that could dominate the signals. We conclude by showing the model’s change in sea-ice and snow cover on land, and by discussing how these results may be reflected in paleoclimatic proxy records.