The influence of human-induced climate change on extreme low Arctic sea ice extent in 2012

 

The Arctic experienced a record-low sea ice extent (SIE) in September 2012. Extreme low SIEs can have impacts on Arctic communities, ecosystems, and economic activities. Determining the role of human-induced climate change in extreme low Arctic SIEs is important, because understanding the role of anthropogenic greenhouse gases compared to natural variability provides a basis for understanding future projections and potential adaptation measures.

Event attribution methods are used to determine the influence of human-induced climate change on the occurrence (or intensity) of extreme events (NASEM, 2016). The probability of a particular extreme event is compared between two different sets of climate model simulations: those that include the contribution from human activities and those that include only natural factors. The difference in these probabilities indicates the effect of human-induced climate change on the event. Attribution studies are described in more detail in Chapter 4, Section 4.4.

Increasing temperatures in the Arctic have been attributed to human-induced factors in many studies (Gillett et al., 2008; Najafi et al., 2015; Min et al., 2008). Furthermore, attribution studies show that the record-low SIE in 2012 was extremely unlikely to be due to natural variability in the climate system alone (Zhang and Knutson, 2013) and that it would not have occurred without human influence on climate (Kirchmeier-Young et al., 2017). Figure 5.9a shows September Arctic SIE over time from climate model simulations using only natural factors (blue line) and simulations that also include the human-induced component (red line). Simulations that include the human-induced component show a strong decreasing trend, similar to the observed decline in SIE (black line). On the other hand, the simulations with only natural forcings show similar year-to-year variability but no trend.

To compare the probability of the 2012 event from each set of simulations, probability distributions are shown in Figure 5.9b. The distributions describe possible values that might be expected in each scenario and how likely they are. The observed 2012 record-low SIE event (vertical dashed line) is within the distribution from the simulations that include the human-induced component and is much lower than any values in the distribution with only natural forcings. With the human-induced component included, there is a 10.3% possibility of a September SIE more extreme than the observed 2012 event. For the natural-only simulations, this probability is extremely small. Therefore, the record-low September SIE in 2012 was extremely unlikely to be due to only natural variability of the climate and would not have occurred without the human influence on climate (Kirchmeier-Young et al., 2017).