Climate Change

Skating on Thin Ice

Reviewing Article: : Guarino, M. V., Sime, L. C., Schröeder, D., Malmierca-Vallet, I., Rosenblum, E., Ringer, M., … & Wolff, E. (2020). Sea-ice-free Arctic during the Last Interglacial supports fast future loss. Nature Climate Change, 1-5.

Twilight in the Arctic

Of all the environments being impacted by climate change, the Arctic is particularly susceptible. For the past few decades, sea ice in the Arctic has been melting at a staggering rate, so much so that it appears the disappearance of permanent Arctic sea ice is quickly upon us.

At the current rate that we are emitting CO2 and increasing the amount of heat we retain in the atmosphere, the debate in the scientific community is not if we will see an ice-free Arctic in the future, but when. Global climate models are one way scientists are developing a better understanding of sea ice loss in the Arctic. Dr. Guarino and colleagues recently published in Nature Climate Change results from a latest generation global climate model, called HadGEM3, that furthers our understanding of Arctic sea ice stability and overturn much of what we previously believed about our timeline for sea ice loss.

Late summer sea ice from 2019 compared to the late summer sea ice at the end of the 20th century in red, highlighting how much sea ice has retreated in just a few decades. (Image Source: NASA/Katie Jepson)

Global climate models like HadGEM3 simulate the oceans, atmosphere, land and ice of the entire Earth based on mathematical equations that describe the natural world (see example of a global atmospheric model here).

Schematic of global climate model. The Earth’s surface is divided into thousands of grids. In each grid, scientists model the important physical interactions (Image Source: NOAA)

As you might guess, these climate models are incredibly complex, requiring thousands of processors and hundreds of hours to run. We can test how accurate these models are by running simulations of past climate conditions and comparing their results to historical observations, a technique called “hindcasting”. Dr. Guarino and her team were confident in HadGEM3’s predictive abilities was because it performed much better than previous models in hindcasting the Last Interglacial, which spanned from 130 to 120 thousand years in past.

Simulating the Last Interglacial Period

So why did the researchers focus on validating their model during the Last Interglacial of all times? The reason is because the Last Interglacial was the most recent time in Earth’s history that the global climate was as warm as it is today. The Last Interglacial is therefore in some ways a good case study to test some of our predictions of global warming.

Based on microfossils found in marine sediments of the region, summers during the Last Interglacial were ice-free in the Arctic. This is supported by pollen and ice core records that also suggest a warmer, ice-free Arctic during this time. However, many models of the previous generation were unable to simulate ice retreat to this extent. In addition, these older models weren’t able to recreate air temperatures as cold as the pollen and ice core records suggested.

A Next Generation Climate Model

Dr. Guarino and others were eager to report that HadGEM3 was able to both simulate the loss of summer sea ice during the Last Interglacial as well as the warmer Arctic temperatures in agreement with paleo-records. Compared to its predecessors, HadGEM3 boasts more accurate cloud physics and an updated ocean model but the authors emphasize that the main reason for the discrepancy between model results is that HadGEM3 explicitly models melt ponds atop the sea ice.

Melt ponds over sea ice in the Arctic. The melt ponds are much darker than the surrounding sea ice and therefore reflect much less energy (Image Credit: Karen Frey, Clark University)

Sea ice has high albedo, which means that because it’s so white and shiny it reflects much of the sunlight it receives back into space. When the sea ice melts into pools of water, it no longer becomes as reflective. These melt ponds have much lower albedo and therefore trap more heat in the ice than if they were not there. Because HadGEM3 explicitly models these melt ponds that contribute to heating the ice, the simulated Arctic more closely resembles the Last Interglacial observations.

Polar bears are one of the many species threatened by sea ice loss in the Arctic (Image credit: Chris Linder, Woods Hole Oceanographic Institution)

According to HadCM3, a predecessor model of HadGEM3, if we make no efforts to reduce our carbon emissions into the atmosphere, then we are likely to see a disappearance of permanent sea ice by the year 2086. However, with the inclusion of the melt ponds and other more accurate climate physics, HadGEM3 predicts the first ice-free summer will occur as early as 2035. This difference highlights how important the work of climate scientists and oceanographers are. A keen understanding of our climate system is crucial for predictions of global warming.

Leave a Reply

Your email address will not be published.