The sea ice data you can find here is collected every day by the National Snow and Ice Data Center (NSIDC). Sea ice extent is divided into the northern and southern hemispheres, and a global measurement which is just the total of both hemispheres. It’s measured as the total area (in millions of square kilometers) that sea ice covers. Anomaly values are based on how different an ice measurement is on a given date compared with the long-term average for that same day. Positve values indicate more ice than normal, and negative values indicate less ice than normal. The farther a value is from 0, the less normal it is.
From looking at the plotly, we can see how the southern and northern hemisphere sea ice trends are about mirror images of each other. In other words, when northern sea ice is at its lowest, sea ice in the southern hemisphere tends to be close to its peak, and vice versa. However, we can see a slight lag between these opposite trends which can likely be attributed to the tilt of the Earth’s axis. Interestingly, we also sea that sea ice has greater seasonal variation in the southern hemishpere than in the northern hemisphere.
From looking at the anomaly plot, it seems like global anomalies are mostly driven by changes in the northern hemisphere. Based on simple linear models, this rings true, as anomalies in the northern hemisphere explain ~64% of variation in global anomalies, while the southern hemisphere anomalies only explains around 28% of global variation> This makes sense when we look at the annual trends in sea ice extent, where global ice extent is mostly driven by declining sea ice in the northern hemisphere.
This plot shows the average sea ice extent for each year. Based on simple linear regression models, global sea ice decreased on average by ~49,000 square kilometers each year from 1979-2021. Sea ice on the northern hemisphere decreased on average by ~54,000 square kilometers. Based on these trends, if mean annual sea ice trends continued at the same rate as they have since 1979, we would predict that sea ice on average will disappear from the northern hemisphere by the year 2210, and globally / in the southern hemisphere, by the year 2468.
This plot shows the same data as the first plot, but using stacked bars instead for easier monthly comparison of southern and northern hemisphere sea ice extent.
Surprisingly, unlike Minnesota lakes, the duration of sea ice expansion and recession in either the northern or southern hemisphere hasn’t really changed over the last 30 years. However, even though durations haven’t changed, we’ve shown from other plots that sea ice is expanding less and receding more when it does happen. Therefore, we suspect that the magnitude of increasing or decreasing is what’s shifting, but not necessarily over the number of days in which those changes occur.
These two temperature plots show the significant effect that global mean temperature change has had on the extent of global sea icem and, even more, on the presence of anomalies in global sea ice. Essentially, this potentially demonstrates much of what we’ve already heard regarding the effects of global warming / climate change, that increasing global temperatures also increase extreme events and create more variation in trends like sea ice.
This plot shows very little association between global sea ice and temperature. However, we constructed a nonlinear model and found a small but significcant effect of sea ice on temperature, with sea level decreasingly slightly with increasing sea ice. It’s important to note that because of data obtained, this plot represents sea ice and levels during the month of March.
This tab shows the variation in sea ice extent on both the northern (top panels) and southern hemisphere (bottom panels) over time. The first column shows how sea ice varies in March in each hemisphere from 1980 to 2020 to demostrate changes over the last 40 years. As we see, there are some losses in sea ice over time but it doesn’t appear to be major at this time of year for either hemisphere. However, when we shift to the second column and look at ice trends from 1980 to 2020 during the month of September, we see a different story.
In the northern hemisphere, sea ice looks to be cut in half over this forty year stretch, while the southern hemisphere still doesn’t appear to decline very much, which explains the trends we saw in annual sea ice extent plot and linear model results, where the northern hemisphere sea ice was significantly declining but there was no significant decline in the southern hemisphere.
Interestingly, when we shift to our final column, which looks at see ice differences between March 2020 and September 2020, we can see how variable sea ice is on a seasonal scale. This roughly depicts how sea ice changes at its maximum and minimum extent in each hemisphere. Of course, these exhibit inverse trends, with sea ice in the northern hemisphere reaching its greatest extent in March, and then its lowest in September, while in the southern hemisphere, sea ice roughly reaches its greatest extent in September and its lowest in March. This column shows that see ice in both hemispheres declines and increases dramatically (> 50%) every six months.