March 2024
Hi! Another winter in the Arctic has passed, and it’s now time to review the conditions over the course of the freeze season and briefly share our new research study. This winter is especially of high interest given that another season has passed without any new daily, monthly, or climatological (maximum) record lows. This includes for all of the key sea ice metrics: ice extent, ice concentration, ice thickness, ice age, and ice volume. While the exact details of the causal drivers are much more complicated, the broad answer as to why Arctic sea ice is relatively high compared to other recent years is pretty simple: thank the “weather.”
My special visualization for this month’s blog shows changes in daily Arctic sea-ice concentration from 15 September 2023 (near the annual minimum) to 25 April 2024 (melt season underway) using the high-resolution (approximately 3 km horizontal grid) AMSR2 satellite instrument algorithm. There are some satellite-related artifacts along coastal regions that return false positive pixels of sea ice, but you can mostly ignore these. The very fast rate of the GIF animation (it may take a moment to load) is designed to help show the type of ice growth and variability that is typical of an Arctic winter.
This year’s annual maximum Arctic sea-ice extent was the overall 14th lowest on record and set on 14 March 2024 at 15.01 million square kilometers (5.80 million square miles) (NSIDC’s Sea Ice Index v3). This was very close to the average date of the sea-ice maximum, which is March 12 over the 1981-2010 climatological period. Keep in mind that all of these records begin in 1979 with the start of the temporarily- and spatially-consistent passive microwave satellite observations. By the way, we won’t know the annual maximum Arctic sea-ice volume from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) for another week or two, which is set around a month after the typical extent maximum. This is because air temperatures remain extremely cold in the far northern latitudes of the Arctic Ocean, allowing ice to continue to thicken. The area-averaged mean Arctic sea-ice thickness maximum is sometimes reached as late as mid May! Anyways, the relatively higher maximum extent in 2023-2024 was in response to favorable conditions for sea ice to expand outward along the marginal ice zone edge in both the Atlantic and Pacific sides of the Arctic Ocean. I will discuss this more throughout the rest of the blog. You can also find a map of the different regions at: https://zacklabe.com/arctic-climate-seasonality-and-variability/.
Keep in mind while reading this blog that it still gets extremely cold during winter across the Arctic Circle, even after considering recent climate change. The entire Arctic Ocean therefore still experiences ice cover given that air temperatures are well below the freezing point. This overall contributes to a much smaller long-term trend in the loss of sea-ice area and volume in winter compared to the summer melt season. In fact, recent studies (Zhang, 2021) have also found that winter ice growth may actually increase due to a negative feedback from thinner ice. This suggests that some Earth system feedbacks could also be a contributing factor to the recent slowdown in Arctic sea ice decline, rather than always contributing to an acceleration (i.e., a positive feedback). Even in the highest future emission scenarios from climate model projections, we still find sea ice reforming during the winter months across the Arctic Ocean.
Arctic sea ice typically reaches its lowest point in the year in early to mid-September, which corresponds to when all of the heat that has finally accumulated during the summer months. This is also right before the air temperatures begin to fall with the dwindling amount of solar insolation. Sea ice begins to reform shortly thereafter, especially in the northernmost latitudes where very thin ice begins to form in sea ice leads (gaps in the ice cover) thus contributing to first a greater mean sea-ice concentration across the ice pack. This past year’s freeze season kicked off with unusually widespread open water across the Pacific side of the Arctic after the September 2023 minimum, which was the 5th lowest in the satellite-era record. There was a really striking amount of open ocean water completely void of ice north of Alaska during this time period. One of my colleagues who was on an icebreaker at the time in the Beaufort Sea reported only seeing any signs of ice crystal formation at 79°N as of early October and actual sea surface temperatures of greater than +6°C were found about 100 nm north of Utqiaġvik, Alaska.
This contributed to well above average near-surface air temperatures in this region of the Arctic during late fall and early winter, as all this heat stored in the upper ocean is then lost to the overlying atmosphere through turbulent heat flux exchanges. Only then can sea ice finally begin to reform. This is one of the key reasons that the largest Arctic amplification factor is observed during late fall.
I also want to point out another key area that observed unusual sea ice conditions in early winter. This was found across the Hudson Bay, which is located in northern Canada. Atmospheric conditions of southerly winds and widespread warmth associated with an upper-level ridge of high pressure contributed to an ice freeze-up that was nearly 20 days later than average. Once the sea ice finally began to return to the Hudson Bay, the overall amount of Arctic sea-ice extent quickly rose in late December with daily rankings at only about the 20th lowest on record. This was quite remarkable considering how low sea ice has been in recent years. For a period in early January, daily Arctic sea-ice extent even rose above the 2000s decadal average for a few days (for that time of year), which was the first time this has happened in over 10 years in the JAXA-derived dataset. Keep in mind that all of this occurred in the background of global air temperatures setting new record highs, and sea surface temperatures in the North Atlantic completely shattering new daily and monthly records. Accordingly, understanding changes in the Arctic cryosphere is complicated. There are many Earth system interactions happening at the same time, and consequently it is not overly informative to make comments that record global warmth automatically means record low sea ice. Clearly this is not the case despite what is frequently shared on social media and gets all of the retweets (it’s also misinformation to use this winter as an example of sea ice recovering to normal by the way). Of course, all of this doesn’t mean that human-caused global warming isn’t real, and it doesn’t mean that the Arctic isn’t undergoing rapid change. It absolutely is and continues to warm (in the long-term) at a rate of more than three times faster than the global average. But we know that there is a lot of variability at these shorter timescales, like in the subseasonal to interannual range.
Now despite the relatively higher levels of sea-ice extent, it remained quite a warm winter compared to average over the Arctic Ocean. This counterintuitive statistic is due to the orientation of the large-scale atmospheric circulation, which favored higher sea level pressure on the Eurasian and Canadian Archipelago sides of the Arctic and lower sea level pressure in the North Atlantic and North Pacific. But remember, for most all areas the air temperature is still well below freezing, even if considering that it was nearly 20°C above average near the North Pole for a few days in early February.
This dynamically-driven circulation pattern creates a sea level pressure gradient which at times causes warmer, humid air to blow toward the North Pole, while on the other side it can enhance northerly winds. As example of this occurred in December of this winter when the position of a deep Aleutian trough caused a relatively rapid ice freeze-up in the Bering and Chukchi Seas. The westward retrograde of this trough then shifted the offshore flow to feature more prominently over the Sea of Okhotsk region. In fact, by March 2024, colder temperatures and favorable winds contributed to ice expanding to ice area that were much more like the 1980s levels for the Sea of Okhotsk. These regional northerly winds actually contributed to an equatorward expansion of ice in both the Atlantic and Pacific sides of the Arctic this winter. A similar situation occurred in the Greenland Sea with a favorable offshore/northerly flow allowing ice to expand away from the Greenland coast, including to nearly the island of Jan Mayen, which has been a rare occurrence in recent decades. In summary, there were several regions of the Arctic that simultaneously witnessed an unusual amount of sea ice compared to recent decades thanks to ideal weather conditions creating a cold offshore flow.
Despite the overall smaller loss of ice extent around the time of the annual maximum in 2024, it’s important to point out that the remaining ice cover has significantly thinned compared to just a few decades ago. Satellite-derived observations and simulated (PIOMAS) ice thickness products from this past winter support this fact and continued to reveal that sea ice was nearly 2 meters (about 6.5 feet) thinner than the 1981-2010 average north of Greenland and the Canadian Arctic Archipelago. This is also reflected in the latest ice age maps, which show little (if any) 4-5+ multi-year ice remining. And yes, this thinner and younger ice cover from the long-term trend is directly linked to human-caused climate change.
Compared to the higher extent “rankings” this winter, Arctic sea-ice volume remained on the lower side of things according to the PIOMAS dataset. February 2024 actually ended up as the 3rd lowest on record for the month due to the widespread coverage of thinner ice anomalies. The only areas of thicker than average ice were found toward the Siberian side of the Arctic by late winter. For this reason, it wouldn’t surprise me if this contributes to a later melt-out and cooler sea surface temperatures in this region during spring.
Now I have talked quite a bit about how favorable weather conditions contributed to widespread ice growth this year, but unfortunately this does not mean the news is all good for the state of Arctic sea ice. A particularly strong pressure gradient, most evident in March, likely contributed to strong ice export in the Fram Strait region of the Greenland Sea. In other words, this strong transpolar drift flow likely resulted in a substantial amount of thicker and older sea ice leaving the Arctic and eventually melting out in the subpolar north Atlantic Ocean. Previous studies (e.g., Lindsay and Zhang, 2005) have shown how this type of strong ice export can precondition the sea ice for extensive melting during the warmer months of the year. This will definitely be something to monitor this summer, especially if we see a corresponding Arctic Dipole pattern (e.g., Overland and Wang, 2016) forming with an anomalous high pressure system.
Now looking at the bigger picture, it is also obvious that we have not set a new record sea-ice minimum since September 2012. This is quite a long time, and the short-term September sea ice trend between 2012 and 2023 is nearly flat. Is this pause/hiatus/slowdown a surprise? NO. When we look across climate model large ensembles (i.e., tools for simulating the range of internal variability in the climate system relative to anthropogenic change; see Phillips et al. 2020), we see that there are periods of short-term (e.g., lasting 10-15 years) accelerations or slowdowns in the rate of the longer ice loss trend (e.g., Swart et al. 2015). The period in the early 2000s until 2012 is an example of one of those accelerated periods, and we now know that this does not necessarily mean that sea ice is declining faster than climate model projections (we have a much better understanding of internal variability using these large ensembles than we did 10-20 years ago).
It is therefore crucial that we continue to understand and communicate the influences of natural/internal climate variability for Arctic climate change. Not every month or year is going to be a new record. As I have mentioned countless times before, this is one of the key reasons that we cannot narrow down the timing for the first ice-free summer. Moving forward, we also should spend time to better understand the predictability of both the atmospheric and oceanic drivers that result in these decadal slowdowns in ice extent, especially for how long it may continue into the future. Now of course my background is in meteorology/atmospheric science, so I admit that my bias for the importance of the weather is outsized… so I will also point out that changes in ocean heat transport via the far northern Atlantic Ocean and Bering strait can also contribute to these temporary ice slowdowns and accelerations (e.g., very rapid ice loss events and bottom melt). We need to understand both the atmosphere (e.g., Francis and Wu, 2020) and ocean (e.g., Polyakov et al. 2023), and climate models remain our tools to conduct these types of causal sensitivity experiments.
A final question for this blog is whether this higher annual maximum ice extent compared to recent years is any indicator for the melt season ahead. The answer is no, probably not. There is little if any predictive value in comparing ice conditions in winter to subsequent annual minimum in September. This is also probably a good time to point out our new community-driven study that was just published last week, which reviews our understanding of predicting Arctic sea ice using statistical and dynamical models. Be sure to check it out below (it’s open access)!
Bushuk, M., S. Ali, D. Bailey, Q. Bao, L. Batte, U.S. Bhatt, E. Blanchard-Wrigglesworth, E. Blockley, G. Cawley, J. Chi, F. Counillon, P. Goulet Coulombe, R. Cullather, F.X. Diebold, A. Dirkson, E. Exarchou, M. Gobel, W. Gregory, V. Guemas, L. Hamilton, B. He, S. Horvath, M. Ionita, J. E. Kay, E. Kim, N. Kimura, D. Kondrashov, Z.M. Labe, W. Lee, Y.J. Lee, C. Li, X. Li, Y. Lin, Y. Liu, W. Maslowski, F. Massonnet, W.N. Meier, W.J. Merryfield, H. Myint, J.C. Acosta Navarro, A. Petty, F. Qiao, D. Schroder, A. Schweiger, Q. Shu, M. Sigmond, M. Steele, J. Stroeve, N. Sun, S. Tietsche, M. Tsamados, K. Wang, J. Wang, W. Wang, Y. Wang, Y. Wang, J. Williams, Q. Yang, X. Yuan, J. Zhang, and Y. Zhang (2024). Predicting September Arctic sea ice: A multi-model seasonal skill comparison. Bulletin of the American Meteorological Society, DOI:10.1175/BAMS-D-23-0163.1
In any case, this year is another reminder that monitoring local weather patterns can still bring surprises, despite rapid Arctic amplification and the dramatic transformation of the landscape at the top of our planet. If you want to understand changes in sea ice, you must follow local weather conditions and take into account natural variability. As of my writing of this blog, neither pole (Arctic or Antarctic) is currently observing record low sea ice levels despite the record warmth globally in the atmosphere and ocean. I guess we will soon see what (unpredictable) Arctic weather extremes are in store for us this summer…
Thanks for reading! March 2024 was fairly quiet again in the Arctic with relatively high levels of Arctic sea-ice extent continuing (13th lowest on record) and mostly near average air temperatures, aside from anomalous warmth over Greenland (thanks to a negative North Atlantic Oscillation circulation). However, total Arctic sea-ice volume continued to be unusually low and was the 4th lowest on record for the month of March. You can find my other monthly blogs for 2024 at https://zacklabe.com/blog-archive-2024/, and the Arctic climate data rankings for 2024 are described for the mean air temperature, sea-ice extent, and sea-ice volume at https://zacklabe.com/archive-2024/. My monthly visualization blogs are lagged by one month in the title. Okay, all done, goodbye for now!
Changes in mean surface air temperature anomalies (GISTEMPv4; 1951-1980 baseline), mean Arctic sea ice extent (NSIDC; Sea Ice Index v3), and mean Arctic sea ice volume (PIOMAS v2.1; Zhang and Rothrock, 2003) over the satellite era. Updated 4/24/2024.
Archive of Page:
Other Blog Years:
Other Climate Year Statistics:
My Visualizations:
My research related to data visualization:
[2] Witt, J.K., Z.M. Labe, A.C. Warden, and B.A. Clegg (2023). Visualizing uncertainty in hurricane forecasts with animated risk trajectories. Weather, Climate, and Society, DOI:10.1175/WCAS-D-21-0173.1
[HTML][BibTeX][Code]
[Blog][Plain Language Summary][CNN]
[1] Witt, J.K., Z.M. Labe, and B.A. Clegg (2022). Comparisons of perceptions of risk for visualizations using animated risk trajectories versus cones of uncertainty. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, DOI:10.1177/1071181322661308
[HTML][BibTeX][Code]
[Plain Language Summary][CNN]
The views presented here only reflect my own. These figures may be freely distributed (with credit). Information about the data can be found on my references page and methods page.