Northern Hemisphere Sea Ice Area – Watts Up With That?

By Andy May

Northern Hemisphere sea ice area is an important climatic indicator because it determines how much of the Arctic Ocean and surrounding seas are open to the atmosphere. Ice is a good insulator and traps heat in the water below it (Peixoto & Oort, 1992, p. 361). Ice is also a good reflector of sunlight (high albedo), whereas water is a good absorber (low albedo). While we have no accurate data on Northern Hemisphere sea ice area (called NH_ice here) before 1978, the first year of good satellite data, it does appear to follow the global 60-70-year global climate oscillation (Wyatt, 2020). This may be because the closely related AMO affects the sea ice area as it warms and cools, of course the reverse could also be true.

Prior to 1978 we have reports on ice extent, which is weakly related to sea ice area, from ship observations. Sea ice extent was very low in the early 20^th century warm period from about 1920 to 1945, this was also within the early 20^th century rise in the AMO, which is analogous to the AMO rise that began in 1977, see here and here. Ship observations frequently mention an ice-free Arctic Ocean from the 1850s through the 1870s, another time of a warming AMO. Silas Bent proposed that the Atlantic Gulf Stream met the warm Pacific Kuroshio current in the Arctic and kept it ice free in 1872 as shown in figure 1 (Luedtke, 2015).

Systematic observations of Arctic ice did not begin until 1885 when the Danish Meteorological Institute (DMI) began to systematically make maps of ice extent using reports from “explorers, whalers, and other navigators of the polar seas.” These reports were published in both Danish and English (Luedtke, 2015). However, by then the AMO was entering a negative state.

By the mid-thirties, during the early 20^th century warm period, which was also a prominent AMO warm period, the Arctic Northeast Passage was unusually open for ordinary steamships (Luedtke, 2015). This part of the Arctic Ocean, around Svalbard, Novaya Zemlya, and Franz Josef Land (see figure 2) is usually closed by ice most of the year.

While the Northeast Passage was open most of the early 1930s, the Northwest Passage, Beaufort and Bering Seas remained ice covered over most of the year. The loss of ice in the Arctic during the 1930s prompted some to warn the world about potentially dangerous climate change (Manley, 1944), but as the world cooled over the next few decades worries about global warming diminished until temperatures began to warm again in the 1980s.

Polar energy fluxes vary by season as shown in figure 3. In summer, outgoing energy to outer space is small because the incoming energy (meridional transport + solar) goes into melting ice and snow and warming the Arctic Ocean. In winter this sequence reverses, there is very little incoming solar energy, and the summer meltwater refreezes which frees the latent heat stored in the water. Thus, most of the energy transport to outer space occurs during the winter months.

During winter the radiative energy loss at the top of the polar atmosphere is very large. About two-thirds of the loss is balanced by the inflow of transported energy from lower latitudes and one-third through or from the surface. The heat released from the surface seems to come about equally from lowering the ocean temperature and from the freezing of surface waters. Both processes are largely a function of Northern Hemisphere sea ice area or NH_ice.

Most energy delivered to the Arctic from the lower latitudes is via storms or “transient eddies” (Barry, Craig, & Thuburn, 2002), but some is from the mean meridional circulation (Kaspi & Schneider, 2013). The energy flux between the atmosphere and the surface in the Arctic is characterized by a large upward transport of energy from the surface during the winter. Water temperature below the ice is no lower than –1.9°C, while the atmosphere can be –30 to –60°C. Sea ice is a very good insulator, but when the ice is thin or absent there is some loss (Peixoto & Oort, 1992, p. 361).

NH_ice probably changes on a periodic basis because the arrow in figure 1 labeled “Meridional Transport in” and/or solar input vary. When the AMO or global temperature are warming (for example from 1975 to 2000, see here or post 1) it can also mean that the Arctic Oscillation is mostly positive and the winter polar vortex is strong. This climatic configuration keeps cold air in the Arctic, and it cools as a result. Since cold air is trapped in the Arctic, the middle latitudes and the Northern Hemisphere, except for the Arctic, are warmer.

All the factors in creating the conditions for high or low NH_ice are not known, but the so-called “Warm Arctic-cold continents” and its opposite the “Cold Arctic-warm continents” oscillation is well known (Overland, Wood, & Wang, 2011). This pattern is related to both the Arctic Oscillation (AO) as discussed above and the North Atlantic Oscillation (NAO). It is even possible that the Warm Arctic-cold continents oscillation is influenced from as far away as Antarctica through the Southern Annular Mode (SAM) (Lin, Yu, & Hall, 2025). I just mention all this to show that all the climate oscillations interact with one another and how it all fits together is very complex and poorly understood. We see the results, the climate oscillations themselves, but we do not know how they work, and they cannot be properly modeled.

Figure 4 shows the longer-term full-year HadCRUT5 and AMO rising, as NH_ice falls from 1987 to 2024. The very sudden rise in the NH_ice from 1984 to 1987 precedes a strengthening of the Polar Vortex, which begins in 1987, just as the sea ice begins a long decline. Before 1997, NH_ice changes precede AO changes, but after 1997 this relationship reverses, the reason for this change is unclear. Nothing is simple.

  The long decline in winter NH_ice from 1987 is mirrored in the Schwabe 11-year solar cycle peaks as shown in figure 5, which is a bit counterintuitive. As solar activity increases after a solar cycle low there is usually an NH_ice peak with a zero to one-year lag. Likewise, as solar activity decreases after a solar cycle high there is an NH_ice low with a one to two-year lag.

Discussion

The Arctic is a major net emitter of radiation during the winter. The combination of little to no solar input, a frozen surface, nearly cloud free skies, and very low humidity mean that almost all surface and atmospheric radiation emissions go to space. As figure 3 shows, two-thirds of the energy entering the Arctic in the winter months is brought in by atmospheric meridional transport and most of that is brought in by storms.

The Arctic Oscillation is positive when the Polar Vortex is strong and negative when it is weak. Thus, we expect NH_ice to increase when the AO is positive and decrease when it is negative, but this relationship is not consistent and often reverses as seen in figure 4. There must be other factors that affect ice cover. I have no explanation for why ice cover decreases when the Sun is less active.

We would expect NH_ice to increase as the sun gets weaker and to decrease when it is stronger, especially in the Arctic, but the opposite has occurred. The Arctic is covered with an ocean and ocean water has a low albedo and is a very strong absorber of solar radiation.

There is no obvious explanation for the relationship between all these factors, but we only have 47 years of good NH_ice data, which is not nearly enough. The overall global climate cycle is 60-70-years as noted in post 1, and maybe when we have that much data the answer will become clear. This is not a very satisfying post, since I can present all the information I could find on NH_ice, and still have no answers as to why things are as they are. All I can say is we see definite patterns, but the patterns cannot be modeled or explained. As I wrote in post 1, “Most of the natural ocean and atmospheric circulation oscillations examined in this post are not modeled properly (some say not modeled at all) in current global climate models.”

So given all of that, let’s imagine a comment like the following from a member of the “consensus:”

Andy,

You are a complete idiot. The Sun isn’t responsible for the NH_ice decline and neither are the other oscillations, it must be increasing CO₂ that is responsible! We get there by process of elimination. What else could it be?

To which I reply:

The CMIP6 and earlier sets of climate models assume that CO₂ is responsible for warming and the ice decline, yet they cannot reproduce the critical NAO, AO, AMO, SAM, and PDO oscillations. They do better with ENSO but still have major problems with it (IPCC, 2021, pp. 489-514). It is unclear that the models adequately reproduce any of the observed long-term climate oscillations. I should add they are not weather models, but climate models, and as such they should reproduce these oscillations if they are accurate. They have a CO₂ bias, yet they do not reproduce them. Process of elimination is not enough to make your case.

Barry, L., Craig, G., & Thuburn, J. (2002). Poleward heat transport by the atmospheric heat engine. Nature, 415, 774-777. doi:10.1038/415774a

IPCC. (2021). Climate Change 2021: The Physical Science Basis. In V. Masson-Delmotte, P. Zhai, A. Pirani, S. L. Connors, C. Péan, S. Berger, . . . B. Zhou (Ed.)., WG1. Retrieved from https://www.ipcc.ch/report/ar6/wg1/

Kaspi, Y., & Schneider, T. (2013). The Role of Stationary Eddies in Shaping Midlatitude Storm Tracks. Journal of the Atmospheric Sciences, 70(8), 2596 – 2613. doi:10.1175/JAS-D-12-082.1

Lin, H., Yu, B., & Hall, N. (2025). Link of the Warm Arctic Cold Eurasian pattern to the Southern Annular Mode variability. npj Climate & Atmospheric Science, 8. doi:10.1038/s41612-025-01102-z

Luedtke, B. (2015). An ice-free Arctic Ocean: history, science, and scepticism. Polar record, 51, 130-139. doi:10.1017/S0032247413000636

Manley, G. (1944). Some recent contributions to the study of climatic change. Q.J.R. Meteorol. Soc., 70, 197-219. doi:10.1002/qj.49707030508

Overland, J. E., Wood, K. R., & Wang, M. (2011). Warm Arctic—cold continents: climate impacts of the newly open Arctic Sea. Polar Research,, 30(1). doi:10.3402/polar.v30i0.15787

Peixoto, J., & Oort, A. (1992). Physics of Climate. New York: Springer-Verlag. Retrieved from https://www.amazon.com/Physics-Climate-Jose-P-Peixoto/dp/0883187124/ref=sr_1_2?crid=3BCKX8A3FIUUH&keywords=Physics+of+Climate%2C+Peixoto&qid=1649270821&sprefix=physics+of+climate%2C+peixoto%2Caps%2C100&sr=8-2

Wyatt, M. (2020). Circulation Patterns: Atmospheric and Oceanic. In Y. Wang, The Handbook of Natural Resources, volume VI (2nd ed., Vol. 6). CRC Press. Retrieved from https://www.taylorfrancis.com/chapters/edit/10.1201/9780429440984-9/circulation-patterns-atmospheric-oceanic-marcia-glaze-wyatt

Related