By Andy May

In post 1, I ranked fourteen climate oscillations in Table 1 by their regression statistics against the HadCRUT5 global surface mean temperature. In this regression study the AMO is number one, the Western Hemisphere Warm Pool Area is #2, and the Northern Hemisphere sea ice area is #3. The fourth in importance is the Length of Day or “LOD.” Longer periods (>10 years) of acceleration in Earth’s rotation speed (shorter LOD) correspond to years of increasing zonal (east-west) circulation and global warming, whereas periods of deceleration (longer LOD) indicate less zonal acceleration and periods of cooling (Lambeck & Cazenave, 1976).

Currently, the variations in the length of day (LOD) are determined by the International Earth Rotation Service (IERS) as the current length of day minus 86,400 seconds (i.e. 24 hours). This difference is very close to zero on 1 January 1958. Annual changes in LOD are almost entirely due to seasonal changes in zonal circulation with tidal, oceanographic, and hydrological phenomena contributing less than 10% to the changes (Lambeck & Hopgood, 1981). At longer timeframes, tides and core-mantle interactions may play a large role in LOD variations. While there is an identifiable tidal influence on LOD, the core-mantle interactions and magnetic field fluctuation effects on LOD are speculative and subjects of debate (Lambeck & Cazenave, 1976).

The LOD since 1650 is shown in figure 1 in milliseconds, this value is often abbreviated as “LOD,” but is technically ΔLOD. The values plotted in figure 1 are uncorrected for changes in tides and possible mantle-core or geomagnetic influences but still correlate to major long-term climatic changes as shown in the plot. The depth of the Little Ice Age temperature minimum in the late 1600s is easily seen, as well as the late 19^th century cooling, the early 20^th century warming, the mid-20^th century cooling, and the late 20^th century warming.

The fact that the rotation speed of Earth changes with time was discovered by Simon Newcomb (Newcomb, 1882, p. 465) in 1882. He carefully measured the transits of Mercury across the face of the Sun, the Moon around the Earth, and the orbit of the largest satellite of Jupiter and observed that they were all slightly off by the same amount of time. From this he concluded that the rotation speed of the Earth must be changing.

Two very important researchers, Klyashtorin and Lyubushin of the Russian Federal Research Institute of Fisheries (Klyashtorin & Lyubushin, 2007), have studied the relationship between the Atmospheric Circulation Index (ACI) and climate changes. The ACI characterizes periods of the relative predominance of zonal (east-west) circulation versus periods of predominantly north-south (meridional) circulation. It is this fundamental climate phenomenon that LOD measures indirectly.

Figure 2 is from Klyashtorin and Lyubushin (page 12, figure 1.4). The thin line is the meridional wind anomaly and the thick line is the zonal wind anomaly. Periods when the zonal line is very positive are warm in the Northern Hemisphere and periods when the meridional line are very positive are cooler. Compare figures 1 and 2, notice the correspondence.

When the polar jet stream is “wavier” (Chalif, Osterberg, & Partridge, 2025) meridional circulation is stronger and cold Arctic air more easily travels to the middle latitudes. When the jet stream is smoother and tighter around the pole, zonal circulation is stronger, and the middle latitudes and the globe are warmer. The two jet stream configurations are illustrated by Thomas Keel in figure 3.

Two thirds of the 20^th century cool period from the 1950s through the 1970s can be attributed to elevated jet stream waviness as shown in orange in figure 3 (Chalif, Osterberg, & Partridge, 2025). Modest waviness occurred between 2000 and 2010, associated with some cooling, but the most severe waviness periods in the 20^th century were 1900-1910, the 1940s, 1960s and the late 1970s to the early 1980s (Chalif, Osterberg, & Partridge, 2025). These periods are also visible, with roughly a 12-year lag in the LOD curve as shown in figure 4.

While the relationship in figure 4 is not perfect, there is some correspondence between HadCRUT5 and LOD after accounting for a 12-year lag. Lambeck and Cazenave call it a 10-15-year lag (Lambeck & Cazenave, 1976). They used this lag to predict that the cool period in the 1970s would end 10-15 years after the downward trend began in 1972 and warming would resume, which it did. I do not think that the change in LOD directly causes the climate change 12-years later, but if the visual lag is real (it may not be real) whatever caused the change in LOD later caused the climate changes.

I should note that the regression in post 1 did not incorporate the 12-year lag and when the lag is included the statistics do not improve significantly and the AMO, WHWP, and SAM (next post) still rank above LOD in the 1950 regression (see table 2, post 1). I take this to mean that LOD doesn’t add much to what is already in the top three oscillations.

As Lambeck & Cazenave explain, some have argued that atmospheric circulation changes are not strong enough to cause the observed changes in LOD, but they have calculated that latitudinal shifts of 10° between zones of maximum and minimum winds are sufficient to explain the changes. Clearly, atmospheric changes do not account for all the changes in LOD, but they can, and probably are, a very significant component.

Discussion

Clearly periods when meridional circulation (see the orange line in figure 3) dominate and the polar jet stream is very wavy, the Northern Hemisphere is cooler because more cold Arctic air spills into the middle latitudes, especially in the winter months. The global average surface temperature tends to follow the Northern Hemisphere because temperatures in the other regions of the planet don’t change much, at least over the past 12,000 years, as shown in figure 5.

The extreme waviness of the northern polar jet stream at times is partially due to the presence of most of the world’s land in the Northern Hemisphere and the fact that much of it is close to the Arctic Ocean. As we will see in the next post on the Southern Annular Mode (SAM), also called the Antarctic Oscillation, the Southern Ocean surrounding Antarctica acts very differently. The circumpolar winds in the Southern Hemisphere are more orderly and are better at confining the cold Antarctic air.

Works Cited

Chalif, J. I., Osterberg, E. C., & Partridge, T. F. (2025). A Wavier Polar Jet Stream Contributed to the Mid-20th Century Winter Warming Hole in the United States. AGU Advances, 6(3). doi:10.1029/2024AV001399

Keel, T. (2018). Examining the link between changes in the mid-latitude jet stream in the northern hemisphere and a recent amplification of surface temperatures in the Arctic. University College London. Retrieved from https://www.researchgate.net/profile/Tom-Keel/publication/349721966_Examining_the_link_between_changes_in_the_mid-latitude_jet_stream_in_the_northern_hemisphere_and_a_recent_amplification_of_surface_temperatures_in_the_Arctic/links/603e735fa6fdcc9c780c5f5

Klyashtorin, L. B., & Lyubushin, A. A. (2007). Cyclic Climate Changes and Fish Productivity. Moscow.

Lambeck, K., & Cazenave, A. (1976). Long Term Variations in the length of Day and Climate Change. Geophysical Journal International, 46(3), 555-573. doi:10.1111/j.1365-246X.1976.tb01248.x

Lambeck, K., & Hopgood, P. (1981). The Earth’s rotation and atmospheric circulation from 1963 to 1973. Geophysical Journal International, 64(1), 67-89. doi:10.1111/j.1365-246X.1981.tb02659.x

McCarthy, D. D., & Babcock, A. K. (1986). The length of day since 1656. Physics of the Earth and Planetary Interiors, 44(3), 281-292. doi:10.1016/0031-9201(86)90077-4

Newcomb, S. (1882). Transits of Mercury. In S. Newcomb, Astronomical Papers – American Ephemeris and Nautical Almanac (Vol. 1, p. 465). Washington: U.S. Navy.

Stephenson, F. R., & Morrison, L. V. (1984). Long-term changes in the rotation of the Earth : 700 B.C. to A.D. 1980. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences, 313(1524), 47-70. doi:10.1098/rsta.1984.0082