- JoNova - https://www.joannenova.com.au -

Can the Moon change our climate? Can tides in the atmosphere solve the mystery of ENSO?

Image by Luc Viatour  www.Lucnix.be

The Moon has such a big effect — moving 70% of the matter on the Earth’s surface every day, that it seems like the bleeding obvious to suggest that just maybe, it also affects the air, the wind, and causes atmospheric tides. Yet the climate models assume the effect is zero or close to it.

Indeed, it seems so obvious, it’s a “surely they have studied this before” moment. Though, as you’ll see, the reason lunar effects may have been ignored is not just “lunar-politics” and a lack of funding, but because it’s also seriously complex. Keep your brain engaged…

Ian Wilson and Nikolay Sidorenkov have published a provocative paper, Long-Term Lunar Atmospheric Tides in the Southern Hemisphere. It’s an epic effort of 14,000 words and a gallery of graphs. As these atmospheric tides swirl around the planet they appear to be creating standing waves of abnormal air-pressure that slowly circle the planet, once every 18 years. If this is right, then it could be the key to finally understanding, and one day predicting, the mysterious Pacific ENSO pattern that so affects the global climate. Even at this early stage, brave predictions are on the table — the atmospheric lunar tides should favor the onset of an El Nino either during the summer of 2018-19 or possibly the following southern summer. Wouldn’t it be a major step forward if we could predict the extremes?

Atmospheric tides* might be seeding the El Niños / La Nina pattern

Each summer, there are four stationary high pressure systems in the air at sea level around the Southern Hemisphere. These large regular patterns are spread evenly around the southern half of the planet, spaced at about 90 degrees longitude to each other. Each summer they reform again in roughly the same spot. (See the points marked “H” in Figure 1 below).

Fig. (1a). The NOAA SST anomaly map for the 25th of January 1981.

About 3km up above the ocean the air has its own patterns of air pressure, and the atmospheric tides appear to be standing waves of abnormal pressure (higher or lower than normal) that slowly circumnavigate the planet. There are four peaks in these standing waves, again spread evenly around the southern hemisphere. In addition, it appears that there are two epochs, it seems, the first is from 1947-1970 when all four abnormal pressure cells were high (and La Nina’s were more common). The second is from 1971 – 1994 when all four were low (and El Ninos were predominant).

There are two alternating patterns – before 1970, and after

1947-1970 — Four high pressure standing waves lead to more La Nina’s

In the first epoch, once every 4.5 years, the four main regions of drifting high-pressure cells would have passed over and strengthened the four large semipermanent high-pressure systems at sea-level.

Possibly the most influential stationary summer high occurs near Easter Island in the Pacific (see the map below). High’s rotate counter-clockwise in the Southern Hemisphere, so having two stacked high pressure systems would presumably make the East-Pacific Trade Winds stronger. These kind of conditions would favour the onset of La Nina events over El Nino events.

1971-1994 — Four low pressure standing waves lead to more El Nino’s

In the second epoch, roughly every 4½ years, the roaming low pressure cells would drift on top of the sea-level high-pressure cells. “Lows” rotate clockwise in the Southern Hemisphere, and “highs” rotate counter-clockwise, so having a low zone parked over a high zone tends to neutralize the prevailing trade winds. Thus the high near Easter Island is weakened by a roaming low pressure pattern in the air above, so the trade winds slow across the Pacific.

Normally the prevailing winds drive water from East to West, allowing cold water to well up near Peru, and warm water to sit over the Western Pacific. As the prevailing trade winds slow, the cold water stops rising, the ocean stops mixing, and the surface heats up. The effect on weather and lives extends for thousands and thousands of miles in all directions. The fishing industry off Peru is devastated by the warming seas and the nearby Peruvian hinterland experiences severe crop losses caused by torrential rains and land-slides. Eastern Australia suffers through a terrible drought at the same time as large areas of north-western Mexico and south-eastern United States experience wetter than normal winters.


When a high pressure cell in the standing wave lies over the high pressure summer semi-stationary high, the anticlockwise circulation is strengthened. This increases the prevailing trade winds (which run East to West across that part of the Pacific). This sets up La Nina type conditions.

The lower three kilometres of the Pacific Ocean is near freezing, and during an El Niño the calm ocean means that water stays cold, locked away under the surface. For a few months, the heat accumulates in the top layers, and the “stored cold” lies in wait for the El Niño to subside.

So why don’t we get an El Niño or La Nina exactly every 4½ years?

The period (technically 4.65 years) is not a whole integer, and ENSO’s form only at certain times of the year. In Spanish El Niño means “the boy child” (specifically Jesus Christ) because the effects on South America usually become obvious around Christmas. So some years the standing wave hovers over the right point, but at the wrong time of year. The overlaying cells of the standing wave and the surface highs have to coincide during the right season to seed an ENSO pattern.

The moon’s orbit not only oscillates up and down, it swings closer and further from Earth, and closer and further from the sun. The track is so complex, it will give you a headache if you try to imagine it in 3D.

Source: ScienceU

What drives these standing waves? Is it the moon?

Wilson and Sidorenkov (2013) studied the pattern of the westward shifts and found that these high pressure standing wave cells would take 18.6 years to circle the planet. In astronomical terms, 18.6 is a magic number — it’s the lunar cycle known as the Draconic Cycle. The moon does a lazy spiral around Earth as both objects go around the Sun. But it’s only after 18.6 laps (or years) around the Sun, that the Earth, Moon and Sun return to at the same position relative to each other, ready for another cycle to begin.

Lunar cycles are even more complex than this, because the orbits are not circles. Sometimes the moon is closer to the Earth (that’s perigee), and sometimes the Earth-Sun distance is at its shortest (called “perhelion”). Perhelion occurs on January 3 at the moment, but it shifts o-so-slowly on a 26,000 year cycle. Perigee makes enough difference to the appearance of the moon that we can see the change, even without a telescope. The moon looks bigger. Perhelion makes a difference to Earth – we get 7% more solar energy in January than we do at aperhelion in July.

Because both the Sun and Moon affect our tides, the largest tides are when the Sun-Moon-Earth system lines up (called “syzygy”) and all three are at the shortest distance. The full tidal cycle takes 186 years. If I tried to accurately draw the Earth-Moon orbit around the sun in 3 dimensions I’d have an ultra-stretched slinky-spring leaning 5 degrees from flat looping 186 times around the sun. Impossibly for a slinky, the 5 degree tilt of the moons orbit has a constantly changing axis. No I can’t picture that in my head either. The tilted lunar orbit means that sometimes the moon crosses the sky in front of the sun (an eclipse), but usually it tracks across the sky without doing that.


The moon spirals around the sun.

Image found on Battle Point Astronomical Association.

The atmospheric standing waves

Fig 2a shows four semi-permanent highs being enhanced by pressure “bumps” in the moving standing wave that passes through these four highs roughly once every 4 – 5 years.

Fig. (2a). A plot of the average MSLP anomaly between the latitudes of 20° S and 50° S, as a function of longitude. The longitudinal profiles are shown for those years between 1947 and 1976 that exhibit an N=4 standing wave-like zonal pattern similar to the one that appeared in the Southern Summer of 1981. Arbitrary fixed offsets have been applied to the MSLP  anomalies to vertically separate the longitudinal profiles.


Wilson and Sidorenkov took years of measurements of atmospheric pressure and stacked the results year after year. With a zero shift (the atmospheric pressure at each spot is not shifted west at all) the resulting average is blurred out mostly. With an annual 10 or 20 degree shift regular patterns appear to form suggesting that high or low pressure cells are drifting at that rate around the world. The full paper contains many shift and add graphs like this (see Fig 6a-c).

Fig. (6). (a-c) The shift-and-add maps of the Southern Hemisphere MSLP anomalies between the latitudes of 20° and 60° S for westerly longitudinal drift rates of 0°, 10°, and 20° per year, respectively. The pressure anomalies are plotted so that lower-than-normal MSLP anomalies are displayed as positive numbers.


Fourier transforms pull out the dominant cycles

This graph (Figure 5) shows how fast the lunar tidal atmospheric highs appear to be drifting around the planet — which is about 10, 20 and 40 degrees per year in longitude. Note the pattern of the first three peaks is repeated in the second three at 55, 65, and 85 degrees per year drift. This is because four high zones drifting at 10 degrees makes a very similar pattern every second year to four highs travelling at 55 degrees per year. (Spots travelling 10 degrees a year would move 20 degrees from their start position by the second year, but spots travelling at 55 degrees would have moved 110 degrees in two years and since the spots start 90 degrees apart the faster spots would have caught up with the slower ones. 110 = 90 + 20.)

The vertical bars mark the lunar cycle lengths.  The black line immediately above the 20 degree point corresponds to an 18 year cycle. The peak above 40 degrees matches a 9 year cycle.

Fig. (5). The relative power spectral density (PSD) for those features that have an N=4 pattern in the average longitudinal profile for latitudes between 30° and 50° S, for the shift-and-add map of the summer MSLP anomalies, plotted against westerly longitudinal drift rate. N.B. The
step in westerly drift rates has been increased to 2.5° per year between 0 and 25° degrees per  year for greater resolution. A solid black line drawn across the lower part of this figure indicates the minimum power spectral density that is required to rule out the possibility that the
signal is generated by noise at the 0.01 (= 99 %) confidence limit. This limit was obtained by applying the Multi-Taper Method (MTM) (number of tapers =3) to each shift-and-add anomaly profile that is associated with a given drift rate in this diagram. Only those points which had spectral densities that could not be generated by chance from either white noise or AR(1) noise at the 0.01 level were accepted as being statistically significant. All points above the 99 % confidence line in this figure are statistically significant while all the points below this line are not (with the exception of the point with a drift-rate of 50° per year). It is important to note that there are strong, low frequency spatial features present in the individual spatial MSLP anomaly profiles. Under these circumstances, simple spectrogram analysis is not good at determining the true spectra noise levels that are need to test the statistical significance of spectral features. One way to circumvent this problem is to use MTM analysis, since it better able to distinguish strong low frequency signals from spectral noise.

The peaks of the 9.3 / 93 year tidal cycle

Fig 14 combines the two dominant long term lunar cycles. The top red line is a combination of the two lower lines and represents the years when lunar tidal effects are strongest because the Earth, Moon and Sun line up and have minimal distances as well. The blue node line represents the tilt of the Moons orbit. A Node is the point of the Moons orbit where the Moon crosses the plane of the solar system.

The Nodes are the points where the Moon crosses the plane of the Solar System. The Line of Nodes joins the two Nodes on either side.

The Moon’s orbit has two nodes on opposite sides of the Earth. The Line-of-Nodes joins these two nodes. In an eclipse (i.e. when lunar tidal effects are strongest), the Line-of-Node of the lunar orbit points directly at the Sun. The blue line in figure 14 represents the angle between the Line-of-Nodes compared to the Earth-Sun direction. The blue line peaks when the Line-of-Nodes points directly at the Sun at perihelion.

When the Moon is at perigee or apogee it travels closer and further away from the Earth. The line joining perigee and apogee is called the Line-of-Apsides. The lower brown line (inverted) shows the angle of the Line-of-Apsides compared to the Earth-Sun direction. The brown line peaks when the Line-of-Apsides points directly at the Sun at perihelion.

Fig. (14). Blue curve: The angle between the line-of-nodes of the lunar orbit and the Earth-Sun line at the time of Perihelion ( θ) is plotted as a function 1/(1+ θ) between the years 1857 and 2024, in order to highlight the years in which these two axes are is close alignment. Brown curve: The angle between the line-of-apse of the lunar orbit and the Earth-Sun line at the time of perihelion (φ) plotted as the function φ1/(1+φ), in order to highlight the years in which these two axes are is close alignment. Red curve: This is an alignment index that is designed to represent the level of reinforcement of the Draconic tidal cycle by the Perigee-Syzygy tidal cycle. This is done by plotting the values of the blue curve at times when there is a close alignment of the line-of-apse and the Earth-Sun line at perihelion (i.e. when φ ≤ 16°).


Note that no one is suggesting that Lunar cycles are the dominant drivers of climate, just that they change conditions on Earth in ways that may favor certain kinds of climate patterns, and may explain the magnitude of some of the extremes. The message is that climate models will not be complete without factoring in the lunar cycles.

Below (fig 15) Wilson and Sidorenkov plot summer temperatures in Melbourne and Adelaide compared to lunar effects.

Fig. (15). The median summer time (December 1st to March 15th) maximum temperature, averaged for the cities of Melbourne and Adelaide, Australia, between 1856 and 2010 (blue curve). The alignment index curve from Fig. (13) is superimposed on this figure (red line).


Do the lunar cycles play a role in explaining why we have had a 30 year warming and then 30 year cooling cycle in the last century or so?  Support for this last contention is supplied by a recent paper by Chris de Frietas and John McLean, confirming earlier work by Bob Tisdale and Ian Wilson.

There are many more graphs in the paper as the authors sought to confirm whether the drift was westward, or eastward, and the various rates possible. It is open access, so keen readers can see for themselves the depth of work that has gone into this.



Wilson, Ian and Sidorenkov, Nikolay (2013) Long-Term Lunar Atmospheric Tides in the Southern Hemisphere, The Open Atmospheric Science Journal, 2013, 7, 29-54 [PDF]

Chris R. de Freitas, John D. McLean (2013) Update of the Chronology of Natural Signals in the Near-Surface Mean Global Temperature Record and the Southern Oscillation Index  International Journal of Geosciences, vol 4, 234-239

Other information

Ian Wilson describes the effect on his blog.

[1] http://www.stormfax.com/elnino.htm

[2] http://www.pmel.noaa.gov/pubs/outstand/mcph2029/text.shtml

[3] http://www.bom.gov.au/climate/enso/history/ln-2010-12/ENSO-when.shtml

Image: Moon  via wikimedia by Luc Viatour.



*Atmospheric tides refers to atmospheric lunar tides in this article, not Rossby waves.

8.7 out of 10 based on 74 ratings