By BILL KEIR
Click Here for Part 1 Click Here for Part 3
In last month’s Journal I examined the misleading astronomy and physics of alternative weather forecaster Ken Ring. In this article I examine his evidence and weather forecasting methods.
To support his claim that the lunar perigee brings disasters Ring gives a table listing 11 disasters which occurred between 1931 and 1999. Two of them are earthquakes, one is a volcanic lahar, and the rest are weather related. Eight of them occurred in New Zealand and three elsewhere.
The table employs a cunning device. To increase the hit rate the definition of a hit is made as broad as possible. Five of the disasters are said to have occurred “in the same week” as perigee. Ring doesn’t specify his protocol but, based on the samples I checked, the date of the disaster is deemed eligible for coincidence with perigee if it occurred within four or five days. This, of course, is approaching half way to apogee (seven days) when the lunar tidal force is on its way to its minimum. This amounts to disproof of Ring’s theory, so the cunning trick backfired.
The table also has several subtle errors which fudge some of the detail, and in some places it is notable for what it doesn’t tell you. For example, cyclones Bola and Dreena are listed under the same date – 10 January 1997. This was the date of Dreena. Bola happened nine years earlier, around 6 March 1988, and very close to apogee not perigee.
Cyclone Dreena had impressive coincidences – the same day as New Moon and one day before perigee. But the table doesn’t mention cyclone Fergus which happened about two weeks earlier than Dreena around apogee and last quarter. Tropical cyclones form at sea many days before they hit land in New Zealand, so their dating is uncertain to the point of being meaningless in a table like Ring’s. To be valid for his theory he would have to give the date the cyclone formed. This is not easy because they develop from weaker systems some weeks earlier and you can debate whether to date the cyclone from the formation of the original system or from the transition point where it developed the characteristics that distinguish a “hurricane” from an ordinary storm.
On my count there are only five out of the eleven disasters on Ring’s table with convincing perigee coincidences (within a day). You could expect such a result by chance given that lunar perigee happens once a month. Like the news media, Ring focuses on the human catastrophes. To prove his point he must analyse all earthquakes and storms, suitably defined, for a defined area over a defined period and derive statistical significance for the degree of fit with the theory that is convincingly above chance. Nothing in his books or website comes remotely near this.
He has more comprehensive lists on his website giving the date of every perigee in the previous year with a list of all the world disasters that happened around each. These, of course, are selected from news reports and don’t include disasters not associated with perigee. He does note that some disasters happen around apogee, but that doesn’t faze him. He simply invents a mechanism to make it fit, waffling on about potential energy being stronger than kinetic energy at apogee because the orbital speed is slower, and invoking astrological talk about the Moon “giving its energy” to the Earth.
A few properly designed studies have been done. A recent one on earthquakes (J. Vidal et al, 1998) studied 13,000 earthquakes over 25 years from 1969 to 1994 along a section of the San Andreas fault. It found that when lunar tidal forces “favour” earthquakes the rate of quakes is only, at most, two per cent higher – a statistically insignificant correlation with no predictive value.
These and other investigators have calculated the stresses present in crustal faults associated with earthquakes and found them to be up to 1000 times stronger than tidal forces (Heaton, 1975, cited by Culver and Ianna, 1988).
Are Ken Ring’s weather predictions accurate? His special boast is that, using the Moon’s position, he can do something mainstream forecasters can’t – very long-term forecasting, as far into the future as you like. His annual Almanac gives daily forecasts for 57 New Zealand towns for a year, including monthly rainfall estimates.
I have not done a rigorous analysis of his forecasts and I don’t propose to. But you don’t have to look hard to find evidence that they are not as impressive as he wants the world to think. Curiously, he has deemed it prudent to admit this in the disclaimer he attaches to his work: “The forecasts in this work are the result of best-of-ability endeavour. They represent the opinions of author and associates and no claim of 100% accuracy is made.”
Monthly Rainfall (millimetres)
Ken Ring’s Rainfall Estimates/Actual rainfall (MetService figures, NZ Herald)
2004 Jan Feb Mar Apr May June
Whangarei 195/78 210/70 no data no data 125/171 100/108
Auckland 200/47 170/274 65/9 25/38 100/219 80/111
Hamilton 215/70 110/332 50/20 20/49 135/130 75/149
Tauranga 70/29 140/237 60/15 10/226 95/196 90/111
Whakatane 75/80 165/148 165/16 25/147 90/173 100/129
Rotorua 335/47 180/298 95/10 10/73 125/179 125/183
Taupo 160/53 110/184 40/21 15/36 160/104 35/132
Gisborne 25/82 50/129 60/30 25/65 40/154 55/126
Napier 25/70 60/81 70/36 10/38 30/109 20/102
New Plymouth 120/82 205/410 160/57 30/68 135/99 80/42
Wellington 120/122 180/296 130/39 70/82 85/84 55/60
Christchurch 45/25 90/91 100/37 80/43 20/52 25/6
Dunedin 85/25 95/63 170/33 70/47 65/38 55/42
This rather dampens his bluster about the superior forecasting capabilities of his theory. He also insists that we allow a 3-4 day latitude when interpreting his predictions. This nicely covers all the possibilities, given New Zealand’s well-known average three-day high-low cycle, but negates his claim to be a reliable consultant for choosing a day to make hay or have a wedding.
He certainly can’t predict rainfall magnitudes successfully, as is obvious in my comparison of his monthly estimates with actual rainfall, for six months in 2004, for the towns for which I had MetService figures (Table 2).
Over five years from 1999 to 2004 I have monitored his forecasts unsystematically and have documented many failures which you never hear him mention. I put on record here some recent examples (Ring’s predictions taken from his New Zealand Almanac long-range daily charts):
From January to July 2004 Ken Ring failed to predict:
- A prolonged dry spell throughout New Zealand 9-18 January. He wrongly predicted heavy rain for 18 North Island towns on 14 January and a rainy three days 13-15 January.
- Significant rain throughout the North Island 20-22 January including 64 mm in Wellington. He wrongly predicted dry almost everywhere.
- Significant rain throughout New Zealand on 30 January. He predicted dry everywhere.
- The devastating floods in Manawatu and Picton 16-17 February and heavy rain in most of the North Island. He predicted dry almost everywhere (ignoring his own principles - perigee was on 16 February)
- A dry spell throughout New Zealand 5-7 March. He predicted heavy rain for 11 North Island towns on 6 March.
- Very heavy rain in the Bay of Plenty 28-29 April. He predicted dry everywhere.
- Floods in Hawkes Bay 30 June. He predicted fog there.
- Widespread rain throughout New Zealand 28-29 July. He predicted dry everywhere.
An intriguing feature of Ken Ring’s annual Almanac is the predictive isobaric maps drawn for every day for a year ahead. They centre on New Zealand and the Tasman Sea over an area identical to that covered by the MetService isobaric maps published daily in newspapers. He implies that he generates his maps mathematically “using algorithms derived from past Moon cycles.” This sounds very impressive, but he doesn’t reveal the algorithms. Exactly how lunar orbital data could yield daily atmospheric pressure data for the Tasman Sea is not clear to me.
He boasts that he can generate an isobaric map for any day in the distant future and it will closely match the isobaric map produced by MetService on the day. I’ve compared his maps with MetService maps over several months and have never found more than superficial similarities. Some are glaring mismatches. Occasionally there is a mildly convincing chance hit.
He employs an engaging trick with his maps. He publishes two maps for each day, deliberately drawn very differently (using “lunar orbital calculations” of course), and invites you to select the one that matches the reality best. Now wait a minute. Aren’t these maps supposed to be a prediction? Or is this a matching exercise after the event?
Given the well-known weather patterns around New Zealand I’m sure it would be easy to draw an arbitrary isobaric map of the area for a specified day and it would be roughly similar to the real situation as long as it had a high and a low in it somewhere. If you drew two different maps you’d cover the possibilities quite well.
Ring obligingly provides hints in his Almanac for doing your own forecasting. Some are akin to hints for fortune telling – couched in terms so general that virtually all possibilities are covered. Some don’t follow the principles of his own theory. For example, he says, “When perigee or apogee is close to new or full Moon, then a dry weather period can be expected (less than 36 hours between). When perigee or apogee is more than two days apart from the nearest new or full Moon then a wet period may be expected.” This contradicts his main argument that perigee and full and new Moon are the lunar positions strongly linked with rain.
It is hard to escape the impression that Ring achieves his claimed 80 per cent forecasting success (about the same success rate as official forecasters) by a combination of luck and educated guesses based on known weather patterns. Nothing in his writings constitutes evidence that Moon positions are a useful weather forecasting tool, or that they are related to weather at all. Much of his writing is little more than fanciful pseudoscience.
McIlveen, Robin, Fundamentals of Weather and Climate, Chapman Hall, 1992.
Zeilik, M, Gregory, S, Smith, E, Introductory Astronomy and Astrophysics, Saunders College, 1992.
Karttunen, H, et al, Fundamental Astronomy, Springer, 2003.
Kane, J, Sternheim, M, Physics, John Wiley & Sons, 1988
Meus, J, Astronomical Algorithms, William Bell Inc, 1998.