|
Working Paper Corrections &
Research Suggestions Most Appreciated A 9/56 year cycle has been established in patterns of US & Western European economic history since the mid 18th century (Funk, 1932; Financial Crises & The Number 56). Financial crises tend to take place every 56 years in sequences, which in turn are inter connected in sub-cycles in multiples of 9 years. Major financial crises fall with statistical significance in these patterns, a finding which contradicts prevailing economic orthodoxy. A big question is whether other phenomena also occur with significance in this cycle. The first link between the 9/56 year cycle and US earthquakes since 1800 was made by McMinn (1994). This paper looks at historical listings of major earthquakes and volcanic eruptions to see if they happen preferentially in patterns of the 9/56 year cycle. Favourable, positive correlates would support the concept of a 9/56 year seismic cycle. Alaskan Earthquakes
Please see 9/56
Year Earthquake Cycles: Alaska Based a paper by Motyka et
al (1993), the ending of major Alaskan eruptions fell with statistical
significance (p < .001) in patterns of the 36 ysc Series 1 & 2 (see Table
A, Appendix 1). This may be spurious as numerous volcanoes
ended their eruption in 1987 and the report was written in 1993. To counter
possible distortions, all historic Alaskan
eruptions were considered. Of the 220 listed by the National Geophysical
Data Center (see Table B, Appendix 1), some
58 occurred in the 12 months ended December 31 of those
years in the 36 ysc Series 1 & 2 (p < .001) (see Table 1). The notable eruption periods
within the 36 ysc Series 1 & 2 were 1929-31, 1976-80, 1987 and 1996,
which accounted for about two thirds of the volcanic eruptions that took
place within Series 1 & 2.
New York State Earthquakes New York City: 1735-2000. Won-Young Kim of Columbia University gave a listing of the significant quakes in the New York City region to 1999. Of the 18 events presented, 8 fell in the 7.5 months beginning August 1 of those years in the Table B, Appendix 8, which compared with the 2.8 expected by chance. Amazingly, of the top 7 earthquakes (mag => 4.0), five occurred in the 9/56 patterns during the four months August to November. About 0.6 could have been expected by chance. New
York State: 1735-1985. The University
of Buffalo published a listing of 25 earthquakes in New York State, of which 16 fell in 9/56 year patterns as shown in Table E, Appendix 8.
This finding was significant (p < .01). Record earthquakes in 50 US states Other US Regions California Earthquakes Servicio
Sismologico Universidad de Chile gave a comprehensive listing of Chilean earthquakes of mag =>7.0
since 1570. Of the total 92 earthquakes
presented since 1800 (see Appendix 3), 28
appeared in the 12 months beginning February 1 of those years in the 36
ysc Series 1 & 2 (see Table 1). The likelihood of this happening by
chance was p < .01.
Chinese Earthquakes The May 12, 2008 Sichuan earthquake occurred within the 9/56 year cycle. Chinese earthquakes can
also be correlated with the 9/56 year cycle. A listing, compiled from the National
Geophysical Data Center, has been given in Appendix
4. For major quakes (magnitude => 7.0) in the period 1900 to 2008,
there were some 34 earthquakes, of which 14 occurred in the 10 months beginning July 15 of those years in the combined 36 ysc Series 1
& 2 (see Table 2). This yielded acceptable significance at p < .01.
Some 21 earthquakes were presented in the List of Some Significant Earthquakes In India & Its Neighborhood, of which some 13 took place in the 11 months beginning December 1 of those years in Table B Appendix 5. The expected frequency was around 4.8. The reliability of this listing is debatable, as it only covered 'some significant earthquakes'. There is also very limited information for pre 1800 Indian quakes. The Calcutta earthquake of October 11, 1737 fell within these patterns, but it is questionable how severe this event actually was. The claimed death toll of 300,000 compared with the estimated 20,000 people living in the city at the time. Notable Earthquakes in South Asia presented another historical listing of earthquakes in South Asia, covering India, Pakistan, Nepal and Myanmar (see Table C Appendix 5). Of the total 23 events listed, 16 appeared within 12.5 months commencing August 15 of those years in the 9/56 year grid in Table D Appendix 5 (significant p < .001). Earthquakes in South East France New Zealand Earthquakes The Institute of Geological & Nuclear Sciences presented a map of major quakes in New Zealand between 1848 and 2005 (see Appendix 7). These were 'notable shallow (generally less than 30 kms) earthquakes'. Of the total 17 given, 10 showed up in the 9.5 months beginning January 15 of those years in the combined 36 ysc Series 1 & 2. In contrast, the expected frequency was only about 3.4. The Geonet listed a total of 58 major New Zealand quakes in its data base for the period 1840-1990. Of this total, some 25 appeared in the 12 months beginning January 1 of those years in the combined 36 ysc Series 1 & 2 (significant p < .01) (see Appendix 7). However, if one included the events post 1990, this significance is negated. Additionally, those quakes with a magnitude of 7.0 or higher also did not yield significance. Italian Earthquakes The United States Geological Survey presented a listing of major Italian seismic events since the mid 17th century. Considering quakes of magnitude => 5.9, gave a listing of 16 quakes (see Table A, Appendix 9), of which 9 occurred in the 6 months ended February 15 of those years in Table B Appendix 9. The expected frequency was a mere 1.6. Philippine Earthquakes Peruvian Earthquakes The National Geophysical Data Center gave 55 quakes for Peru since 1600 (see Table A, Appendix 11). Of this total, 25 were found in those 9/56 year patterns presented in Table B, Appendix 11. This was significant at p < .01. Iranian Earthquakes The USGS
listed 20 major earthquakes
for Iran during the 20th century, of which 15 occurred in the 9/56 year
patterns as shown in Table B, Appendix 12 (highly
significant p < .001). Canadian Earthquakes Australian Earthquakes Mexican Earthquakes Kostoglodov & Pacheco (1999) gave a listing of the major Mexican earthquakes (mag =>6.5) during the 20th century. Of the 182 quakes listed - 66 occurred in the 12 months ending December 31 of those 9/56 years presented in Table B, Appendix 16, which was significant (p < .01). The researchers also presented a map of major earthquakes in southern Mexico. Of the 44 quakes noted on the map - 20 occurred in the year ended December 31 of those years Table C, Appendix 16 (significant p < .01). Icelandic Earthquakes The Icelandic
Meteorological Office published a
listing of major Icelandic earthquakes between 1706 and 2000. There were
25 events with a magnitude =>6.0, none of which occurred in the four
months between September 15 and January 15. Of the total figure, 19 fell
in the 8 months beginning January 15 of those years in Table B Appendix
17, a finding that was extremely significant (p < 10-6).
Additionally, 10 major quakes were recorded in only three 56 year
sequences (Sqs 06, 15 & 24) in the 5.5 months between April 1 and September 15, while only 0.6 could have
been expected by chance. Of these 10 events, an amazing 8 happened in
the month to September 10, with five in 1896. Eastern Mediterranean Earthquakes According to the The Geophysical Institute of Israel data base, 26 major quakes (mag => 5.50) happened in the eastern Mediterranean region between 1900 and 2005. Of this total, 13 occurred in the 11.5 months ending October 15 of those years in the 9/56 year cycle presented in Table B Appendix 18, a finding which was significant at p < .01. South African Earthquakes Algerian Earthquakes Greek Earthquakes Argentine Earthquakes Major World Earthquakes Other Countries. Listings for several other countries assessed could not be linked statistically (p > .01) to 9/56 year grid patterns. This applied to Japan, Taiwan, Turkey, United Kingdom and several others. Thus there is an anomaly, which is difficult to explain. Why should the 9/56 year seismic trends show up in most countries/regions, but there are various anomalies that do not support the 9/56 year cycle hypothesis. Alas, this issue cannot be resolved based on current limited knowledge. It could be reasonably hypothesised that a 9/56 year seismic cycle shows up worldwide, but with variations according to the location on the Earth's surface. The 9/56 Year Moon - Sun Cycle Financial crises give far more regular cycles than seismic upheavals. Within the sequences, financial distress often repeats every 56 years and often about the same month within a particular sequence. Then there are the various sub-cycles in multiples of 9 years and associated artifact sub-cycles. Such regularity does not appear to be applicable to earthquakes or volcanic eruptions. This discrepancy may possibly be explained in terms of the different mechanisms involved. Financial crises are believed to arise by lunisolar cycles influencing mass human physiological cycles and thereby activating cyclic changes in the collective mood ranging between optimism and fear. Such a process would give fairly repetitive cycles, with financial crises occurring when there is a sudden shift in market sentiment from optimism to fear. However, geophysical upheavals require a build up of stresses either along active fault lines or under a volcanic vent, before they can be released suddenly in a major cataclysm. It may take decades or even centuries before stresses build up sufficiently before another earthquake/eruption can be activated in a particular region by Moon - Sun tidal resonance. Nutation Cycle. The ecliptical position of the north (ascending) node on a particular date can be correlated perfectly with the 9/56 year grids. On July 1 for ALL years in the 36 year sub-cycles Series 1 (see Table 1), the north (ascending) node was always sited between 328 and 58 Eo, while, for ALL years in Series 2, the north node appeared between 142 and 231 Eo (see Diagram 1). For such patterns to occur by chance would be infinitely small. One would expect an even ecliptical distribution of the north node, if there was no relationship with 9/56 year patterns. Thus these 9/56 year patterns can be strongly related with the 18.6 year nutation cycle and, by implication, Moon - Sun tidal effects.
DIAGRAM 1 NORTH NODE ECLIPTICAL POSITION & THE 36 YEAR SUB-CYCLES As on July 1 - All 44
years in the 36 ysc Series 1 & 2 - 1760-2000 Moon - Sun cycles may also be linked to eclipse cycles. Based on integral and half integral numbers, several Moon - Sun cycles align very closely at 9.0 and 56.0 tropical years.
56 Year Cycle. On the same date every 56, the ecliptical position of the lunar north node moves only 3 Eo clockwise (eg as on July 1: 1761 - NN at 48 Eo; 1817 - 45 Eo; 1873 - 42 Eo; 1929 - 39 Eo; 1985 - 36 Eo). This is a very close alignment of two cycles - the tropical year and lunar nutation. Furthermore, every 56.0 tropical years or 692.5 synodic months, the Sun's relative position forms the same angle to the north lunar with the Moon 180o on the opposite side of the ecliptic circle. The two cycles of 9.0 (half Saros) and 56.0 tropical years result in alternating full/new moons every 111.5 and 692.5 synodic months respectively. Every 9.0 and 56.0 years the relative angles between the Moon, Sun and lunar north node repeat to within one degree at either 0 or 180 degrees. The Saros and Half Saros cycles show
up in Moon - Sun eclipse cycles as listed by Robert
van Gent. He also gave a 112 year cycle, which divided by two
yields the 56 year eclipse cycle. Record quakes
and volcanic eruptions often occurred within plus/minus 6 months of January 1
in the 36 ysc Series 1 & 2, as follows:
The Great Sumatran Quake Was Predicted The
9.0 magnitude earthquake on December 26 caused utter devastation around
the Indian Ocean. This event was accurately predicted
by geologists from the University of Madras. The
research team, lead by N Venkatanathan and N Rajeshwara
Rao,
developed a new technique for earthquake forecasting. According
to Venkatanathan, "we predicted that the disaster will
occur on 26 December 2004 at 00:30 (GMT) with 3.54 N latitude and 97.17E
longitude, which is located near the coast of Banyak Island, Sumatra,
Indonesia, with a magnitude at around 6 to 7. The actual calamity
occurred on 26 December 2004 at 00:58 (GMT), with 3.298 N latitude and
95.779 E longitude, located off the west coast of northern Sumatra".
Remarkably, the predicted epicentre was out by a mere 157 km,
while the timing was out by only 28 minutes. The main failure was the
anticipated 7 magnitude, rather than the realised 9.0. As
Venkatanathan commented "we didn't expect the extent of damage
it will cause to the Tamilnadu coast (south east India), since we
expected the magnitude might be around 7.0, which cannot damage
Tamilnadu. We never expected the consequent tidal waves that would have
such a devastating effect on the coastal areas of Tamilnadu". The
forecast was made four days prior to the event in a paper submitted to
the Department of Science and Technology, New Delhi. The approach is based on the Moon, Sun and one or more of the planets aligning in a line at 0 and 180 degrees (ie around the new or full Moons respectively). The increased tidal pull of these configurations affects the angular momentum of the Earth, resulting in a decrease in the speed of rotation of the Earth's spin and a triggering of an earthquake. However, for an earthquake to occur, two additional factors needed to be assessed - firstly, the distance of the planetary configurations and, secondly, the direction of forces acting at the possible epicenter. Only the planets Mercury to Jupiter were assessed, as Uranus, Neptune and Pluto were considered to be too far away from Earth to be relevant. The researchers are currently appraising three key parameters to perfect their methodology - 1) the distance of the epicentre from the planetary position. 2) the direction of the force acting on the possible epicentre. 3) the increase/decrease in the rotational speed of the Earth (angular momentum). Rajeshwara Rao said "we are in the process of refining the technique so as to achieve a better success rate for which we should have a network of inputs from various international research organisations. For this to happen, there was a need for large-scale funding, which could be done through the government". If accuracy can be improved, the technique can potentially save the lives of thousands and reduce property losses substantially. It offers huge benefits to people living in coastal and earthquake prone regions around the world. Tsunamis Chile. Curso Geografia Del Mar listed major tsunamis that occurred on coastal Chile since the mid 16th century. The biggest tsunami events (categories 3 & 4) gave 14 events (see Table A, Appendix T1), of which 8 occurred in the 8.5 months ended those years in Table B, Appendix T1. In contrast, 2.1 could have been expected by chance. Strangely, Region del Biobio experienced no events within the 9/56 patterns. If the full 35 events are considered (all categories 1 to 4), significance is extinguished. This is a reasonably consistent feature - the 9/56 year effect seems most applicable to extreme events whether they are earthquakes, volcanic eruptions or tsunamis. Peru. The Tsunami Research Center (University of South Carolina) listed 13 major historic tsunamis along the Peruvian coast. Of this total, some 7 happened in the years in Table B Appendix T2, which compares with the expected frequency of 2.8. Dates were given where ever possible. July 9, 1586 "Of those listed, five were particularly destructive. These include the 1586, 1604, 1687 and 1746 tsunamis, as well as the 1868 Arica tsunami". Of these, only the 1746 event did not fall in Table B, Appendix T2. Hawaii. The map titled Hawaii Tsunamis listed some 26 historic tsunamis for the island of Hawaii in the period 1815 to 1975, of which 13 fell in the 9/56 year grid shown in Table B Appendix T3. This finding was highly significant (p < .001). Unfortunately, a source for this map could not be given despite a thorough internet search. If anyone is able to assist, please email the author. Conclusions The 9/56 year cycle in financial patterns is based on lunisolar cycles, a topic covered extensively in The Moon, The Sun & The Number 56. Thus major earthquakes could be expected to occur preferentially in the 9/56 year cycle, because of Moon - Sun tidal triggering. Both Alaskan seismic and volcanic upheavals happened preferentially in 9/56 year patterns. A major problem is to explain why Alaskan eruptions fall within the 36 ysc Series 1 & 2, but not Alaskan earthquakes. There is an anomaly, which is difficult to clarify. 9/56 year cycles could also be found in seismic patterns in China, Chile, India, USA, New Zealand and various other countries. Some seismic listings were tested, but yielded no significance between the timing of earthquakes and 9/56 year patterns (p > .01). This failure applied to Japan, UK, Greece, Turkey, Taiwan and various other countries for whatever reason. More research is warranted in this field to clarify such anomalies. The key patterns are the 9/56 year cycles as shown in Table 1 & 2. These appear wholly or partly in most of the data samples so far assessed - California, Chile, China, south east France, Hawaii, Iran, Italy, New York State, New Zealand, Peru, South Africa (Cape Province) and so forth. The exceptions were Iceland and Mexico, where a year had to be added to the years in Table 2 and Alaska & the Philippines where a year had to be subtracted. No overall world pattern could be deciphered to date, but hopefully that will become possible as more results come in. Earthquakes take place with statistical significance in 9/56 year patterns for most countries assessed. When listed chronologically, quakes would appear to be just a random collection of dates without any regular periodicity. Hence the 9/56 year seismic cycle cannot be considered a 'cycle' in the traditional sense of events happening every so many years, months, days and so forth. Rather the This cycle offers little in the way of forecasting earthquakes accurately as it is only statistically valid. Thus, a 9/56 year seismic/eruption cycle, even if proven convincingly, would be more of scientific interest rather than a good predictive tool. At best it may offer insights into windows periods when earthquakes are most likely to happen. Only time and much more research can tell. Copyright
© 2002-2007. David McMinn. All Rights Reserved. Glossary
REFERENCES |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
