by: Dr. John C. Freeman, CCM and Jill F. Hasling, CCM
Weather Research Center
The following analysis was performed for this study. First, the strong El Nino events occurring during the earliest two orbits (1460 to 1819) were marked and used as prediction periods. Second, because the convention used by Quinn usually allotted two years for a strong El Nino, the year following the single-year El Nino event were also included in these predictions. Using this method, there were a total of twenty-two strong El Nino events in the earlier two orbits. These events were used to predict El Nino events for the most recent 178.7-year period [1819-1997].
Table 1a,Table 1b,Table 1c,and Table 1d, represents the three most recent orbital periods of the sun. These periods were chosen to begin in the years 1460, 1641 and 1819 so that the most recent period ends in 1997. As previously mentioned, data on El Nino events does not begin until 1521 so the forecast period is actually 1521-1818. The orbital period between 1819-1997 is set as the verification period. El Nino event years are denoted with an asterisk (*) and are marked with the following intensities:
WM Weak to Moderate
M, M+ Moderate
MS Moderate to Strong
S, S+ Strong
VS Very Strong
E Designated El Nino year by Climate Analysis Center (C.A.S.)
ES Designated Strong El Nino year by C.A.S.
Of the twenty-two prediction events, six were verified with strong El Nino events occurring during the same phase of the orbit for the 1819-1997 period. This is only a 27 percent forecast accuracy. However, the work of Quinn and Neal also found moderate El Ninos beginning in 1806. If these moderate El Nino events during the last orbital period are also included in the verification period, nine more "hits" are added to the forecast so that the numbers changed to fifteen out of twenty-two with an accuracy of 68 percent.
Depending on the leniency of the time period included before and after the El Nino event, the accuracy will also change. For example, if the years prior to single-year El Nino events and the years following two-year El Nino events are included in the prediction phases (denoted by ~), there are seven -teen verified El Nino predictions with an 85 percent accuracy.
This seems to indicate that there would be an El Nino event in the orbital phase in which at least one strong El Nino had occurred previously. Figure 1a,Figure 1b,Figure 1c, indicates the same information from Table 1 in graphical form. The longer lines indicate the forecast periods while the shorter lines represent actual El Nino events from 1819 to 1997. This figure better represents how the El Nino events tend to group in and around the predicted periods.
This paper is one of the results of 13 years of research by the principal authors and we would like to thank Dr. Rhodes Fairbridge for his writings on the sun's motion and encouragement and Anna Moran-Magwood for making a list of dates from which a calculation of the sun's orbit could be made. The authors would also like to acknowledge Miles E. Standish for providing the program KEPOS which was modified to compute the orbit of the sun.
Quinn, William H and Victor T. Neal : El Nino Occurrences Over the Past Four and a Half Centuries; Journal of Geophysical Research, 92,13 pp. 14,449-14,461
Jose, Paul D  Sun's Motion and Sun Spots, Astronomical Journal, Vol. 70 No.3 pp. 193-200