Saturday, April 8, 2017

 Confirmation and explanation of some of the climate cycles.

Re my opus on solar cycles, Wilson published a new paper in Dec 2013 that provides a good description of the orbital mechanisms involved and provides more accurate estimates of the probable cycle lengths. See  http://www.pattern-recogn-phys.net/1/147/2013/prp-1-147-2013.pdf  . His computation would put the Deep Grand Minima cycle at 353 years compared to my estimate of 363 years, which changes the average solar cycle length by 0.3 years which is within measurement error. Similarly he estimates the long cycle I used as 1156 years vs the 1050 years that I used. Since we don't know the peak of the MWP within an accuracy of +-100 years that is also simply a refinement.

Note, in July 2015 I wrote the following, but did not post it to the blog. I wish I had known of Wilson's paper at that time. He provides an explanation for and better definition of some of the cycles discussed. One takeaway from these analyses is that there may not be a Dansgaard-Oeschger cycle- it might simply be an average of two other long cycles (McCrackens's 1301+-96 years, not confirmed by Wilson, and 1768+-174 years giving 1535 years, or the following 1339 and 1681 giving an average of 1510 years) much like the possible highly variable Bond cycle.
Confirmation and refinement of the key cycles
From this source http://scienceandpublicpolicy.org/reprint/holocene_temperature_records_show_millennial_scale_peroidicity.html  we can elucidate quite a bit more detail about periodicities. Also see  https://www.ras.org.uk/news-and-press/2680-irregular-heartbeat-of-the-sun-driven-by-double-dynamo  .

The analyses of several different proxy records of past temperatures provide a large set of apparent cycles that seem to group into a near 1100 year cycle and a near 1500 year cycle.
All values presented (extracted by the method of analysis from the data presented) for the near 1000 year cycle are 1000 (faint), 1030, 1067, 1089, 1152, 1190, 1200 and 1230 years. The authors find a 1013 year mean without saying exactly how.  The average of the values given above is 1120 years. Excluding the questionable 1000 year case, the median value is  1089 years.  Using the extremes  a value could be inferred of 1130 +- 100 years.  Perhaps 1100 years would be the best estimate to use.
For the near 1500 year cycle we find mentions of  a possible 1339 year case and more likely 1408, 1436, 1470, 1479, 1486, 1527, 1552, 1571, 1650, 1660, 1667, and 1681 year examples. The authors find a 1525 year mean. Using all of the above I find a 1533 year mean, or excluding the doubtful 1339 years, a mean of 1549 years. Using all of the values, the median is 1527 years and the range would be 1510+-170 years. Excluding the 1339 case the range would be 1540+-140 years. If we leave out the 4 values greater than 1600 which seem rather anomalous we are left with a mean of 1491 very close to the 1470 yr  Dansgaard/Oeschger  (Bond event?) cycle.
Given the margin of error on any of the estimates, and the difficulty of identifying when a real peak or bottom takes place,  the simple original value of 1500 years seems good enough.
We seem to have 2 possible cycles, 1100 years and 1500 years, which would peak in phase about every 4400 to 4500 years. Is there a 4500 year cycle? The DGM would also peak in phase with these two at intervals of 12 DGM cycles. If all of this speculation is valid, all 3 (and the Jose cycle) were last in phase at the Dark Ages or Migration minimum  about 570 AD.
The authors also find 2 groupings for the date of the MWP peak. The first is AD 798, 833, or 948, giving a mean about AD 860. The second group is AD 1036, 1047, 1088 and 1138 giving a mean of 1077. The latter value seems more consistent with most other findings. The range for this group would be 1087+-51 years. Was there an earlier peak near AD 860?? . Loehle 2007 finds the main peak about 880-900 and the second lower peak about 1000. Using a Jose cycle of very near 180 years could give a peak about  870 and 1050, and then exactly 5 cycles later at 1950. (The mid 20th century peak was 1944) That seems fairly compelling.
 My 365 - 370 yr DGM cycle gives a pre-DGM peak about 900. A 60 year cycle peak about 890 and the Jose cycle at near 870, with the 1100 year cycle about 60% of the way from valley to peak. We would then get the next Jose cycle peak about 1050, the 60 year cycle peak about 1070 and the 1100 year cycle peak near 1100, giving an extended plateau, with its maximum value suppressed by the Oort SGM.  This likely arrangement would support the 2 apparent peaks at about 880 and 1060.
Note: Between these two peaks, the valley would be the bottom of the DGM about 950, and the Jose cycle at 960. Thus the two peaks seen by the millennium study are explained, consistent with the Jose cycle and SGM/DGM cycle.
Taking the bottom of the 1100 year cycle as AD550, right at the bottom of the Migration Cold Period, the next bottom would be about 1650, and would correspond with a Jose cycle bottom at  near 1680, 60 year cycle bottoms at 1640 and 1700 and a DGM at 1680. Having all of these cycles in such close proximity explains the depth and duration of the Little Ice Age.
The  solar “deep grand minimum cycle” (DGM) theory has a DGM cycle of very near 365 years split into a section of 13 solar cycles of averaging slightly less than 11 years (140 years) from a DGM to a shallow grand minimum (SGM), followed by 20 solar cycles averaging slightly more than 11 years (225 years) to the next DGM. The SGM to DGM relationship, given the uncertainties involved, can be be considered as   the “golden ratio” and the average solar cycle length across all 33 cycles is 11.06 years. Three DGM cycles = 1095 years and 4 DGM cycles = 1460 years, very close to the 1100 and 1500 year cycles. (Note: the 1500 year cycle as proposed by Dansgaard/Oeschger was 1470 years). Also the 1500 year cycle does not show up consistently in the Holocene. Why every 3rd DGM cycle should be prominent remains unexplained.
The DGM theory also posits a sharp temperature plunge from the peak at the start of the DGM to a bottom about 30 to 40 years later, followed by a long rise (330yrs?), with the long rise punctuated by an SGM. The earliest dates estimated for the start point of or peak prior to grand minima are Wolf (DGM)-1280, Sporer (SGM)-1420,  Maunder (DGM) -1640,  and Dalton (SGM)- 1790. The DGM theory fits exactly with start dates as Wolf-1280, Sporer-1420, Maunder-1650, Dalton-1790. This correspondence is well within the margins of error of the dating. This dating would confirm the Oort (SGM) at ca 1050, very near the second peak of the MWP, making the Oort very hard to discern and of uncertain dating in the proxy records.
 Going backwards, we would have had a DGM starting in 908 and bottoming about 940, right between the peaks of the MWP (870 and 1050), and this, with the Oort,  would account for the 890 peak being higher than the 1050 peak. Going farther back we have a DGM beginning at about 550 and bottoming about 580, and coinciding closely with a bottom of the 1100 year cycle that bottomed near 1650 in the LIA. That looks just in time for the Dark Ages or Migration Minimum. Going back further we would have a DGM at 182 which corresponds to a bottom in the proxy records at near 200 AD.

These figures give a Jose cycle bottom at 2033, corresponding with a DGM starting in 2012 and reaching its coldest point about 2032-2042. The approximately 60 year cycle should also bottom at about 2035. The 1100 year cycle should peak about 2200 given all of this conjecture, so the conjunction of the other cycle bottoms will be close to the top of the 1100 year cycle, meaning that this cold period will not be as cold as the LIA.