Long term geophysical cycles in Scandinavia (based on the analysis of tree rings and post-glacial clay deposits in Lapland, South Finland and Estonia)
Title:
Long term geophysical cycles in Scandinavia (based on the analysis of tree rings and post-glacial clay deposits in Lapland, South Finland and Estonia)
Author:
Siren, G. Hari, P.
Appeared in:
Biological rhythm research
Paging:
Volume 3 (1972) nr. 3-4 pages 425-427
Year:
1972-12
Contents:
The authors have recently published a paper on periodicity in three time series of completely different origin. Surprisingly coinciding cycles were detected. The length of the most distinct cycles were 2.24, 3.60, 4.30, 6.06, and 32 years. The background of these studies can be summarized as follows: Tree ring research reveals that fluctuations of cyclic character occur in the tree ring width. The original causes of the periodicities observed seems in most cases to be only partially known. If the climate governs the main pattern of tree growth the findings could indicate some periodicity in the climatic pattern. Reliable information about the constancy of the fluctuations seems, however, difficult to obtain, because of the short span of the tree ring series collected in sub-arctic Scandinavia, where the meteorological data series are rather short. In order to surmount at least some of the obstacles, it seemed appropriate to compare the tree ring material with other types of serial material. The post-glacial clay sediments offer such a possibility. In principle, these sediment series should reflect climatic changes in the same way as the tree rings of the sub-arctic region. The basic series analysed in the author's recent study are from three distinctly different regions, and are also of different origin. The tree ring series from the forest tree limit in Lapland has previously been described by Siren, 1961. It covers the time span from 1181 to 1960. Collection, crossdating and standardization have been carried out conventionally. A tentative serial analysis gave indication of some cycles coinciding with earlier suggested cycles of medium length. In the light of later research, the method of analysis seemed, however, unsatisfactory (Siren, 1963). The clay sediment series from Estonia (provided by the Marine Research Institute of Estonia) consisting of 748 single annual layers, was extracted from sediments accumulated in the Baltic Ice Lake about 12,000 B.C. The structure of the layers was distinct enough to enable a separate measurement of sub-layers originating from summer and winter sedimentation. In the present study the summer layer only was subjected to statistical analysis. Data from the second clay sediment have been provided by the Finnish Research Institute of Geology. The original sedimentation rate and morphology of the cores will be described by Niemela (1971). The core was taken from a clay deposit in southern Finland. Its oldest part is estimated to be from 11,000 B.C. The number of annual sediment layers totals 780. Thus the time-series studied originated not only from different geological epochs (1181-1960, about 12,000 B.C. and 11,000 B.C.) but also from completely different regions representing even different ways of formation. These facts entitle the series to be considered independent. They may, however, reflect the climatic fluctuations of the period they represent. A comparison of the independent periodicity of the series was assumed to reveal periods of either coinciding or non-coinciding length. In the case of coinciding, a preliminary hypothesis of a common generating agent appeared acceptable. The time series were analysed according to the power spectrum method. The estimates of Parzen were used (cf. Granger and Hatanaka, 1964). The significance of the cycles was tested according to a method described by the same authors. This test is originally constructed for the Tukey-Hanning estimates. According to Parzen (1961), the latter test method is, however, not the most powerful, which indicates that the limits of confidence used in this study are wider than necessary. This means that some of the peaks of the estimated power spectrum not reaching the confidence level would potentially pass the level if a more powerful test method could be exploited. The main result of the study is given in Table 1. The analysis revealed that in all three series studied, periods of different length occurred. The following information can be given about the cycles encountered: The 2.24-year occurring in both of the clay sediments coincides with the quasi-biennial oscillation caused by the changes of the movements of the moon (cf. e.g.Brier, 1968: Laulaja, 1970). The about 11-year, slightly irregular sunspot cycle is well known from many reviews on cycle research and astrophysics. The weak 13.7-year period may be caused by the influence of the moon (Brier, 1968). The 22-year cycle seems, according to Willett (1961), to be generated by sun-spot activity. The 32-year cycle probably corresponds to the 33-35-year cycles mentioned by many authors since Bruckner. No physical explanation has been presented except the possibility of a multiple of the sun-spot cycle. The undistinct 80-96 year period resembles a cycle, which Willett (1961) suggested to be a multiple of the normal sun-spot cycle; it may correspond to the 92-93-year cycle proposed by Siren (1961, 1963). Further details will be given in a later publication.
Journal description