Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
SECULAR CHANGES IN THE CONCENTRATION OF ATMOSPHERIC RADIOCARBON Hans E. Suess School of Science and Engineering University of California at La Jolla The technique of radiocarbon dating has proved to be an extremely valuable tool of research. The method is based on the following prem- ises: (1) constancy of the cosmic ray flux; (2) constancy of the various carbon exchange reservoirs. It was found empirically that these assumptions were correct within certain limits of error of the measurements, but refinements in the technique of C^ determination have now shown the existence of devi- ations which demonstrate that one or both of the above assumptions are not precisely correct. The first observations were made by deVries (1958) who believed that his measurements indicated a correlation be- tween the C14 activity of wood dated from tree rings, and glacial advances and retreats. Broecker, Olson, and Bird (1959) have confirmed the devi- ations observed by deVries for the past two or three centuries. They also found that a specific C^ activity one to two percent above that of the 19th century level must have prevailed during the 16th and 17th centuries. Other observers have stated that using Libby's half-life, radiocarbon dates of material from Roman times are accurate within less than 100 years. Further observations by European workers were discussed during the last radiocarbon conference in Groningen in 1959. These observations were made on tree rings of the sequoia gigantea from the British Museum dating back about 1500 years. Deviations in radiocarbon ages have also been observed by Elizabeth Ralph (1959) on samples of historically known age from early Egyptian dynasties. The samples give too high a Cl4 con- centration by about five to ten percent, or in other words look too young by about 400 to 800 years. Through the courtesy of Dr. Terah L. Smiley of the Tree Ring Laboratory, University of Arizona, Tucson, we have obtained wood samples from sequoia gigantea dating back to 1100 B. C. The first thou- sand years in this tree produced wide rings, and ample sample material was available so that it was possible to make one determination for each century. For the younger periods (from 100 to 1700 A. D.) the wood was scarce and only one sample every 200 years could be measured. Some of 90
the runs had to be made at a reduced pressure. As this period will be covered by measurements in Europe, we have concentrated on measure- ments of material older than 200 A. D. In addition to the sequoia gigantea wood samples, two samples of two different kinds of woods (acacia and sycamore) from the tomb of King Zoser (Egypt) have been measured. According to Professor Wilson of the Oriental Institute in Chicago, who originally made this wood available to Dr. Willard Libby, King Zoser's tomb should date from about 2700 B.C., but could be as young as 2500 B.C. The wood samples of 10 to 15 grams each were first split into small fragments, then boiled with dilute NaOH and then with 3 molar HC1 and washed and dried. About one-third of the material, presumably gum and cellulose, was lost by the treatment. The counting was done in an Oeschger counter (Houtermans and Oeschger, 1958) that had a background of 1.63 cpm and a counting rate of 22. 44 cpm when filled with C2H2 of 940 mm Hg pressure prepared from wood grown around 1880. C13 determinations were made in the laboratories of Dr. S. Epstein of the California Institute of Technology and Dr. Harmon Craig of the University of California, La Jolla. The results of our measurements and the deviations from the theoretical values, calculated with a half-life of 5568 years, are shown in Figures 1 and 2. The results of the measurements by Broecker, et al. (1959) are also shown in the figures. Other laboratories have obtained results that fit ours in a similar manner. There seems to be fluctuations in the C14 concentration through time of the order of one percent, presumably outside the limits of error of the measurements, that are superimposed on a more general trend. A maximum of radiocarbon concentration existed around 1500 A. D. During the first millenium A. D. radiocarbon concentrations fluctuated around the 19th century values, as observed by other workers. During the first millenium B.C., however, the C14 concentration seemed to have dropped almost linearly. The deviations found through measurements of historically known samples by Elizabeth Ralph, as well as the results from our wood from the tomb of King Zoser seem to follow the trend indicated by our measurements on sequoia wood for the time from 1100 B.C. to 100 A. D. Obviously at this time no conclusive explanation can be given for the cause of C14 fluctuations with time. As noted by deVries (1958) the maximum in the 16th century coincides with a general advance of glaciers and cold winters. Theoretically one can expect that the average climate of the Earth will affect the Cl4 activity in two different ways which will tend to cancel each other. One is the increased downward mixing of sur- face ocean water because of a weaker thermocline in higher latitudes during periods of cold weather. The second is lower CO2 concentration in air due to a lower partial CO-, pressure at lower temperatures, which would tend to increase the specific C14 activity of the CO2 in air. Without a detailed analysis it is impossible to predict which effect would predominate. 91
KX» AGE (YEARS BP) 2000 3000 WOO 5000 â¢8 »6 i. I'^ll A 600 +400 +200 ; -2 I -200 2000 AD 1000AO 0 1000BC CALENDAR DATE 2000 BC 3000 BC Figure 1. Logarithm of C^ activity of wood samples (mostly sequoia gigantica) versus time of growths. Crosses represent LJ measure- ments. Solid circles represent values by Broecker et al. Straight line represents decay line of Cl4 assuming 5568 years half-life. The changes in the cosmic ray flux can be due to changes in (1) the galactic component, (2) the solar activity, and (3) the Earth's magnetic field. Cosmic ray induced radioactivities in meteorites seem to indicate no major change in cosmic ray activity during geologic time (Arnold, 1960). There is, however, some evidence for secular changes in the Earth's magnetic field. According to E. and O. Thellier (1959; see Elsasser et al., 1956 for other references) the Earth's magnetic field was about 65 percent stronger in Roman times than it is now. Elsasser, Ney, and Winkler (1956) have computed the change of cosmic ray intensity which might have reached the Earth's surface, using Stormer's theory. 92
0- o 1000 AD 0 1000 BC CALENDAR DATE 2000 BC 3000 BC Figure 2. Deviation of empirical results from theoretical line as given in Figure 1. Solid circles show values by Broecker et al. for comparison. 93
and have concluded that the radiocarbon concentration in Roman times should have been about three percent lower than in the 19th century, if the Earth's magnetic field only began to decline 2000 years ago, which would mean that an object ~2000 years old would be dated about 240 years too old. If the decay in the magnetic field has been an exponential one, and has persisted for the last 4000 years, then the C14 inventory two thousand years ago would have been so much lower that the samples of this age would give an age about 1000 years too old. The authors have not considered the possibility of more rapid fluctuations of the Earth's magnetic field and effects of such fluctuations on the Cl4 activity in atmospheric CO£ due to the slowness of isotopic exchange with the bi- carbonate in the oceans. Such effects might possibly explain some of the observed data. The results of the measurements indicate that for samples older than 2000 years the conventional radiocarbon age may appear perhaps as much as ten percent younger than the actual age of the material. For example, the famous Two Creeks wood, dated by Labby (1955) and by other investigators as 11, 350 years old, may actually be some 12, 500 years old; or the maximum extent of the North American ice sheet during the last glaciation measured by the author (Suess 1954) to have occurred between eighteen and nineteen thousand radiocarbon years ago may actually have been reached twenty to twenty-one thousand years ago. How- ever, it is also possible that the observed deviations represent fluctuations from a mean which on the average deviates to a much lesser degree from the true ages. I am grateful to the following people for their assistance: Dr. G. Bien atid Mrs. P. Sandoval for laboratory work; Dr. S. Epstein and Dr. H. Craig for Cl3 determinations; Dr. Terah Smiley for supplying the wood samples. This work was carried out under contract with the Division of Biology and Medicine of the U. S. Atomic Energy Commission. Broecker: deVries attempted to detect the effects of the eleven-year suns pot cycle in his samples. It is uncertain whether solar activity should enhance or reduce C14 production (see papers by Olbert and by Meyer in this volume), but in any event there was certainly too long an integration time in the over-all terrestrial Cl4 inventory, due to the two thousand year average lifetime of a C14 nucleus, for deVries to have noticed an effect. There should be a much larger effect on the Cl4 level due to bomb testing--the C14 level in this year's rings should be about 20 percent higher than in the 1953 rings. We have already compared 1953 rings with somewhat older rings and have found a one and a half percent contamination, which is not beyond possible experimental error. 94
REFERENCES Arnold, J. R. (1960) private communication. Broecker, W. S., Olson, E. A., and Bird, J. (1959) Nature 183. 1582. deVries, H. (1958) Proc. Amsterdam Acad. Sci. B6l, 1. Elsasser, W. M., Ney, E. P., and Winkler, J. R. (1956) Nature 178, 1226. Libby, W. F. (1955) Radiocarbon Dating (University of Chicago Press, Chicago). Ralph, E. K. (1959) Am. J. of Science. Radiocarbon Suppl. 1, 45. Suess, H. E. (1954) Science 120, 467. Thellier, E., and Thellier, O. (1959) Annales de Geophysique 15, 285. 95
