Árbók Hins íslenzka fornleifafélags - 01.01.1991, Síða 71

ALDURSGREININGAR MEÐ GEISLAKOLI
75
Summary
Radiocarbon dating limitations and prospects
A considerable number of 14-C datings from the early settlement of Iceland have resulted
in an age 100-200 years higher than expected from historical records. These high dates are
generally rejected by Icelandic archaeologists and historians, as all other archaeological evid-
ence supports a settlement in the late 9th century. The discrepancy can be explained by
various sources of error of the 14-C method: 1) the limited accuracy of the method, 2) age of
charcoal or wood sample at time of burning, 3) the oceanic or volcanic effect postulated by
Ingrid U. Olsson according to which the 14-C concentration in the atmosphere over Iceland
could be slightly lower than the global mean, leading to too high 14-C dates.
The fundamental principles of the 14-C method is explained directly from the dendro-
chronological calibration work, rather than from Libby's original assumption of constant 14-C
concentration in the atmosphere in the past.
The present general accuracy of 14-C dating is discussed in light of the results of two recent
international intercalibration studies. These showed that a number of the participating labor-
atories had a bias of 100 years or more and that a large part of the laboratories underestimate
their quoted standard error. The results of the best laboratories show that their accuracy is
within 30 years.
The most common standard error in general radiocarbon dating is 60-80 years. Icelandic
archaeology, however, needs higher precision. An accuracy of 30 years is aimed at. This calls
for long counting times and frequent measurements of standard and background samples.
High counting capacity will be secured with a new generation of counting systems that are
being developed at the Science Institute of tlie University of Iceland. This work has resulted
in a new multi-sample liquid scintillation counting system of a radical new design. Further,
an improved new multi-counter system with gas proportional detectors will be tested late in
1992. Plans for setting up a dating Iaboratory in Iceland, using the new generation of counting
systems, are described. It is hoped that the new laboratory will start dating in 1993.