A new method of calculating beta doses for one-dimensional geometries,
based on ''one-group'' theory, is presented. In typical situations it
is found that this approach leads to average beta doses which are sig
nificantly lower (by up to 50%) than those from Grun's DATA software.
For various natural sources, our beta dose rate estimates are slightly
greater than expected from calculations based on Monte Carlo data, wh
ile the DATA results agree well with the experimental results of Aitke
n et al. (1985) for the uranium and thorium decay chains. Further expe
rimental evidence is desirable to resolve this issue. We compare age e
stimates calculated using new ESR dating software called ROSY, which i
ncorporates this approach, with those from DATA. Because of the differ
ence in beta dose estimates, ROSY ages are generally older than the co
rresponding DATA ages, there being little difference in gamma and alph
a dose estimates. We do however observe significant discrepancies in c
osmic dose estimates. Examination of results for El Castillo Cave, Spa
in, by Rink et al. (1996) suggests that the ROSY age estimates are the
more accurate. (C) 1997 Elsevier Science Ltd.