Law Dome d18O data 2100BP to 1996 as 4-year averages

LD2.1ky-grid4y - Law Dome d18O data 2100BP (Present=2000AD) to 1996 as 4-year averages.Data set used in:    Jones, P.D. and M.E. Mann, Climate Over Past Millennia, Rev. Geo. 42(2), RG2002, doi:10.1029/2003RG000143,     Mann, M.E. and Jones, P.D. Global Surface Temperatures over the Past Two Millennia, GRL 30(15), 1820, doi: 10.1029/2003GL017814        Records for this site are in active development: check with [e-mail] for information on updated or improved Law Dome isotope data sets.    Notes:    Ages older than 700BP estimated from flow-model constrained by layer-thickness data (similar to Morgan et al., J. Glaciol, 43(143), 3-10, 1997).    Estimated error at 2ky BP on the order of a few decades.    Ages younger than 700BP from layer counting. Estimated error +/-1 year at 700BP scaling to absolute (zero error) from 193-0BP.    d18O values in per mille. Approximate accuracy of individual measurements 0.1 per mille, but each datum presented here is an integration of between 4 and 40 individual measurements, dependent upon    age and sampling scheme.    For conversion to temperature, values of ~0.2-0.3 per mille/degree C are deduced from short-term comparisons. Longer term comparison of seasonal amplitude (van Ommen and Morgan, JGR, 102(D8), 1997) indicates conversion coefficient of 0.44 per mille/degree C. Spatially derived conversion coefficients are on the order of 0.6-0.7 pe r mille/degree C, but are unlikely to be appropriate.        Note on annual phasing:    A year for the purposes of averaging of the resolved-annual cycles, is by convention marked by the isotope maximum. Thus, calendar year 1994, for example, starts with the isotope peak in the Austral summer of 1993/1994 and concludes at the 1994/1995 peak. Studies show that on the average, the isotope peak occurs around January 10 (van Ommen and Morgan, JGR, 102(D8), 9351-9357, 1997).This work was completed as part of ASAC/AAS project 757.