Nature and origin of variations in late-glacial and Holocene atmospheric carbon-14 as revealed by global carbon cycle modeling

Simulations with a global box-diffusion {dollar}sp{lcub}14{rcub}{dollar}C model indicate that the millennium- and century-scale atmospheric {dollar}Deltasp{lcub}14{rcub}{dollar}C variations during the Holocene are more likely explained by fluctuations in {dollar}sp{lcub}14{rcub}{dollar}C production rate (Q) than by changes in air-sea CO{dollar}sb2{dollar} exchange rate (F) or internal ocean mixing (parameterized as an "eddy diffusivity" K{dollar}sb{lcub}rm z{rcub}{dollar}. The {dollar}sp{lcub}14{rcub}{dollar}C reservoir model deconvolves histories for each of these three processes that are compatible with a 96-yr bi-decadal atmospheric (tree-ring){dollar}Deltasp{lcub}14{rcub}{dollar}C record assuming alternative pre-Holocene {dollar}sp{lcub}14{rcub}{dollar}C conditions. Holocene microparticle concentrations in ice cores and dust grain sizes in marine sediment cores disagree with the model-derived global wind speeds necessary to explain (through F variations) the millennium-scale trends in atmospheric {dollar}Deltasp{lcub}14{rcub}{dollar}C. Alternately, foram {dollar}sp{lcub}14{rcub}{dollar}C data do not support the history in the oceanic ventilation index generated by millennium-scale K{dollar}sb{lcub}rm z{rcub}{dollar} variations. Coral {dollar}sp{lcub}14{rcub}{dollar}C data for recent centuries conflict with the marine {dollar}Deltasp{lcub}14{rcub}{dollar}C history associated with century-scale variations in F or K{dollar}sb{lcub}rm z{rcub}{dollar} but are consistent with changes in {dollar}sp{lcub}14{rcub}{dollar}C production rate.; The {dollar}sp{lcub}14{rcub}{dollar}C production rates derived theoretically from an 11,000-yr record of averaged global dipole moments strongly correlate with the Q history required to explain tree-ring {dollar}Deltasp{lcub}14{rcub}{dollar}C. Several pre-Holocene Q histories were calculated from limited dipole moment data available for the past 30,000 yrs and do not contradict {dollar}sp{lcub}234{rcub}{dollar}U/{dollar}sp{lcub}230{rcub}{dollar}Th-calibrated coral {dollar}sp{lcub}14{rcub}{dollar}C measurements. Relative variations in Greenland ice-core {dollar}sp{lcub}10{rcub}{dollar}Be concentrations (reflecting changes in {dollar}sp{lcub}10{rcub}{dollar}Be production) over the past 9000 yrs also correlate strongly with tree-ring Q fluctuations except for a 4500-3500 BC discrepancy.; Simulations of transient variations in Q, F, and K{dollar}sb{lcub}rm z{rcub}{dollar} supplement previous studies of alternative steady-state {dollar}sp{lcub}14{rcub}{dollar}C situations. The modeling of combined climate and production rate scenarios (i.e. F + K{dollar}sb{lcub}rm z{rcub}{dollar}, F + Q, K{dollar}sb{lcub}rm z{rcub}{dollar} + Q) incorporates "feedback" effects which depend on the instantaneous marine {dollar}sp{lcub}14{rcub}{dollar}C profile and atmosphere/surface ocean {dollar}sp{lcub}14{rcub}{dollar}C disparity.; Spectral analyses of the 9600-yr tree-ring Q history generally produce power at or near harmonics of a 420-yr cycle. Century-scale Q periodicities may reflect harmonic or nonsinusoidal solar processes. Residual {dollar}sp{lcub}14{rcub}{dollar}C production and bi-decadal sunspot numbers from AD 1700 to 1840 define a preliminary history of relative sunspot numbers back to 7730 BC. "Fine-tuned" Q periodicities are used to forecast future solar behavior.