Glacial–interglacial changes of continental weathering: estimates of the related CO 2 and HCO 3− flux variations and their uncertainties

Abstract

A range of estimates for the glacial–interglacial variations in CO 2 consumption and HCO 3 − production rates by continental weathering processes were calculated with two models of continental weathering: the Gibbs and Kump Weathering Model (GKWM) [Paleoceanography 9(4) (1994) 529] and an adapted version of Amiotte Suchet and Probst's Global Erosion Model for CO 2 Consumption (GEM-CO 2) [C. R. Acad. Sci. Paris, Ser. II 317 (1993) 615; Tellus 47B (1995) 273]. Both models link CO 2 consumption and HCO 3 − production rates to the global distributions of lithology and runoff. A spectrum of 16 estimates for the runoff distribution at the Last Glacial Maximum (LGM) was constructed on the basis of two different data sets for present-day runoff and climate results from eight GCM climate simulation experiments carried out in the framework of the Paleo Modelling Intercomparison Project (PMIP). With these forcings, GKWM produced 3.55–9.0 Tmol/year higher and GEM-CO 2 4.7–13.25 Tmol/year higher global HCO 3 − (1 Tmol=10 12 mol) production rates at the LGM. Mean variations (plus/minus one standard error of the mean with 7 df) were 6.2±0.6 and 9.4±1.0 Tmol/year, respectively. The global CO 2 consumption rates obtained with GKWM were 1.05–4.5 Tmol/year (mean: 2.8±0.4 Tmol/year) higher at the LGM than at present. With GEM-CO 2, this increase was 1.95–7.15 Tmol/year (mean: 4.8±0.6 Tmol/year). The large variability in the changes obtained with each weathering model was primarily due to the variability in the GCM results. The increase in the CO 2 consumption rate due to continental shelf exposure at the LGM was always more than 60% larger than its reduction due to ice cover. For HCO 3 − production rates, the increase related to shelf exposure was always more than twice as large as the decrease due to ice cover. Flux variations in the areas exposed both now and at the LGM were, in absolute value, always more than 3.5 times lower than those in the shelf environment. The calculated CO 2 consumption rates by carbonate weathering were consistently higher at the LGM, by 2.45–4.5 Tmol/year (mean: 3.4±0.2 Tmol/year) according to GKWM and by 2.75–6.25 Tmol/year (mean: 4.6±0.4 Tmol/year) according to GEM-CO 2. For silicate weathering, GKWM produced variations ranging between a 1.9 Tmol/year decrease and a 0.4 Tmol/year increase for the LGM (mean variation: −0.7±0.2 Tmol/year); GEM-CO 2 produced variations ranging between a 0.8 Tmol/year decrease and a 1.05 Tmol/year increase (mean variation: +0.2±0.2 Tmol/year). In the mean, the calculated variations of CO 2 and HCO 3 − fluxes would contribute to reduce atmospheric p CO 2 by 5.7±1.3 ppmv (GKWM) or 12.1±1.7 ppmv (GEM-CO 2), which might thus represent a non-negligible part of the observed glacial–interglacial variation of ∼75 ppmv.