Does δ 18 O of O 2 record meridional shifts in tropical rainfall?

• Published in: Climate of the Past
• Main Authors: A. M. Seltzer, C. Buizert, D. Baggenstos, E. J. Brook, J. Ahn, J.-W. Yang, J. P. Severinghaus
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2017
• Subjects: Environmental pollution; TD172-193.5; Environmental protection; TD169-171.8; Environmental sciences; GE1-350; Antarctic; Siple; Siple Dome; Antarc*; ice core
• Online Access: https://doi.org/10.5194/cp-13-1323-2017; https://doaj.org/article/0fc8364e88cf45339195d1a55cff4ba3

Marine sediments, speleothems, paleo-lake elevations, and ice core methane and δ 18 O of O 2 ( δ 18 O atm ) records provide ample evidence for repeated abrupt meridional shifts in tropical rainfall belts throughout the last glacial cycle. To improve understanding of the impact of abrupt events on the global terrestrial biosphere, we present composite records of δ 18 O atm and inferred changes in fractionation by the global terrestrial biosphere (Δ ε LAND ) from discrete gas measurements in the WAIS Divide (WD) and Siple Dome (SD) Antarctic ice cores. On the common WD timescale, it is evident that maxima in Δ ε LAND are synchronous with or shortly follow small-amplitude WD CH 4 peaks that occur within Heinrich stadials 1, 2, 4, and 5 – periods of low atmospheric CH 4 concentrations. These local CH 4 maxima have been suggested as markers of abrupt climate responses to Heinrich events. Based on our analysis of the modern seasonal cycle of gross primary productivity (GPP)-weighted δ 18 O of terrestrial precipitation (the source water for atmospheric O 2 production), we propose a simple mechanism by which Δ ε LAND tracks the centroid latitude of terrestrial oxygen production. As intense rainfall and oxygen production migrate northward, Δ ε LAND should decrease due to the underlying meridional gradient in rainfall δ 18 O. A southward shift should increase Δ ε LAND . Monsoon intensity also influences δ 18 O of precipitation, and although we cannot determine the relative contributions of the two mechanisms, both act in the same direction. Therefore, we suggest that abrupt increases in Δ ε LAND unambiguously imply a southward shift of tropical rainfall. The exact magnitude of this shift, however, remains under-constrained by Δ ε LAND .