Reconstructing rainfall variability in Northeast Mexico in the late-Pleistocene and Holocene using multiple speleothem geochemical proxies

Northeast Mexico is projected to become drier in the future, however, the timing, magnitude and spatial extent of precipitation is poorly constrained at present. Although paleoclimate records can help inform us about the range and mechanism of natural precipitation variability, few records are currently available. To address this, we have developed two multiproxy U-Th dated speleothem records of past hydroclimate variability from the northeast state of Tamaulipas, Mexico, with interpretations of speleothem geochemistry supported by cave monitoring data and geochemical modeling. We utilize multiple geochemical proxies (d18O, d13C and Mg/Ca) in a speleothem sample (CB2) that spans from 4.6 ka to 58.5 ka to provide one of the highest temporally resolved records of precipitation variability in Mexico over this time-period. This record demonstrates a dominant thermodynamic control on precipitation via changes in Atlantic SSTs during major paleoclimate events, including the interglacial-glacial transition, the Last Glacial Maximum, the Younger Dryas, and Heinrich Stadials 1-5, as well as on longer, orbital, timescales. Lastly, we demonstrate drying in response to cooler Atlantic SSTs is consistent across large parts of Mexico and Central America, suggesting the entire region is likely to respond similarly to precipitation change in the future. This thesis also presents the first speleothem record (CB4) from NE Mexico covering the last millennium, an essential time-period to evaluate the role of internal climate variability and for improving projections of future rainfall. While Pacific SST variability has consistently been suggested as the central influence of Northern Mexico precipitation, we instead demonstrate that mean annual rainfall in NE Mexico is more heavily influenced by Atlantic SST variability, which is further supported by forced-SST climate model simulations. We suggest previous interpretations have primarily utilized tree ring records, which tend to record winter or early summer precipitation. In contrast to future predictions of a drier climate, we demonstrate precipitation in NE Mexico has become wetter over the industrial period, and that this trend is likely to continue in the future