Investigations of paleoclimate variations using accelerator mass spectrometry

摘要

This project has used Accelerator Mass Spectrometry (AMS) {sup 14}C measurements to study climate and carbon cycle variations on time scales from decades to millennia over the past 30,000 years, primarily in the western US and the North Pacific. {sup 14}C dates provide a temporal framework for records of climate change, and natural radiocarbon acts as a carbon cycle tracer in independently dated records. The overall basis for the study is the observation that attempts to model future climate and carbon cycle changes cannot be taken seriously if the models have not been adequately tested. Paleoclimate studies are unique because they provide realistic test data under climate conditions significantly different from those of the present, whereas instrumental results can only sample the system as it is today. The aim of this project has been to better establish the extent, timing, and causes of past climate perturbations, and the carbon cycle changes with which they are linked. This provides real-world data for model testing, both for the development of individual models and also for inter-model diagnosis and comparison activities such as those of LLNL's PCMDI program; it helps us achieve a better basic understanding of how the climate system works so that models can be improved; and it gives an indication of the natural variability in the climate system underlying any anthropogenically-driven changes. The research has involved four projects which test hypotheses concerning the overall behavior of the North Pacific climate system. All are aspects of an overall theme that climate linkages are strong and direct, so that regional climate records are correlated, details of fine structure are important, and accurate and precise dating is critical for establishing correlations and even causality. An important requirement for such studies is the requirement for an accurate and precise radiocarbon calibration, to allow better correlation of radiocarbon-dated records with calendric paleoclimate archives such as ice cores. The extension of the radiocarbon calibration back into the late Pleistocene (the period of deglaciation) thus constitutes a fifth project. This project has been Institute-oriented in that it was only possible through collaborations with researchers from several UC campuses, and other US and foreign institutions. These collaborators have provided expertise in sampling and access to the best available paleoclimate records. In turn, CAMS scientists have provided expertise in selecting the best samples for {sup 14}C measurements, and in interpreting radiocarbon results in terms of climate and carbon cycle changes, plus access to unmatched {sup 14}C measurement capabilities. Project output has included climate model test data plus fundamental information on the carbon cycle and the climate system, assisting LLNL modelers (and the modeling community as a whole) to improve their simulations.