One of the key premises in paleoclimatology is that understanding of past climate can give us insight into present day and future climate. A corollary is that the ability of climate models to faithfully simulate past climate, particularly during extreme periods such as the Last Glacial Maximum (LGM; ~20,000 years ago), will help us improve these models and give us more confidence in their projections of future climate change. Over the past few decades scientists around the world have periodically carried out an intercomparison of the latest model simulations of climate over the past few centuries and the future. In its most recent iteration (the Intergovernmental Panel of Climate Change Fifth Assessment Report, or IPCC AR5), this exercise also includes simulations of climate during the last glacial period (the Paleoclimate Model Intercomparison Project- Phase 3, or PMIP3).
The primary goal of this project, a collaboration between researchers from Columbia University's Lamont Doherty Earth Observatory and Oregon State University, is to evaluate the ability of the AR5 and PMIP3 suite of climate models to simulate the modern and LGM global ocean circulation. The goal is to both assess the state-of-the-art models used to project future climate change, as well as to contribute to a better understanding of climate and ocean circulation during the LGM. Previous assessments have been hampered by the lack of direct observations of past climates. Fortunately, the interaction of ocean circulation with carbon dioxide and other biogeochemical trace material leaves behind chemical signatures in ocean sediments that record the climate and ocean circulation of that period. Unfortunately, while scientists have become quite adept at measuring them, these chemical signatures are only indirect and imprecise recorders of ambient climate. To get around this problem, the researchers have developed a model capable of directly simulating biogeochemical tracers and their interaction with ocean circulation. Coupled with a novel computational technique, the researchers will simulate these tracers in the AR5 and PMIP3 models, and then compare them directly to the chemical signatures measured in sediments. By analyzing the results the researchers will be able to characterize the circulation of each model, identify the underlying factors responsible, and assess its performance against observed data.
The work is highly relevant to ongoing efforts to more accurately constrain past climates, to project the future impact of human activity on the climate system, and to develop models capable of simulating different climate regimes. The work involves collaboration with climate modeling groups around the world (Germany, Japan, France, and Australia).
