This project is directed at a quantitative understanding of the ocean circulation during the last glacial maximum, a period during which the climate was a radically different state from today with an ocean inferred to also have behaved quite differently. The central approach is to combine an ocean general circulation model found to be useful in the modern world, with the paleo-oceanographic proxy data for the North Atlantic during the last glacial maximum. The numerical technique for combining the two is based upon the least squared method of Lagrange multipliers as implemented using an adjoint model. Proxy data will be individually weighted according to best estimates of the relative errors. Calculation of the oceanic state involves adjusting "first-guess" initial and boundary conditions that are then modified to reduce the model-proxy-data misfits. The most abundant useful proxy data include isotopic oxygen and carbon, inferred pore water salinities and various sea surface temperature reconstructions. Estimates will be of an assumed perpetual yearly cycle. Supporting work is directed at inferences, mainly global ones, of circulation properties in the modern ocean that can most readily be compared to the much more poorly constrained last glacial maximum period.
Intellectual Merit: At one level, determination of the ocean circulation is a fundamental intellectual challenge in the earth sciences - apart from any practical applications. Determination in the modern world is very difficult, even given the comparatively large data sets available. The circulation of the last glacial maximum presents a far greater intellectual challenge, dependent as it is on indirect data sets (called "proxies") and which are sparse in space and time. In addition to this scientific curiosity aspect, the ocean is widely regarded as a major contributor in determining past and future climate states differing radically from the present one. As long as the oceanic state during the last glacial maximum remains as obscure as it is, it will be extremely difficult to claim understanding how the climate system works, and it will undermine predictions of what the future might bring as anthropogenic forcing increases.
Broader Impacts: This project is a part of a wide community effort aimed at understanding the very different climate state known to have occurred in the past, including periods of full glaciation and ones of essentially no ice. A consensus exists that the influence of the ocean in bring about these different states and/or sustaining them for extended periods of time is very important, but extremely poorly understood. This effort is unique in attempting to understand the ocean of the last glacial maximum by the fully quantitative use of all the proxy data plus a general circulation model thought to well describe the modern world. With a successfully outcome, this approach could be used to the deglacial periods as well as other intervals in the past, perhaps inter-glacial, where an adequate data exists. Most to the proposed research will be done by a graduate student as part of her Ph.D. thesis.
The work accomplised under this award has focussed on using proxies of ocean circulation in order to constrain ocean circulation prior to the instrumental record. Three of the most abundant indicators of temporal changes in ocean state are alkenones, Mg/Ca ratios, and corals, which can each be interpreted as indicators of changes in Sea Surface Temperature (SST). Along with collaborator Thomas Laepple, we have undertaken to explore the degree to which these proxies are mutually consistent with one another and with the instrumental record. Two papers on these intercomparisons have been published under the auspices of other grants. As part of this grant, we have extended these foregoing results to intercompare instrumental and proxy results, correct for systematic biases between them, and then systematically compares results against recent model simulations of SST. We find that the proxies contain an order of magnitude more variability than the model simulations at centennial timescales and that the mismatch increases to two order of magnitude at millennial timescales. This indicates that either there are continued substantial systematic biases related to interpreting SST proxies or that model simulations substantially underestimate regional SST at centennial and longer timescales. In seperate work conducting in 2011, I also demonstrated that obliquity and precession together pace the timing of deglaciation during the late Pleistocene. This demonstration was made possible throught the introduction of a new statistical procedure that is robust to errors in timing. Results are consistent with the Milankovitch hypothesis that Northern Hemisphere summer insolation controls deglaciation, but are also consistent with the duration of Antarctic summer controlling deglaciaiton.
