Coupling to global climate
We will use GENIE and FAMOUS climate models to provide climatically determined boundary conditions for the components developed and studied in the other WPs, and as the framework for complete models of the Earth System, including all these components, to simulate and investigate their interactions through the last glacial cycle and hence the evolution of the global climate system.
Periods for time-slice integrations will include intervals: (1) during the Eemian, (2) during the early glacial drawdown of CO2, (2) during the later glacial as CO2 and sea level continue to fall, and (4) shortly before the termination to investigate the rise in CO2. These periods are chosen for their important differences in the nature of the changes going on at the time. HadAM3 (an atmosphere-only model at 2.5°x3.75° resolution) will be driven with SST anomalies and ice sheets from GENIE integrations to provide surface climate data (temperature, precipitation, winds). For each period, these integrations will be for 20 years. FAMOUS (an AOGCM at 5°x7.5° resolution in the atmosphere and 2.5°x3.75° in the ocean, derived from HadCM3), coupled to other ESM components, will be integrated for a few centuries for each period to provide 3D ocean conditions and variability on longer timescales.
The GENIE model will be used to simulate the coupled physical and biogeochemical evolution of the atmosphere-ocean-ice system. The computational efficiency of GENIE, especially in its EMBM mode, allows the thorough exploration of parameter space over entire glacial-interglacial cycles. Results from these experiments will therefore inform that selection of parameter values associated the slowly-varying, biogeochemical system components in the more expensive GCM-based work packages. A second objective of this work package will be to simulate multiple glacial-interglacial cycles with the aims of (a) determining how representative the last glacial cycle was and (b) exploring the differences between cycles and whether these differences can attributed to varying orbital forcing alone or whether endogenic changes also need to be invoked.
FAMOUS is not fast enough to run 100 kyr with all components interactively coupled. Instead we will accelerate the boundary conditions and slow components of the system by a factor of about 20, allowing integration in ~100 days on the supercomputer. Acceleration of ice sheet changes could be achieved by integrating the ice sheet for a number of years using each year of AOGCM climate . The ice sheet will not experience variability on timescales of a few years, while its longer-period climate forcing will have exaggerated variability, but this distortion appears to have little effect on the results. A better established technique is to run the climate model episodically, storing information from a number of years which are then recycled many times. Acceleration of astronomical changes in insolation should be unproblematic so long as the atmosphere-ocean climate system can keep pace with the orbital forcing. Deep ocean adjustment requires O(1000) years, which is also the period of a precession cycle if accelerated by a factor of 20. However, more important for many aspects of coupling are the surface climate and upper ocean, which adjust on much shorter timescales of decades. Acceleration techniques would need to be tested for use of other slow components in FAMOUS. The acceleration may have to adapt during the simulation; for example, if the ice sheet and climate were evolving interactively over decades through a strong effect of freshwater discharge on ocean circulation, synchronous coupling might be necessary for a period. If successful, the experiment would not only produce a simulation of Earth system evolution, but also provide insight into the various timescales of coupling between components and their dependence on internally generated variability.