On the sensitivity of the ocean response to LGM and MIS3-forcings

The AMOC has undergone abrupt and quasi-periodic changes during the MIS-3. The prevailing background climatic conditions that produce such behavior in AMOC have yet to be fully understood. Previous studies have demonstrated that some climate models tend to have an oscillatory behavior in their AMOC under specific conditions that vary from model to model. A systematic study that compares these conditions across models is missing. Moreover, the relative impact of greenhouse gas and icesheet forcings on the mean strength of AMOC remain unresolved.

 

Here, we present our results from CMIP/PMIP style simulations with MIS-3 boundary conditions. This study has been carried out under the PalMOD project. Based on the minimum and maximum ice sheet extent and greenhouse gas radiative forcing, we carried out a set of 4 experiments. These experiments are the LGM, 38ka, LGM_38kaghg (LGM topography with 38ka greenhouse gas concentrations), and 38ka_LGMghg (38ka topography with LGM greenhouse gas concentrations). We have used two Earth system models (ESM), Viz. the MPI-ESM and the CESM. The experiments in MPI-ESM were carried out with two versions of the river run-off directions - one in which run-off directions are compatible with the topography and the other where run-off directions are set to that of the modern-day. Thus, we have three sets of simulations for each experiment.

 

A robust feature across these simulations is that during the MIS-3, the mean strength of AMOC is sensitive to changes in greenhouse gases, and the changes in ice sheets do not significantly affect the AMOC. The density of water in the North Atlantic Deep-Water formation (NADW) region does not change significantly in response to these forcings. However, the variations in the density in the Arctic and Southern Ocean deep-water formation region drive variations in AMOC strength. The AMOC in CESM undergoes Dansgaard-Oeschger (DO) like oscillations in the 38ka LGMghg simulation. No oscillations are found in any MPI-ESM experiments with the run-off adapted for topography. However, Bo-like oscillations appear in the LGM simulation with modern run-off. This highlights the importance of model parameters and the location of freshwater input into the ocean in determining the conditions that lead to oscillatory behavior in AMOC.