Internal oscillations of the thermohaline circulation and abrupt climate changes during the last ice age and perhaps in the future.

摘要

The first half of this dissertation presents a hypothesis to explain the abrupt climate changes that occurred during the last ice age. The Dansgaard-Oeschger (DO) oscillations are a prime example of abrupt climate changes in the paleoclimate record, clear evidence that large and extremely rapid climate fluctuations have repeatedly happened in the recent past. There is wide agreement that a likely driver for the DO oscillations may involve changes in the strength of the Atlantic thermohaline circulation (THC). A 3D coupled global atmosphere-ocean-sea-ice model of intermediate complexity, ECBilt-Clio, is used here to study the natural variability of the THC and associated climate changes in the North Atlantic. Under boundary conditions appropriate for the last glacial period the model simulations produce large amplitude, DO-like oscillations. Since no varying external forcing is applied, this is an internally driven, self-sustained, nonlinear free oscillation of the THC. It is argued that the free oscillation is caused by energy imbalance between the advection of cold water and the diffusion of heat by vertical turbulent mixing in the high latitude deep ocean. In addition, another hypothesis is proposed to explain the timing of the DO events. It is argued that the pattern is created by the Northern Hemisphere summer insolation, which modulates the free oscillating mode of the THC. This speculation is supported by the ECBilt-Clio model, as well as an independent conceptual model SVO (Saltzman-van der Pol Oscillator). This hypothesis suggests that abrupt climate change is most likely to happen when insolation is rapidly changing. Since astronomical forcing is readily predictable, it is in principle possible to anticipate long-term (centennial to millennial), large amplitude, abrupt climate change events.; In the second half, the work from my secondary PhD project, which is totally independent from the aforementioned study, is presented. An inversion scheme is introduced to deduce subsurface crack model, namely crack strike, dip and crack density, by using shear-wave splitting polarization and delay time observations.