Project OC5 - Methods

The adjoint method computes a gradient of an arbitrary scalar quantity calculated from model output with respect to model input. In our case, this scalar quantity, often called the objective function, is defined as the difference between data and corresponding model results. We can use the gradient information from the adjoint method to search for the minimum of the objective function, which corresponds to an optimized model state with the smallest model-data misfit. In practice, we integrate ocean model forward in time, compare to paleo-proxy data, adjust control variables (e.g., air-sea fluxes, initial conditions, mixing parameters) to minimize misfit, and iterate until convergence.