MASS Seminar - Douglas G. MacMartin (Caltech)
Date Time Location
October 14th, 2014 12:30pm-1:30pm 54-915
Frequency-domain analysis of ENSO and AMOC variability: Frequency-domain system identification tools are relevant to evaluate process dynamics for modes of climate variability such as El Niño/Southern Oscillation (ENSO) and the Atlantic Meridional Overturning Circulation (AMOC). For example, AMOC dynamics vary considerably between different GCMs, with some models showing clear peaks in the power spectra at multi-decadal frequencies, while others do not. To better understand these frequency-dependent differences in the response spectrum, we apply a tool to resolve subsystem process dynamics in the frequency domain. The transfer function describes the frequency-dependent input-output relationship between any pair of causally-related variables, and can be estimated from time series obtained from model simulations, or from data. This offers substantial advantage over a traditional time-domain regression, including the ability to infer the analytical form of the relationship between the input and output variables in some cases. For ENSO where it is possible to compare models with data, estimating process parameters allows the identification of compensating model errors that may lead to a seemingly realistic simulation for the wrong reason.

Time permitting; I will touch on the application of tools from engineering feedback analysis to climate science – specifically solar geoengineering, where optimization and the use of feedback to manage uncertainty can contribute towards a more systematic assessment of the potential and the limitations.