The Neyman Seminar: 1011 Evans, 4:10-5:00 pm Wednesday, November 12, 2003

Using Independent Component Analysis to Separate Signals in Climate Data

Imola K. Fodor

Lawrence Liverman National Laboratory

Abstract

One problem in the atmospheric sciences is to identify the main sources of variability affecting the climate and to quantify the contributions of the different sources on the climate. For example, volcano eruptions and El Nino Southern Oscillation (ENSO) variations both influence global temperatures. Recent eruptions of the El Chicon and Mt. Pinatubo volcanoes in 1982 and 1991, respectively, coincided temporally with large ENSO events, a fact that complicates the separation of the effect of the volcanoes on temperatures from the effect of ENSO.

Current methods for separating volcano and ENSO signals involve parametric models and iterative techniques. We describe a complementary, non-parametric approach based on independent component analysis.