Exploring Non-Linear Dependencies in Atmospheric Data with Mutual Information

Relations between atmospheric variables are often non-linear, which complicates research efforts to explore and understand multivariable datasets. We describe a mutual information approach screen for the most significant associations in this setting. This method robustly detects linear non-linear dependencies after minor data quality checking. Confounding factors seasonal cycles can be taken into account without predefined models. present two case studies of method. The first one illustrates deseasonalization simple time series, with results identical classical second explores larger dataset many variables, some them lognormal (trace gas concentrations) or circular (wind direction). examples use our Python package ‘ennemi’.