Using neural networks to describe tracer correlations

D. J. Lary (1, 2, 3) , M. D. Müller (2, 4) , H. Y. Mussa (3)

1 GEST at the University of Maryland Baltimore County
2 GMAO - Global Modeling and Assimilation Office
3 Unilever R&D
4 National Research Council

Neural networks are ideally suited to describe the spatial and temporal dependence of tracer-tracer correlations. The neural network performs well even in regions where the correlations are less compact and normally a family of correlation curves would be required. For example, the CH₄-N₂O correlation can be well described using a neural network trained with the latitude, pressure, time of year, and \methane\ volume mixing ratio (v.m.r.). In this study a neural network using Quickprop learning and one hidden layer with eight nodes was able to reproduce the CH₄-N₂O correlation with a correlation coefficient between simulated and training values of 0.9995. Such an accurate representation of tracer-tracer correlations allows more use to be made of long-term datasets to constrain chemical models. Such as the dataset from the Halogen Occultation Experiment (HALOE) which has continuously observed CH₄ (but not N₂O) from 1991 till the present. The neural network Fortran code used is available for download.

Domaines: Océan, Atmosphère

D. J. Lary, M. D. Müller, H. Y. Mussa. Using neural networks to describe tracer correlations. Atmospheric Chemistry and Physics, 2004, 4 (1), pp.143-146. ⟨hal-00295389⟩