ACP Atmospheric Chemistry and Physics ACP 1680-7324 Copernicus GmbH Göttingen, Germany 10.5194/acp-4-143-2004 Using neural networks to describe tracer correlations Lary D. J. 1 2 3 Müller M. D. 1 4 Mussa H. Y. 3 Global Modelling and Assimilation Office, NASA Goddard Space Flight Center, USA GEST at the University of Maryland Baltimore County, MD, USA Unilever Cambridge Centre, Department of Chemistry, University of Cambridge, UK National Research Council, Washington DC, USA 31 01 2004 4 1 143 146 This is an open-access article ditributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. This article is available from http://www.atmos-chem-phys.net/4/143/2004/acp-4-143-2004.html The full text article is available as a PDF file from http://www.atmos-chem-phys.net/4/143/2004/acp-4-143-2004.pdf

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 CH<sub>4</sub>-N<sub>2</sub>O 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 CH<sub>4</sub>-N<sub>2</sub>O 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 CH<sub>4&nbsp;</sub> (but not N<sub>2</sub>O) from 1991 till the present. The neural network Fortran code used is available for download.