ACP Atmospheric Chemistry and Physics ACP 1680-7324 Copernicus GmbH Göttingen, Germany 10.5194/acp-5-2311-2005 Estimates of lightning NO<sub>x</sub> production from GOME satellite observations Boersma K. F. 1 Eskes H. J. 1 Meijer E. W. 1 Kelder H. M. 1 Royal Netherlands Meteorological Institute, De Bilt, The Netherlands 01 09 2005 5 9 2311 2331 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/5/2311/2005/acp-5-2311-2005.html The full text article is available as a PDF file from http://www.atmos-chem-phys.net/5/2311/2005/acp-5-2311-2005.pdf

Tropospheric NO<sub>2</sub> column retrievals from the Global Ozone Monitoring Experiment (GOME) satellite spectrometer are used to quantify the source strength and 3-D distribution of lightning produced nitrogen oxides (NO<sub>x</sub>=NO+NO<sub>2</sub>). A sharp increase of NO<sub>2</sub> is observed at convective cloud tops with increasing cloud top height, consistent with a power-law behaviour with power 5&plusmn;2. Convective production of clouds with the same cloud height are found to produce NO<sub>2</sub> with a ratio 1.6/1 for continents compared to oceans. This relation between cloud properties and NO<sub>2</sub> is used to construct a 10:30 local time global lightning NO<sub>2</sub> production map for 1997. An extensive statistical comparison is conducted to investigate the capability of the TM3 chemistry transport model to reproduce observed patterns of lightning NO<sub>2</sub> in time and space. This comparison uses the averaging kernel to relate modelled profiles of NO<sub>2</sub> to observed NO<sub>2</sub> columns. It exploits a masking scheme to minimise the interference of other NO<sub>x</sub> sources on the observed total columns. Simulations are performed with two lightning parameterizations, one relating convective preciptation (CP scheme) to lightning flash distributions, and the other relating the fifth power of the cloud top height (H5 scheme) to lightning distributions. The satellite-retrieved NO<sub>2</sub> fields show significant correlations with the simulated lightning contribution to the NO<sub>2</sub> concentrations for both parameterizations. Over tropical continents modelled lightning NO<sub>2</sub> shows remarkable quantitative agreement with observations. Over the oceans however, the two model lightning parameterizations overestimate the retrieved NO<sub>2</sub> attributed to lightning. Possible explanations for these overestimations are discussed. The ratio between satellite-retrieved NO<sub>2</sub> and modelled lightning NO<sub>2</sub> is used to rescale the original modelled lightning NO<sub>x</sub> production. Eight estimates of the lightning NO<sub>x</sub> production in 1997 are obtained from spatial and temporal correlation methods, from cloud-free and cloud-covered observations, and from two different lightning parameterizations. Accounting for a wide variety of random and possible systematic errors, we estimate the global NO<sub>x</sub> production from lightning to be in the range 1.1&ndash;6.4 Tg N in 1997.