| Volumes and Issues Contents of Issue 2 |
www.biogeosciences-discuss.net/6/3373/2009/
doi:10.5194/bgd-6-3373-2009
© Author(s) 2009. This work is distributed
under the Creative Commons Attribution 3.0 License.
Estimating the monthly pCO2 distribution in the North Atlantic using a self-organizing neural network
1School of Environmental Sciences, University of East Anglia, Norwich, UK
2L'Institut Pierre-Simon Laplace/Laboratoire des Sciences du Climat et de l'Environnement, Centre National de la Recherche Scientifique - Commissariat à l'Énergie Atomique, Gif-sur-Yvette, France
3Department of Marine Chemistry, Universidad de Las Palmas de Gran Canaria, Las Palmas, Gran Canaria, Spain
4Geophysical Institute, University of Bergen, Bergen, Norway
5Leibniz Institute of Marine Sciences, Kiel, Germany
6Atlantic Oceanographic and Meteorological Laboratory, National Oceanic and Atmospheric Administration, Miami, Florida, USA
7Instituto de Investigacions Marinas (Consejo Superior de Investigaciones Científicas), Vigo, Spain
8Bjerknes Centre for Climate Research, UNIFOB AS, Bergen, Norway
9Marine Chemistry, Departement of Chemistry, University of Göterborg, Göteborg, Sweden
Abstract. Here we present monthly, basin-wide maps of the partial pressure of carbon dioxide (pCO2) for the North Atlantic on a 1° latitude by 1° longitude grid for years 2004 through 2006 inclusive, constructed using a neural network technique which reconstructs the non-linear relationships between 3 biogeochemical parameters and marine pCO2. A self organizing map (SOM) neural network has been trained using the SeaWiFS-MODIS chlorophyll a concentration, the NCEP/NCAR reanalysis sea surface temperature, and the FOAM mixed layer depth. 389 000 such triplets were used. The trained SOM was labelled with 137 000 underway pCO2 measurements collected in situ during 2004, 2005 and 2006 in the North Atlantic, which span the range of 208 and 437 μatm. The root mean square (RMS) deviation of the neural network fits from the data is 11.55 μatm, which equals to just above 3 per cent of an average pCO2 value in the in situ dataset. The seasonal pCO2 cycle as well as the interannual variability estimates in the major biogeochemical provinces is presented and spatial and temporal variability of the estimated fields is discussed. High resolution combined with basin-wide cover makes the maps a useful tool for several applications such as monitoring of basin-wide air-sea CO2 fluxes or improvement of seasonal and interannual marine CO2 cycles in future model predictions. The method itself is a valuable alternative to traditional statistical modelling techniques used in geosciences.
Discussion Paper (PDF, 3276 KB) Interactive Discussion (Closed, 4 Comments) Final Revised Paper (BG)
Citation: Telszewski, M., Chazottes, A., Schuster, U., Watson, A. J., Moulin, C., Bakker, D. C. E., González-Dávila, M., Johannessen, T., Körtzinger, A., Lüger, H., Olsen, A., Omar, A., Padin, X. A., Ríos, A., Steinhoff, T., Santana-Casiano, M., Wallace, D. W. R., and Wanninkhof, R.: Estimating the monthly pCO2 distribution in the North Atlantic using a self-organizing neural network, Biogeosciences Discuss., 6, 3373-3414, doi:10.5194/bgd-6-3373-2009, 2009. Bibtex EndNote Reference Manager XML
