Methane and nitrous oxide emissions from the ocean: A reassessment using basin-wide observations in the Atlantic

We measured the concentrations of nitrous oxide (N2O) and methane (CH4) in the marine boundary layer and surface waters of the Atlantic Ocean from similar to 50 degrees N to similar to 50 degrees S during the Atlantic Meridional Transect expedition (AMT-7) in 1998. The cruise track transects a variety of meteorological and oceanographic regimes. Unusually high mixing ratios of atmospheric CH4 were observed in the extratropical Northern Hemisphere, coinciding with globally high levels of CH4 associated with the El Nino event of 1998. Atmospheric N2O remained nearly invariable during the expedition, with only a small hemispheric difference (0.82 ppb). Throughout the cruise, these gases were saturated or supersaturated in the water. The coastal region was observed to be a significant source of CH4, while upwelling regions acted as strong N2O emission sources. We estimated the global oceanic emission of CH4 to be 0.6-1.2 Tg a(-1), comparable to previous estimates from basin-wide observations. However, our estimate turns out to be similar to 10 times lower than the value in the 1990 to 2007 Intergovernmental Panel on Climate Change (IPCC) reports, which essentially all relied on the estimate by Ehhalt (1974). A bias toward high CH4 saturation anomalies is probably responsible for the overestimation of the marine CH4 source in the IPCC reports. The CH4 saturation anomaly in the ocean appears to have remained constant over an interval of 20 years in spite of the increase of atmospheric CH4, suggesting that the increase of the surface water temperature driven by global warming may be a major factor. Meanwhile, the N2O emission from the ocean, estimated in the present study to be 0.9-1.7 Tg N a(-1), is similar to 3 times lower than the value in the recent IPCC report [Denman et al., 2007], implying either weak upwelling activity or low amounts of dissolved N2O in upwelling subsurface waters, or both, in the Atlantic.