Testing models of aquatic N2O flux for inland waters

Abstract

The current Intergovernmental Panel on Climate Change (IPCC) method for tabulating agricultural nitrous oxide (N2O) emissions suggests that aquatic ecosystems may be important N2O sources. However, estimates are highly uncertain, and the method to estimate emissions is rarely tested. The default IPCC emission factor for groundwater and surface drainage (EF5-g; defined as N2O-N:nitrate-N) has been lowered recently. Our data support further reduction. Notably, we present the first EF5-gdata under ice, which reflects groundwater inputs not confounded by gas exchange. These under-ice data suggest an EF5-gof 0.13%, approximately half the current IPCC default value. Our data from the open-water season also suggest a reduction to 0.11% (based on annual means). The data highlight major problems with the IPCC method. EF5-gwas extremely variable, with the highest values observed in a stream that was a net sink of N2O. Testing the overall method for freshwater emissions is more problematic because of the multidecade transit time from headwaters in our study region to the Atlantic Ocean and limited information on emissions from lakes.