Fischer, Tim; Kock, Annette; Arévalo-Martínez, Damian L; Dengler, Marcus; Brandt, Peter; Bange, Hermann Werner (2019): Gas exchange estimates in the Peruvian upwelling regime biased by multi-day near-surface stratification [dataset publication series]. PANGAEA, https://doi.org/10.1594/PANGAEA.902689,

---

Supplement to: Fischer, T et al. (2019): Gas exchange estimates in the Peruvian upwelling regime biased by multi-day near-surface stratification. Biogeosciences, 16(11), 2307-2328, https://doi.org/10.5194/bg-16-2307-2019

The coastal upwelling regime off Peru in December 2012 showed considerable vertical concentration gradients of dissolved nitrous oxide (N2O) across the top few meters of the ocean. The gradients were predominantly downward, i.e., concentrations decreased toward the surface. Ignoring these gradients causes a systematic error in regionally integrated gas exchange estimates, when using observed concentrations at several meters below the surface as input for bulk flux parameterizations – as is routinely practiced. Here we propose that multi-day near-surface stratification events are responsible for the observed near-surface N2O gradients, and that the gradients induce the strongest bias in gas exchange estimates at winds of about 3 to 6 m s−1. Glider hydrographic time series reveal that events of multi-day near-surface stratification are a common feature in the study region. In the same way as shorter events of near-surface stratification (e.g., the diurnal warm layer cycle), they preferentially exist under calm to moderate wind conditions, suppress turbulent mixing, and thus lead to isolation of the top layer from the waters below (surface trapping). Our observational data in combination with a simple gas-transfer model of the surface trapping mechanism show that multi-day near-surface stratification can produce near-surface N2O gradients comparable to observations. They further indicate that N2O gradients created by diurnal or shorter stratification cycles are weaker and do not substantially impact bulk emission estimates. Quantitatively, we estimate that the integrated bias for the entire Peruvian upwelling region in December 2012 represents an overestimation of the total N2O emission by about a third, if concentrations at 5 or 10 m depth are used as surrogate for bulk water N2O concentration. Locally, gradients exist which would lead to emission rates overestimated by a factor of two or more. As the Peruvian upwelling region is an N2O source of global importance, and other strong N2O source regions could tend to develop multi-day near-surface stratification as well, the bias resulting from multi-day near-surface stratification may also impact global oceanic N2O emission estimates.