The importance of short-term variability for constraining methane air–sea exchange in a coastal upwelling region

Short-term variability plays a key role in controlling air–sea CH₄ exchange in coastal upwelling systems, yet it is largely unresolved by conventional low-frequency sampling. Here, we quantify the influence of synoptic-scale variability on CH₄ content and its air–sea exchange using a buoy-based sensor system in a coastal upwelling bay off central Chile (Coliumo Bay, 36.5°S) during the upwelling season (September 2024–February 2025).

Spectral and wavelet analyses revealed a multiscale structure in surface CH₄ levels and alongshore winds, with variance dominated by periods >10 d and 3–10 d in about ~52-21% and ~40-31%, respectively. The latter variability, comprising synoptic oscillations, was mainly associated with alternating periods of active upwelling and relaxation/downwelling events.

At the synoptic scale, during active upwelling events, CH₄ effluxes averaged 25.38 ± 17.74 μmol m⁻² d⁻¹ whereas during relaxation periods effluxes were reduced by almost half (mean ± SD: 9.16 ± 9.58 μmol m⁻² d^-1). These results indicate that during active upwelling events, the advection of subsurface waters rich in CH₄ and wind-driven gas transfer are key factors triggering the highest CH₄ effluxes.

When the high-frequency time series is compared with a long-term (2007–2025) monthly time series from the same upwelling system, clear differences in capturing real variability emerge. Based on monthly sampling over 18 years, air–sea CH₄ fluxes were on average 9.43 ± 6.95 μmol m⁻² d^-1, with a weak seasonal contrast between upwelling-favorable and non-upwelling seasons (10.5 vs. 7.5 μmol m⁻² d⁻¹). These results demonstrate that synoptic variability in CH₄ concentration and air–sea exchange exceeds seasonal variability.

An uncertainty analysis accounting for aliasing under coastal upwelling conditions indicates that high-frequency observations capture CH₄ dynamics that are otherwise missed, thereby reducing bias in coastal CH₄ emission estimates. Our results underscore the need to incorporate high-frequency observations, as episodic events such as wind pulses, extreme rainfall, or atmospheric rivers, together with non-linear surface biogeochemical CH₄ production, are required to achieve a more realistic quantification of CH₄ emissions from coastal upwelling systems. Main funding FONDECYT (Chile) N° 1250210

.