Research
Methane
Methane is a potent greenhouse gas whose role in future climate warming will hinge on both anthropogenic emissions and perturbation of the natural methane budget. This motivates efforts to quantify and understand natural methane sources to the atmosphere.
Marine emissions: The global ocean is a small but highly uncertain source of methane to the atmosphere. Our work has narrowed this uncertainty range by employing machine learning methods to map methane supersaturation in the surface ocean and compute global sea-air fluxes (Weber et al. 2019, Nature Communications). The modeled methane distribution is consistent with a previously hypothesized aerobic source in the surface mixed layer, related to primary production.
Probability density functions for oceanic methane emissions, integrated regionally (left) and globally (right). The methane flux is estimated using machine learning mapping methods: Artificial Neural Networks (ANN) and Random Regression Forests (RRF). Adapted from Weber et al., 2019.