Estimation of country-scale methane emissions by airborne and ground-based in situ observations and inverse modeling

Abstract:

Methane (CH₄) is the second most important long-lived greenhouse gas from both natural and anthropogenic sources. In Switzerland, CH₄ sources are dominated by agricultural activities (>80%) while natural emissions from wetlands and wild animals are thought to represent a minor source (~3 %). Except for leakage from the natural gas distribution network, all relevant emissions in Switzerland are associated with microbial processes which, due to their diffuse nature and sensitivity to ambient conditions, are associated with comparatively large uncertainties. The Swiss National Greenhouse Gas Inventory, for example, assigns an uncertainty of 18% to the country total CH₄ emissions as compared to only 3% for CO₂.

To verify the Swiss national CH₄ emission estimate and to reduce its uncertainty, “top-down” methods combining atmospheric observations and regional scale transport simulations can be used. Here, we present two independent analyses of the Swiss CH₄ emission budget using inverse modeling. The first is based on airborne observations during more than 20 flights of a motor glider conducted in the framework of the MAIOLICA project. The second is based on near surface measurements from the newly established CarboCount CH measurement network (http://www.carbocount.ch). A Bayesian inversion framework is applied to these observations in combination with source sensitivities (footprints) calculated with the Lagrangian Particle Dispersion Model FLEXPART. To account for the complex terrain and flow in Switzerland, FLEXPART is driven by meteorological fields from the non-hydrostatic numerical weather forecast model COSMO at horizontal resolutions of up to 2 km x 2 km. Due to the critical role of the transport simulations, we first present an analysis of the sensitivity to different model configurations (e.g. resolution, time-averaged winds versus instantaneous fields, ECMWF versus COSMO fields). We then present the inverse modeling results separately for the airborne and the ground-based observations contrasting the numbers to the official National Greenhouse Gas Inventory.