A new, 4-dimensional variational (4DVAR) inverse modelling system has been developed for inverse modelling of atmospheric methane. The main advantage of the new system is that it allows optimizing emissions of individual model grid cells. At the same time very large sets of observational data can be used (e.g. high-frequency in situ measurements or satellite data). In contrast, the previously widely used synthesis inversions were restricted to the optimization of emissions of larger, predefined regions (e.g. continents, or countries), but could not further optimize spatial emission distributions within these pre-defined regions. Therefore, these approaches were prone to the so-called aggregation error [Kaminski et al., 2001].
P Bergamaschi, JF Meirink, M Krol, MG Villani. New TM5-4DVAR inverse modelling system to estimate global and European CH4 sources
2007, 2007, JRC