Original Research Papers

Methane budget estimates in Finland from the CarbonTracker Europe-CH4 data assimilation system

Authors:

Abstract

We estimated the CH4 budget in Finland for 2004–2014 using the CTE-CH4 data assimilation system with an extended atmospheric CH4 observation network of seven sites from Finland to surrounding regions (Hyytiälä, Kjølnes, Kumpula, Pallas, Puijo, Sodankylä, and Utö). The estimated average annual total emission for Finland is 0.6 ± 0.5 Tg CH4 yr−1. Sensitivity experiments show that the posterior biospheric emission estimates for Finland are between 0.3 and 0.9 Tg CH4 yr−1, which lies between the LPX-Bern-DYPTOP (0.2 Tg CH4 yr−1) and LPJG-WHyMe (2.2 Tg CH4 yr−1) process-based model estimates. For anthropogenic emissions, we found that the EDGAR v4.2 FT2010 inventory (0.4 Tg CH4 yr−1) is likely to overestimate emissions in southernmost Finland, but the extent of overestimation and possible relocation of emissions are difficult to derive from the current observation network. The posterior emission estimates were especially reliant on prior information in central Finland. However, based on analysis of posterior atmospheric CH4, we found that the anthropogenic emission distribution based on a national inventory is more reliable than the one based on EDGAR v4.2 FT2010. The contribution of total emissions in Finland to global total emissions is only about 0.13%, and the derived total emissions in Finland showed no trend during 2004–2014. The model using optimized emissions was able to reproduce observed atmospheric CH4 at the sites in Finland and surrounding regions fairly well (correlation > 0:75, bias < ±7 ppb), supporting adequacy of the observations to be used in atmospheric inversion studies. In addition to global budget estimates, we found that CTE-CH4 is also applicable for regional budget estimates, where small scale (1° × 1° in this case) optimization is possible with a dense observation network.

Keywords:

CH<sub>4</sub> fluxatmospheric CH<sub>4</sub>Finlanddata assimilationflux estimation
  • Year: 2019
  • Volume: 71 Issue: 1
  • Page/Article: 1565030
  • DOI: 10.1080/16000889.2018.1565030
  • Published on 1 Jan 2019
  • Peer Reviewed