On the need for assuming imperfect prior knowledge of emissions in regional CO2 inversions
•
0 likes•224 views
Presented by Christoph Gerbig at the 2nd ICOS Science Conference in Helsinki, Finland, 27-29 Sep 2016.
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Viewers also liked
Viewers also liked (19)
Similar to On the need for assuming imperfect prior knowledge of emissions in regional CO2 inversions
Similar to On the need for assuming imperfect prior knowledge of emissions in regional CO2 inversions (20)
More from Integrated Carbon Observation System (ICOS)
More from Integrated Carbon Observation System (ICOS) (20)
Recently uploaded
Recently uploaded (20)
On the need for assuming imperfect prior knowledge of emissions in regional CO2 inversions
- 1. ICOS science conference, Sept. 27-29 2016, Helsinki On the need for assuming imperfect prior knowledge of emissions in regional CO2 inversions Christoph Gerbig, Panagiotis Kountouris, Fabio Boschetti, Christian Rödenbeck (MPI-BGC, Jena), Thomas Koch (DWD), Ute Karstens (ICOS-CP, Lund) Max Planck Institute for Biogeochemistry ICOS science conference, Sept. 27-29 2016, Helsinki
- 2. ICOS science conference, Sept. 27-29 2016, Helsinki Motivation • Current regional inverse modeling of CO2: ‣ inverse transport modeling targeted biosphere- atmosphere exchange only ‣ fossil fuel emissions assumed much better known than biospheric fluxes ‣ uncertainty in transport models more important (Peylin et al., 2011) ‣ ICOS atmospheric network not targeted at emissions
- 3. ICOS science conference, Sept. 27-29 2016, Helsinki Motivation • Future regional inverse modeling of CO2: ‣ political pressure from stakeholders to assess emissions ‣ INDCs (Intended Nationally Determined Contributions) need regular verification ‣ more ICOS stations with emission influence ‣ spatial resolution of inversion transport models increases ‣ uncertainty in increases with decreasing scales for spatiotemporal disaggregation ‣ observations „see“ these uncertain fluxes
- 4. ICOS science conference, Sept. 27-29 2016, Helsinki Approach • Assessment of differences in various emission datasets • Analysis of resulting emission signals at atmospheric stations • Inverse transport modeling of biosphere- atmosphere exchange using different emission datasets
- 5. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions EDGAR v4.1 + BP2012 EDGAR v4.3 + BP2014 EDGv4.3 + BP + IER (D+F) ODIAC 2015a
- 6. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions EDGAR v4.1 + BP2012 EDGAR v4.3 + BP2014 ODIAC 2015aEDGv4.3 + BP + IER (D+F) (EDGv4.3blend - ODIAC) _____________________ 0.5 x (EDGv4.3blend + ODIAC)
- 7. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions IER Stuttgart Germany IER Stuttgart Germany
- 8. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions IER Stuttgart Germany 90% largest emissions IER Stuttgart Germany 90% largest emissions
- 9. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions IER Stuttgart Emissions FranceIER Stuttgart Emissions Germany 90% of emissions from 20% of the area 90% of emissions from 30% of the area 1.5 km resolution: 1.5 km resolution:
- 10. ICOS science conference, Sept. 27-29 2016, Helsinki Spatial distribution of CO2 emissions IER Stuttgart Emissions FranceIER Stuttgart Emissions Germany 90% of emissions from 65% of the area 90% of emissions from 75% of the area 96 km resolution:96 km resolution:
- 11. • STILT-ECMWF using 0.25 deg. resolution met fields • STILT Footprints to provide sensitivity of observations to upstream emissions • Linking footprints to different emission inventories at different spatial resolution - EDGAR v4.3 + IER (D + F) blend @ 80, 10, 6, and 1.5 km - ODIAC 2015a @ 0.75 km STILT simulations of CO2 fossil fuel signals STILT footprints Boschetti et al., 2015, Tellus-B
- 12. • STILT-ECMWF using 0.25 deg. resolution met fields • STILT Footprints to provide sensitivity of observations to upstream emissions • Linking footprints to different emission inventories at different spatial resolution - EDGAR v4.3 + IER (D + F) blend @ 80, 10, 6, and 1.5 km - ODIAC 2015a @ 0.75 km STILT simulations of CO2 fossil fuel signals STILT footprints Boschetti et al., 2015, Tellus-B
- 13. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions regionalCO2emissionsignal[ppm] afternoon values (11:00-17:00) only
- 14. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions regionalCO2emissionsignal[ppm] afternoon values (11:00-17:00) only
- 15. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions regionalCO2emissionsignal[ppm] afternoon values (11:00-17:00) only
- 16. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions regionalCO2emissionsignal[ppm] afternoon values (11:00-17:00) only
- 17. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions regionalCO2emissionsignal[ppm] afternoon values (11:00-17:00) only
- 18. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions
- 19. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions
- 20. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions CO2 emissions in 2009 : (www.carma.org) Weissweiler 19.200.000 Tons CO2/yr Niederaussem 26.300.000 Tons CO2/yr Neurath 90.650. 000 Tons CO2/yr Frimmersdorf 2.119.600 Tons CO2/yr 10 km JUE Future ICOS station Juelich
- 21. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions CO2 emissions in 2009 : (www.carma.org) Weissweiler 19.200.000 Tons CO2/yr Niederaussem 26.300.000 Tons CO2/yr Neurath 90.650. 000 Tons CO2/yr Frimmersdorf 2.119.600 Tons CO2/yr 10 km JUE Future ICOS station Juelich
- 22. ICOS science conference, Sept. 27-29 2016, Helsinki CO2 fossil fuel signals @ different resolutions CO2 emissions in 2009 : (www.carma.org) Weissweiler 19.200.000 Tons CO2/yr Niederaussem 26.300.000 Tons CO2/yr Neurath 90.650. 000 Tons CO2/yr Frimmersdorf 2.119.600 Tons CO2/yr 10 km JUE Future ICOS station Juelich
- 23. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes June 2014 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories using EDGAR v4.1 + BP2012 Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 24. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes June 2014 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories using EDGAR v4.3 + BP2014 Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 25. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes June 2014 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories using EDGAR v4.3 + BP2014 + IER (D + F) Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 26. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes January 2014 using EDGAR v4.1 + BP2012 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 27. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes January 2014 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories using EDGAR v4.3 + BP2014 Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 28. ICOS science conference, Sept. 27-29 2016, Helsinki Posterior fluxes January 2014 biosphere-atmosphere flux [PgC/a] STILT-TM3 inversions using different emission inventories using EDGAR v4.3 + BP2014 + IER (D + F) Inversion system: P. Kountouris et al. 2016a and 2016b (ACPD)
- 29. ICOS science conference, Sept. 27-29 2016, Helsinki STILT-TM3 inversions using different emission inventories using EDGAR v4.3 + BP2014 + IER (D + F) using EDGAR v4.3 + BP2014 using EDGAR v4.1 + BP2012 PRIOR annual NEE (GtC/a): POSTERIOR annual NEE (GtC/a): annual fossil fuel emissions (GtC/a): 1.374 1.416 1.416 -0.49 -0.53 -0.74 -1.24 -1.24 -1.24 Fraction of ∆FF „recovered“ as NEE: 95% 595%
- 30. ICOS science conference, Sept. 27-29 2016, Helsinki Synergy of tracers CO2CH4 Photo- synth. Resp. Anthr. emiss. CO
- 31. ICOS science conference, Sept. 27-29 2016, Helsinki Synergy of tracers CO2CH4 Photo- synth. Resp. Anthr. emiss. CO • shared atmospheric transport
- 32. ICOS science conference, Sept. 27-29 2016, Helsinki Synergy of tracers CO2CH4 Photo- synth. Resp. Anthr. emiss. Oil Anthr. emiss. Coal Anthr. emiss. Gas CO • shared atmospheric transport • shared fuel types
- 33. ICOS science conference, Sept. 27-29 2016, Helsinki Synergy of tracers Anthr. emiss. Gas Anthr. emiss. Coal Anthr. emiss. Oil Anthr. emiss. Gas Anthr. emiss. Coal Anthr. emiss. Oil Anthr. emiss. Gas Anthr. emiss. Coal Anthr. emiss. Oil Anthr. emiss. Gas Anthr. emiss. Coal Anthr. emiss. Oil Anthr. emiss. Gas Anthr. emiss. Coal Anthr. emiss. Oil CO2CH4 Photo- synth. Resp. Anthr. emiss. Oil Anthr. emiss. Coal Anthr. emiss. Gas CO • shared atmospheric transport • shared fuel types • shared emission sectors
- 34. ICOS science conference, Sept. 27-29 2016, Helsinki Synergy of tracers EDGAR v4.3 + BP2014 + IER (D + F) CO2 CO CH4 micro-molesC/m2/s
- 35. ICOS science conference, Sept. 27-29 2016, Helsinki tracer correlations for sub-grid variations correlation CO2 - CO correlation CO2 - CH4 • Linking footprints to emission inventories for CO2, CO and CH4 • different spatial resolution 80km, 10km, 6km, 1.5km • Difference to 80km
- 36. ICOS science conference, Sept. 27-29 2016, Helsinki Conclusion/Outlook • Emissions have increasing uncertainty with decreasing spatial disaggregation scale, leading to potential bias in retrieved biosphere-atmosphere exchange ‣ in combination with increasing number of stations under influence from emissions ‣ in combination with increasing resolution of inverse transport models • Potential benefit from multi-species inversions ‣ provide a clearer link between inverse modeling and UNFCCC reporting (sectors, fuel types) ‣ shared uncertainties (shown here: influence from sub grid variability) • Implementation @ ICOS-CP (Carbon Portal): ‣ STILT footprints for ICOS atmospheric stations ‣ EDGARv4.3 + BP2015 emission estimates (sector- + fuel-specific)