ULg contributions to AGACC-II: main results for 2013 Team members: W. Bader, B. Bovy, P. Demoulin,...

39
ULg contributions to AGACC- ULg contributions to AGACC- II: II: main results for 2013 main results for 2013 Team members: W. Bader, B. Bovy, P. Demoulin, O. Flock, B. Franco, B. Lejeune, E. Mahieu, C. Servais, D. Zander [email protected]. be

Transcript of ULg contributions to AGACC-II: main results for 2013 Team members: W. Bader, B. Bovy, P. Demoulin,...

ULg contributions to AGACC-II: ULg contributions to AGACC-II: main results for 2013main results for 2013

Team members:

W. Bader, B. Bovy, P. Demoulin, O. Flock, B. Franco, B. Lejeune, E. Mahieu, C. Servais, D. Zander

G. Roland and R. Zander [email protected]

ULg involvement• Contribution to WP1 (monitoring of direct GHGs):

– Deliver long-term time series of major GHGs (CO2, CH4, N2O), including the exploitation of historical grating observations

– Investigations of CCl4 and CF4: retrieval strategies and trends• Contribution to WP2 (VOCs & CFC substitutes):

– Development or optimization of retrieval strategies for CH3Cl, CH3OH, C2H6, C2H4 (see Corinne’s talk) ,H2CO and CFC-substitutes

• Contribution to WP5 (outreach): publications, archiving…• Essentially based on the exploitation of long-term (FT)IR

measurements under clear-sky conditions at the Jungfraujoch station (NDACC site, Swiss Alps, 46.5ºN, 8.0ºE, 3580 m asl):– Grating IR spectra available since 1976– FTIR spectra since 1984, current measurement density: >130 d/yr

WP1 results

Analysis of historical grating spectra: search for appropriate

CO2 lines/windows

The Jungfraujoch grating spectrometer

• A 7.3 m focal length Ebert/Fastie grating spectrometer was installed at the Jungfraujoch in the late 1950s, it was first used to record solar atlases (IR, vis and near-UV)

• As of 1976, the instrument was used in double-pass mode to record narrow atmospheric solar spectra (InSb detector), encompassing absorption lines of specific target gases, e.g. HF, HCl, N2O, CH4

• Various diffraction orders have been used over 1976 to 1989 => several observational subsets

Searching for CO2 lines…

• Conversely to CH4 and N2O, CO2 was not a specific target of the pioneering observational effort

• Four windows encompassing CO2 lines have been identified (resolution between 7 and 13 mK):

• BUT first FTIR data in 1984 => no ext. back in time…

Spectral range (cm-1) Time period # days (sp.) Comment

2481.3 – 2484.9 (A) 11/1983 – 07/1985 24 (61) Weak feature

2437.2 – 2442.1 (B) 08/1985 – 10/1989 119 (789)

2416.0 – 2419.9 (C) 10/1985 – 09/1989 93 (1016)

2386.8 – 2394.3 (D) 10/1985 – 09/1989 74 (412)

Preliminary fittings (B-D windows)

WP1 results (ctd)

Retrievals of carbon tetrafluoride (CF4) from ground-based FTIR spectra

Carbon tetrafluoride – CF4

• The retrieval strategy (simple scaling) has been finalized: two consecutive runs allow improving the precision of the retrievals; the 2nd window focuses on the most interference-free features of CF4 (i.e. avoiding the center of the window (HDO & H2O))

• A two-decade time series has been produced and analyzed, the “wet-spectra” are excluded

• Mean relative standard deviation around the daily averages (inv. ≥ 4 meas.): <1%!

Carbon tetrafluoride – CF4

• First FTIR time series for this very long-lived GHG

• Total column available from 1989 onwards

• Error budget dominated by the uncertainty affecting the CF4 spectroscopic parameters

• Continuous increase (+35% over 1989-2012), but with two regimes

Derived global emissions: 15.8 ± 1.3 Gg/yr over 1989-1997 11.1 ± 0.2 Gg/yr over 1998-2012

In very good agreement with evaluations based on AGAGE measurements

Carbon tetrafluoride – CF4

• Good agreement with in situ, balloon and satellite data, when accounting for the uncertainties affecting the data sets and for the time needed for CF4 to propagate and mix throughout the atmosphere

• Anthropogenic contribution larger than the natural background since about the late 90s

Mahieu et al., accepted for publication in AMT, 2013

End of WP1

WP2 results – VOCs

Development of a retrieval strategy for methanol (CH3OH) from

Jungfraujoch

Methanol – Finalized retrieval strategy

Simulations for Jungfraujoch, 80°, 6.1 mK using HITRAN08

Rinsland et al., 2009

992 – 998.7 cm-1

Stavrakou et al., 2011

1029 – 1037 cm-1

Mahieu et al., 2012

+ Atmospheric Chemistry Experiment, 2011

995.5 – 1008.3 cm-1

Methanol – Finalized retrieval strategy

Simulations for Jungfraujoch, 80°, 6.1 mK using HITRAN08

Rinsland et al., 2009 Stavrakou et al., 2011

1029 - 1037 cm-1

Mahieu et al., 2012

+ Atmospheric Chemistry Experiment, 2011

992 – 1008.3 cm-1 Improvement of the DOFSfrom 1.5 to 1.8

CH3OH: typical information content

• DOFS ≈ 1.8• 1st Eigen vector: tropospheric column• 2nd Eigen vector: some vertical resolution (LT & UTLS)

Methanol Seasonal Modulation and Variability• No long-term trend over 1995-2012• Minimum : December to February• Maximum : June-July• Peak-to-peak amplitude : ≈130 % of the CH3OH yearly mean

Lower-Tropospheric mixing ratios

• Intercomparison with in situ campaign-type measurements in 2005 (Empa)• Very few coincident measurement

UTLS Partial Columns

• Intercomparison with occultation measurements by ACE-FTS (7-14 km) between 41 and 51ºN• Time period: 2004-2012

WP2 results – VOCs (ctd)

Retrieval of formaldehyde (H2CO) from FTIR observations and comparison with

MAX-DOAS and model results

- Retrieval SFIT-2 algorithm (v3.91)

- 6 microwindows(Vigouroux et al., Atmos. Chem. Phys., 9, 9523-9544, 2009)

- Spectral resolution: 0.005 cm-1

- Spectroscopic parameters: HITRAN 2008

- Data set: 9481 solar spectra from January 2005 to April 2013

- SZA: 0° - 88°

- Tikhonov type L1 regularization

- A priori profile from 1980 – 2020 WACCM v.6 simulations

Constraint matrix R = α L1T x L1

Microwindows (cm-1) Interfering Species

2763.425 – 2763.600 HDO, CH4, O3, N2O, CO2

2765.725 – 2765.975 HDO, CH4, O3, N2O, CO2

2778.200 – 2778.590 HDO, CH4, O3, N2O, CO2

2780.800 – 2781.150 HDO, CH4, O3, N2O, CO2

2810.000 – 2810.350 HDO, CH4, N2O, CO2

2855.650 – 2856.400 HDO, CH4, O3, N2O

Retrieval of formaldehyde – H2CO

A priori profile from 1980 – 2020 WACCM simulations (v6)

TOTAL AVERAGING KERNEL x 0.10

• Limited information content: typical DOFS ≈ 1 • Sensitivity range: lower troposphere

Retrieval of formaldehyde – H2CO

Individual measurements

Retrieval of formaldehyde – H2CO

Monthly means

Retrieval of formaldehyde – H2CO

Retrieval of formaldehyde – H2CO

HITRAN 2008 2012

FITTING RESID. (%) 0.283 0.277

DOFS 1.043 1.038

Relative StDev (%) 11.99 12.85

Rejected spectra (%) 3.34 27.19

AVERAGED BIAS = -37.70%!!

Test of HITRAN 2012…

WP2 results – CFC substitutes

Definition of a retrieval strategy for HCFC-142b

Retrievals of HCFC-142b (CH3CClF2)• Lifetime of 18 years, GWP of 2310• Its atmospheric content has been increasing to reach ≈ 21 ppt

in 2011, with variable rates over the recent years: 1.1 ppt/yr in 2005-2008; 0.7 ppt/yr in 2009-2012, annual emissions on the order of 30-40 Gg

• Four candidate windows have been evaluated, around 903, 967, 1133 and 1193 cm-1 (used for ACE-FTS retrievals)

• HITRAN08 was used• For HCFC-142b, pseudolines based on Clerbaux et al. (1993)

cross-sections were assumed. Since then, improved laboratory measurements have been produced by Le Bris & Strong (2009), but no pseudolines available…

Two very unfavorable cases…

Here, strong CO2 or H2O lines prevent the retrieval of F142b from the ground…

“Strong” signal, but HDO and N2O… Weaker signal, but less critical interferences…

Simulation & fit for the 900-906 cm-1 window

Weak absorption, even at low sun (0.5% for 85°), but relatively clean window

Main interferences by HNO3 manifolds, CO2 and H2O lines

Red residuals when assuming no F142b in the atmosphere

Retrieval of F142b• Retrievals in the 903 and 1193 windows look

reasonable, but are in disagreement (the 1193 columns are too high: interferences or inconsistent cross-sections?)

• The 900-906 cm-1 microwindow was finally selected despite a very weak absorption (0.5% on average)

• A time series has been produced for the Jungfraujoch, from 1990 onwards (daily mean data from 2000 onwards)

• Comparison with in situ surface data (Mace Head and Jungfraujoch) shows good overall agreement (absolute values and trend)

Jungfraujoch time series

Trends (2-σ):

(1.05±0.08) 1013 molec./cm2/yr

(6.5±0.5)%/yr (2000.0 as ref.)

WP5 – Outreach

Publications, communications, assessments, archiving…

Long-term time series of halogenated source gases above Jungfraujoch

• Time series and trends solicited for inclusion in the next WMO assessment on ozone depletion

• Not shown (scale): SF6 and HCFC-142b

• CCl4, CF4 and F142b thanks to AGACC-II funding!

AGACC-II publications & communications 2013

Peer-reviewed:• Mahieu et al., Spectrometric monitoring of atmospheric carbon tetrafluoride (CF4)

above the Jungfraujoch station since 1989: evidence of continued increase but at slowing rate, accepted for publication in Atmos. Meas. Tech., 2013.

Communications:• Bader et al., Evolution of methanol (CH3OH) above the Jungfraujoch station

(46.5ºN): variability, seasonal modulation and long-term trend, poster at the EGU 2013 General Assembly, Vienna, 2013.

• Kolonjari et al., Validation of ACE-FTS measurements of CFC-11, CFC-12 and HCFC-22 using ground-based FTIRs, poster at AGU Fall meeting, San Fransisco, 2013.

• Mahieu et al., First retrievals of HCFC-142b from ground-based high-resolution FTIR solar observations: application to high-altitude Jungfraujoch spectra, poster at the EGU 2013 General Assembly, Vienna, 2013.

• Mahieu et al., Overview of the geosphysical data derived from long-term FTIR monitoring at the Jungfraujoch NDACC Site (46.5ºN), poster at the 6th GEOS-Chem meeting, Harvard University, Cambridge, 2013.

Thank you!

Including contributions from BIRA-IASB and Empa (CH) colleagues as well as from ACE-science team members (UWaterloo).

Pseudolines from G.C. Toon (NASA-JPL)