Science highlights from the Cape Verde Observatory (CVAO) Lee, J.D. a , Read, K.A. b , Carpenter, L.J. a , Lewis, A.C. a , Moller, S.J. a , Mendes, L.M. c , Fleming Z. d , Evans, M. J. e a National Centre for Atmospheric Science (NCAS), University of York, UK b Facility for Ground based Atmospheric Measurements (FGAM), NCAS, University of York, UK c Instituto de Naçional de Meteorologia e Geofísica (INMG), Mindelo, Cape Verde d NCAS, University of Leicester, UK e University of Leeds, UK
Cape Verde Observatory 0 15 330 30 10 300 60 wind speed / ms-1 5 0 0 270 90 5 240 120 10 210 150 15 180 Using Met office NAME model, Carpenter et al., 2011
Overview of Talk Atmospheric photochemistry • Discussion of O 3 measurements O 3 < 340nm • Comparison to GEOSCHEM • Influence of radiation and NO x on O 3 depletion O 3 O 1 D • CO record H 2 O NO • Comparison to GEOSCHEM NO 2 OH CH 4/ VOCs CO HO 2 HO 2 Peroxides CH3O2 HO 2 H 2 O 2 CO 2 NO O 2 HCHO CH3O O 3 NO 2
Tropospheric ozone record 60 O3 (monthly ave) 50 40 O3 (ppbV) 30 20 10 0 Aug/2006 Aug/2007 Aug/2008 Aug/2009 Aug/2010 winter spring winter spring 45.0 45.0 summer autumn summer autumn 40.0 40.0 O3 (ppbV) O3 (ppbV) 35.0 35.0 30.0 30.0 25.0 25.0 20.0 20.0 00:00 04:48 09:36 14:24 19:12 00:00 00:00 04:48 09:36 14:24 19:12 00:00 Diurnal cycle of O 3 by season for 2007 (left) and 2010 (right).
Ozone diurnals and halogen influence In 2008 we published a paper where we modelled the depletion and found that we had to invoke halogen chemistry to reproduce the measurements. 4 2 3 IO / pptv 2 BrO / pptv 1 1 0 0 -1 00 06 12 18 00 00 06 12 18 00 Measured ∆ O 3 GEOSCHEM ∆ O 3 Box model ∆ O 3 without halogens (MCM 3.1) 2 Box model ∆ O 3 with IO and BrO chemistry 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep -2 delta O 3 / ppb day -1 -4 -6 ∆ O 3 – ppb O 3 loss per day -8 Read et al., Nature , 2008 -10
Tropospheric O 3 budget • Photolysis plays a significant role although balanced to some extent by the entrainment term. • Competition between photochemical loss and production of O 3 through: - NO 2 + h ν (+O 2 ) → NO + O 3 NO + HO 2 → OH + NO 2 OH + RO (+O 2 ) → HO 2 + RO 2 ….leads to a nonlinear dependency of O 3 on NO x . • Lee et al. looked at this relationship and found that in this region there is a level of NO x at which there is a switch from net destruction to production of O 3 , the “O 3 compensation point” (Lee et al., J. Geophys. Res., 2009). • In the MBL bromine and iodine atoms have been shown 8 to be important in the role of O 3 destruction in this region 6 (Read et al., Nature, 2008). 4 2 delta O3 / ppbv in 8 hours X + O 3 → XO + O 2 0 XO + XO → 2X + O2 0 5 10 15 20 25 30 35 40 -2 XO + IO → X + OIO y = 0.2139x - 4.658 -4 R 2 = 0.368 XO + HO 2 → HOX + O 2 HOX + hu → OH + X -6 XO + NO → NO 2 + X -8 -10 • XO causes a decrease in the NO/NO 2 ratio but O 3 is not -12 increased because the X formed will likely destroy O 3. -14 NO / pptv
Delta O 3 , the role of photolysis and air mass history 2 2 1000 1000 • The modelled and 900 900 measured measured 0 0 monthly monthly measured follow the 800 800 average average GEOSCHEM GEOSCHEM same trend most of the 700 700 monthly monthly -2 -2 average average Delta O3 / ppbv Delta O3 / ppbv 600 600 time in-line with the Slr_W_Avg Slr_W_Avg -4 -4 500 500 dominance of the O 3 400 400 photolysis term. -6 -6 300 300 200 200 -8 -8 100 100 -10 -10 0 0 Oct/06 Oct/06 Apr/07 Apr/07 Nov/07 Nov/07 May/08 May/08 Dec/08 Dec/08 Jul/09 Jul/09 Jan/10 Jan/10 African (dust) Polluted Coastal African Atlantic Continental Atlantic Marine 100 90 • Better agreement when 80 % contribution of each sector 70 sampling Atlantic conditions than 60 coastal African conditions. 50 40 30 20 10 0 Mar/08 Apr/08 May/08 Jun/08 Jul/08 Aug/08 Sep/08 Oct/08 Nov/08
GEOS CHEM comparisons O 3 60 GEOSCHEM O3 CVAO O3 50 40 O3 (ppbV) 30 20 10 0 Oct/06 Feb/07 Jul/07 Dec/07 May/08 Oct/08 Mar/09 Aug/09 Jan/10
NOx 100 NO2 measured NO2 GEOS 75 NO2 / pptv 50 25 0 30 NO measured NO GEOS 25 20 NO / pptv 15 10 5 0 Oct-06 Apr-07 Oct-07 Apr-08 Oct-08 Apr-09 Oct-09
Delta O 3 and NO/NO 2 ratio • The modelled NO/NO 2 ratio is 2 0.75 generally too high although agrees better in 2009. measured 0.25 monthly average 0 • However NO 2 (and often NO- -0.25 GEOSCHEM see plot below) are also much Delta O3 / ppbv monthly average higher in the measurements -2 -0.75 compared to the model measured NO/NO2 ratio -1.25 (sometimes by 5x). -4 GEOSCHEM • NO correlates with delta O 3 in -1.75 NO/NO2 ratio the measurements and model -6 -2.25 Oct/06 Apr/07 Nov/07 May/08 Dec/08 Jul/09 Jan/10 8 6 y = 0.3531x - 2.8144 4 R 2 = 0.3537 • Not enough XO to explain the 2 delta O3 / ppbv in 8 hours difference in modelled to measured NO / 0 0 5 10 15 20 25 30 35 40 -2 NO 2 ratio y = 0.2139x - 4.658 -4 R 2 = 0.368 • Other processes going on -6 -8 -10 -12 -14 NO / pptv
GEOS / meas comparisons NOy 1000 NOy meas 900 NOy GEOS 800 700 600 NOy / pptv 500 400 300 200 100 0 100 90 80 % contribution of each sector 70 60 50 40 30 20 10 0 Oct/06 Mar/07 Aug/07 Jan/08 Jun/08 Nov/08 Apr/09 Sep/09 African (dust) Polluted Coastal African Atlantic Continental Atlantic Marine NOy = Sum NOx, NO 3 , HNO 3 , N 2 O 5 , PANs, HONO, alkyl nitrates and particulate nitrate.
CO and ethane comparison to GEOS CHEM 200 CO meas CO GEOS 175 CO (ppb) 150 125 CO / ppbv 100 75 50 25 0 Sep-06 Feb-07 Jul-07 Dec-07 May-08 Oct-08 Mar-09 Aug-09 Dec-09 May-10 Oct-10 2500 2500 ethane measured GEOS ethane 2000 2000 ethane (pptv) ethane (pptv) 1500 1500 1000 1000 500 500 0 0 Sep/06 Sep/06 Feb/07 Feb/07 Jul/07 Jul/07 Dec/07 Dec/07 May/08 May/0 Oct/08 Oct/08 Mar/0 Mar/09 Aug/09 Jan/10 Aug/09 Jan/10 Jun/10 Jun/10 Oct/10 Oct/10 Is the amplitude of the CO correct for the wrong reasons? Can the emissions be underestimated all year? If we increase the Spring emissions to similar to those needed for ethane then the amplitude is too large, need additional sources in summer.
Other sites CO with GFDL GCTM model Purple measured Black-modelled Red-Fossil fuel Blue- Biogenic hydrocarbon Green-Biomass Orange - Methane oxidation The relative contributions show that fossil fuel emissions dominate in late winter/early spring but other sources compete in summer. No seasonal distribution assumed in fossil fuel emissions so changes arise from changes in transport and OH conc. Methane oxidation is similar all year. We fall between Tenerife and Ascension, you can see the difference in the contributions between the two sites…… Biogenic hydrocarbon influence in summer…. Holloway et al., 2000
Using ethane to interpret CO • CO and ethane share some common sources e.g. fossil fuel emissions, and biomass burning. The ratio of CO : ethane OH dependent fit to CO 3500 emissions from these sources is not Sinusoidal fit to ethane 200 OH dependent fit to ethane 3000 thought to vary seasonally. 2500 150 ethane (pptV) • “n[OH]” can be calculated from ethane CO (ppbV) 2000 and applied to calculate a theoretical CO 100 1500 assuming no other sources exist and all loss routes are also shared. 1000 50 500 • If the modelled CO amplitude is correct 0 0 then is there a year-round underestimate Oct/06 Dec/06 Jan/07 Mar/07 May/07 Jul/07 Sep/07 Nov/07 Jan/08 Mar/08 of emissions? Read et al., J. Geophys. Res., 2009
Seasonal distribution 1500 'NCA COASTAL NAA AM AFR propane 24 per. Mov. Avg. (AirTC_Avg) 40 • Methanol shows variability with air 1000 mass trajectory. 35 acetone and propane (pptV) • Higher levels during Spring (Mar- Air temperature (°C) 500 Apr) in-line with the precursor NMHC 30 and their shared sources (later for acetone 2500 methanol than acetone). OH is also 0 acetaldehyde y = 13.52x + 622.33 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec lowest at this time. methanol R 2 = 0.64 2000 25 • June-radiation decreases. -500 y = 13.32x + 461.80 ovoc (pptv) 1500 R 2 = 0.53 • (Elevated levels are also observed in Aug-September correlating with y = 7.36x + 628.90 -1000 20 1000 3000 1980 maximum in air temperature R 2 = 0.83 NCA COASTAL NAA AM AFR methane (increased biogenic sources, 2500 1960 500 photochemical influence on 2000 1940 secondary chemistry). 0 1500 1920 • Source of CO from NMVOC is a -40 -20 0 20 40 60 80 100 methanol (pptV) methane (ppbV) cause of high uncertainty (>factor of 1000 1900 CO measurements - CO fit (ppbv) 3) in models (Holloway et al., 2000) 500 1880 Potentially up to 60% of the observed 0 1860 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec summer CO is from summertime -500 1840 increases in the seasonal trend of NMVOC (and methane) oxidation. -1000 1820 Read et al., J. Geophys. Res., 2009 -1500 1800
Recommend
More recommend
