Presentation is loading. Please wait.

Presentation is loading. Please wait.

Toward a more accurate estimate of global stratospheric aerosol surface area density. Is it important? T. Deshler, J. L. Mercer, M. Kovilakam, J. M. Rosen.

Similar presentations


Presentation on theme: "Toward a more accurate estimate of global stratospheric aerosol surface area density. Is it important? T. Deshler, J. L. Mercer, M. Kovilakam, J. M. Rosen."— Presentation transcript:

1 Toward a more accurate estimate of global stratospheric aerosol surface area density. Is it important? T. Deshler, J. L. Mercer, M. Kovilakam, J. M. Rosen Univ. of Wyoming, Laramie, WY D. J. Hofmann, S. Solomon, J. F. Lamarque, P. J. Young NOAA Earth System Research Laboratory, Boulder, CO History of stratospheric aerosol (UW in situ instruments)History of stratospheric aerosolUW in situ instruments Present climatology – SAGE II+ and times of concern –1971 - 1984 –Post Pinatubo – low aerosol loading Why we care Comparison with in situ measurements and fixing the climatologyComparison with in situ measurements and fixing the climatology –Comparisons over Laramie –Broadcasting (Surface area discrepancies (SAGE/UW)Surface area discrepancies (SAGE/UW Using the new climatology – results from Chem-CAM (3D) – 1970sUsing the new climatology – results from Chem-CAM (3D) – 1970s –NOx, ClOx, Ozone Conclusions (What’s new)What’s new

2 Summary Present aerosol surface area density climatology has some deficiencies, 2000 A new climatology corrected with in situ measurements is available along with the in situ measurements used to develop it. See: http://www-das.uwyo.edu/~deshler/ http://www-das.uwyo.edu/~deshler/ First results with the new climatology show differences in NOx, ClOx and OH leading to ± 0.5% for global average ozone in the 1970s when stratoshperic chlorine was 1500 – 2000 ppt. A total of 1% change does not sound like much, but recall globally the mid latitude ozone loss through the 1980s was on the order of 5%. Future work –Fix blanks at pressures > 100 hPa in El Chichon period –Smooth climatology at pressures < 20 hPa, where signal is very weak –Use the new climatology in model runs, 1980-1985, and 1991-2010 There are many people and agencies to thank for these results –Funding over the last 40 years – NSF, NSF, NSF, NASA, NRL,... –The pioneers Jim Rosen and Dave Hofmann, and their (and my) engineers, technicians, scientists and students necessary to complete the measurements. –Susan Solomon for the invitation to a sabbatical and providing the means to get introduced to the world of atmospheric models, and to Jean Francois Lamarque and Paul Young for guiding me into the details

3

4 Update through 2009 of Figure 1 from Deshler, T. (2008), A Review of Global Stratospheric Aerosol: Measurements, Importance, Life Cycle, and Local Stratospheric Aerosol, Atmos. Res., 90, 223-232. SAGESAGE II Satellite Record Four Lidar Records In Situ Record Two altitudes

5 Periods of concern Current Surface area density used in Atmospheric models

6 Why we care N 2 O 5 + H 2 O (aer)  2HNO 3 Consequences of this conversion –Less N 2 O 5 for: N 2 O 5 + hν  NO 2 + NO 3 –Less NO 2 for: NO 2 + ClO + M  ClONO 2 NO 2 + OH + M  HNO 3 –More ClO for HO 2 + ClO  HOCl + 0 2, HOCl + hv  OH + Cl Cl + 0 3  ClO + 0 2 –More OH for: OH + 0 3  H0 2 + O 2 HO 2 + 0 3  OH + 0 2 + 0 2 Net result –Less ozone from reactions with ClO and OH –More ozone from reduction in loss from NOx

7 Revising the climatology using In Situ Aerosol Profiles University of Wyoming – with particular thanks to Jim Rosen and Dave Hofmann http://www-das.uwyo.edu/~deshler/ –US_Laramie_41N_105W (1971 - 2009) –AU_Mildura_34S_142W (1972 - 1980) –NZ_Lauder_45S_170E (1991 - 2001) –Ant_McMurdo_78S_167E (1989 - 2008) –SE_Kiruna_68N_21E (1991 - 2004) –Miscellaneous Brazil, Niger, France

8 Aerosol concentration for particles > 0.01, 0.15, 0.25 µm Oct 1971 Oct 1974 Fuego

9 Mar 1991 July 1991 Pinatubo Aerosol concentration for particles > 0.01, 0.15, 0.25, … 2.0/10.0 µm

10 Fixing the climatology –Surface Area Comparisons over Laramie with SAGE II –Resultant ratios – UW : SAGE –Broadcasting Comparison with far flung measurements –Mildura Australia (1972-1980) –Lauder New Zealand (1991-2001) Led to determining rate of latitudinal spread and weighting functions for dispersal from eruptions

11 Fuego

12 Post - Pinatubo Moderate aerosol load Particles large enough to Allow extinction Measurements represent Surface area

13 Low aerosol load Extinction measurements Can’t see the small particles Controlling surface area

14 1988 – 1991 Pre Pinatubo Aerosol load high enough for extinction meas. to obtain good estimates of surface area Peak of Pinatubo SAGE blinded Decay of Pinatubo aerosol in good range for extinction measurements Post Pinatubo low aerosol load

15 Comparison of revised climatology with measurements

16

17 Periods of concern Period investigated With CAM - chem

18

19 NOx New / Old Climatology Results CAM – chem run through 1970s

20 Revised Climatology Fuego NOx Profiles 40 N New / Old Climatology Results CAM – chem run through 1970s

21 Revised Climatology Fuego NOx profiles 40 S New / Old Climatology Results CAM – chem run through 1970s

22 Results CAM – chem run through 1970s

23 ClO New / Old Climatology Results CAM – chem run through 1970s

24 Revised Climatology Fuego ClO 40 N New / Old Climatology Results CAM – chem run through 1970s

25 Revised Climatology Fuego ClO 40 S New / Old Climatology Results CAM – chem run through 1970s

26 Ozone New / Old Climatology Results CAM – chem run through 1970s

27 Revised Climatology Fuego Ozone 40 N New / Old Climatology 40 S is similar Results CAM – chem run through 1970s

28 Revised Climatology Fuego Total ozone global average New / Old Climatology Results CAM – chem run through 1970s

29 Wyoming Optical Particle Counter, r > 0.2 - 10.0  m White light Pump PMTs PHA  Processor Optical Chamber Aerosol in

30 Low Gain Mid Gain High Gain

31 Initial Measurement Sizes 1971 - 19901990 - present Post El Chichon

32

33 Sources of Error Size Aerosol index of refraction and shape Uniformity of illumination PMT response – pulse width broadening Number concentration Flow rate Coincidence Poisson counting uncertainty

34

35

36 Estimates of differential surface area (  m 2 cm -3 ), extinction (km -1 ), and volume (  m 3 cm -3 ) for in situ measurements in 1993 and 1999 Evolution of Pinatubo Aerosol

37 Radius dependence of SAD and surface area kernel

38 a) b

39

40

41 Comparison of SAGE v6.0 and 6.1 with UWOPC extinction – 525 1020 nm

42


Download ppt "Toward a more accurate estimate of global stratospheric aerosol surface area density. Is it important? T. Deshler, J. L. Mercer, M. Kovilakam, J. M. Rosen."

Similar presentations


Ads by Google