1 Recovery of the Antarctic ozone hole P. Newman 1, E. Nash 1, S. R. Kawa 1, S. Montzka 2, Susan Schauffler 3, R. Stolarski 1, S. Pawson 1, A. Douglass 1, J. E. Nielsen 1, S. Frith 1 University College Dublin, Sept. 21, 2006 Introduction Ozone Hole trends CCM model prediction of ozone hole Parametric model Controlling factors Model outline Predictions of Recovery Estimating recovery Uncertainties Climate Change and Recovery Summary 1 NASA/GSFC, 2 NOAA/ESRL, 3 NCAR

2 Introduction

3 Why is understanding ozone hole recovery important? The ozone hole is the poster child of atmospheric ozone depletion Scientists staked their reputations on ozone depletion - international regulations were implemented. We need to carry our predictions through. Severe ozone holes lead to acute UV events in mid-latitudes Possible regulation changes could accelerate the phase out of ozone depleting chemicals. The ozone hole is a fundamental example of mankind’s ability to alter our atmosphere and climate - forming a useful example on climate change policy

4 Ozone Hole Trends

5 TOMS 1984 South America Antarctica Green-blue indicates low ozone values, while orange-red indicated high values High values are normally found in the mid-latitudes Extremely low values Extremely cold temperatures are found in the lower stratosphere in spring and fall Strong jet stream (the polar vortex) acts to confine ozone losses over Antarctica October 1984 TOMS total ozone

6 October Average Ozone Hole Low Ozone High Ozone Halley Bay Station

7 October Antarctic Ozone

8 ozonewatch.gsfc.nasa.gov

9 Defining the Hole Antarctic Ozone Hole on Oct. 4, 1998 Ozone hole area is defined by the area coverage of ozone values less than 220 DU = 24.7 M km DU located near strong gradient 220 DU is lower than values observed prior to 1979 Values of 220 tend to appear in early September. TOMS doesn’t make measurements in polar night! Values of 220 tend to disappear in late November Ozone hole minimum is 94 DU

10 Daily Ozone Hole Area 24.7 M km 2 on Oct. 4, 1998 Derive average size from an average of daily values: Sep. 7-Oct. 13

11 Seasonal Ozone Hole Area

12 Current Conditions

13 Sept. 17, 2006 Aura OMI Ozone < 220 DU

14 Current Conditions

15 Assessment of the ozone hole’s recovery (WMO, 2003) Chapter 3 - Polar Ozone

16 Model area estimates WMO Fig. 3-47

17 Model area estimates WMO Fig. 3-47

18 Model area estimates WMO Fig. 3-47

19 Minimum Ozone WMO Fig. 3-47

20 Model Predictions Summary WMO assessment (2004): “These models suggest that the minimum column ozone may have already occurred or should occur within the next decade, and that recovery to 1980 levels may be expected in the 2045 to 2055 period.” CCM losses tend to be too small All of the CCMs underestimate the ozone hole area. In general, the CCMs overestimate the depth of the ozone hole.

21 What controls Antarctic ozone losses?

22 PSCs 1. PSC composition & phase are key to heterogeneous reaction rates II - Crystaline water Ice ~ 188 K Ia - Crystaline particles above frost point ~ 195 K Ib - liquid particles above the frost point ~ 192 K 2. PSCs control de-nitrification and de-hydration, which influences ozone loss 1. PSC composition & phase are key to heterogeneous reaction rates II - Crystaline water Ice ~ 188 K Ia - Crystaline particles above frost point ~ 195 K Ib - liquid particles above the frost point ~ 192 K 2. PSCs control de-nitrification and de-hydration, which influences ozone loss Photo: Paul A. Newman - Jan. 14, Southern Scandanavia

23 Solomon et al. (1986), Wofsy and McElroy (1986), and Crutzen and Arnold (1986) suggest reactions on cloud particle surfaces as mechanism for activating Chlorine HCl ClONO 2 HNO 3 Cl 2 Cl 2 is easily photolyzed by UV & blue/green light HNO 3 is sequestered on PSC Antarctic ozone hole theory

24 Polar Ozone Destruction Only visible light (blue/green) needed for photolyzing ClOOCl No oxygen atoms required Net: 2  O 3 + h  3  O 2 2 O 3 3. ClOOCl+h  2 Cl+O 2 3 O 2 1. O 3 + Cl  ClO + O ClO + M  ClOOCl + M

25 Chlorine and Bromine NOZE 1 & 2 missions in 1986: High- concentrations of chlorine monoxide at low altitudes in the Antarctic spring stratosphere - diurnal-variations, R. Dezafra, M. Jaramillo, A. Parrish, P. Solomon, B. Connor, J. Barrett, Nature, 1987 AAOE mission in August-September 1987: observations inside the polar vortex show high ClO is related to a strong decrease of ozone over the course of the Antarctic spring: J. Anderson et al., JGR, 1989 Latitude (˚S) Ozone (ppmv)

26 Ozone Hole Area Versus Year Polar vortex ≈ 33 Million km 2

27 Ozone Hole Residual Area Vs. T O 3 residual area: 9/21-9/30 T: 9/11 - 9/20, 50 hPa, 55-75ºS If the temperature is 1 K below normal, then ozone hole’s area will be 1.1 Million km 2 larger than normal. See Newman and Nash, GRL, 2004

28 Problem We have reasonably good estimates of temperatures over Antarctica from radiosondes and satellite temperature retrievals We only have snapshots of Cl and Br over Antarctica How can we estimate Cl and Br over Antarctica for all of our observed ozone holes?

29 Chlorine over Antarctica

30 Ozone Loss Source Chemicals Surface concentrations ~ 1998 Cl is much more abundant than Br Br is about 50 times more effective at O 3 destruction From Ozone FAQ - see

31 Atmospheric Chlorine Trends from NOAA/ERL - Climate Monitoring Division Updated Figure made by Dr. James Elkins from Trends of the Commonly Used Halons Below Published by Butler et al. [1998], All CFC-113 from Steve Montzka (flasks by GC/MS), and recent updates of all other gases from Geoff Dutton (in situ GC). 50 years 102 years 5 years 42 years 85 years Steady growth of CFCs up to 1992 CFC-11 CH 3 CCl 3 CCl 4 CFC-113 CFC-12 Production fully banned in US by Pres. Bush

32 CFC-12 released in troposphere Carried into stratosphere in the tropics by slow rising circulation CFC-12 photolyzed in stratosphere by solar UV, releasing Cl Cl catalytically destroys O 3 Cl reacts with CH 4 or NO 2 to form HCl or ClONO 2 HCl and ClONO 2 react on the surfaces of PSCs CFC-12 (CCl 2 F 2 ) pathway to Antarctica Latitude Pressure (hPa) 16 Altitude (km)

33 Mean Age-of-air

34 CCM mean age-of-air (Sept.) GSFC GEOS-4 mean age-of-air derived from advected age tracer. Magenta line is the tropopause, white lines are zonal mean zonal wind Grey lines schematically show mean flow.

35 CCM mean age-of-air (Sept.) Air at a particular point in the stratosphere is a mixture of air parcels that have come together from a multitude of pathways with different times of transit. This “spectrum” of transit time forms an “age-spectrum” that has a mean value and a spectrum “width”

36 Age Spectra The spectrum is convolved with the surface observation time series to yield the stratospheric time series.

37 Fractional Release

38 CCM mean age-of-air (Sept.) CFC-11 Inorganic 0-year CFC-11 Inorganic 1-year CFC-11 Inorganic 2-year CFC-11 Inorganic 3-year CFC-11 Inorganic 4-year CFC-11 Inorganic 5-year

39 CCM mean age-of-air (Sept.) CFC-11 Inorganic 0-year CFC-11 Inorganic 5-year If we know the mean age of air (  ), and we know the fractional release rate as a function of , then we can estimate the chlorine available from CFC-11 for ozone loss

40 CFC-11 break down Schauffler et al. (2003)

41 Estimating chlorine over Antarctica

42 Estimating halogen (Cl & Br) levels over Antarctica Observations show that it takes about 5.5 years for air to get to the Antarctic stratosphere - tropospheric CFCs in January 2000 yield Antarctic stratospheric Cl in July 2005! We use observed CFCs & mean age-of-air estimates to calculate fractional release rates as a fcn. of age EESC = equivalent effective stratospheric chlorine n = # Cl or Br atoms, f = release rate,  = chemical mixing ratio,  = scaling factor to account for Br efficiency for ozone loss

43 EESC Observed total chlorine* (surface) Estimated stratospheric chlorine

44 Parametric model of the ozone hole Method  fit ozone hole size to quadratic functions of EESC and temperature

45 Ozone Hole Parametric Model EESC max = ppbv a 0 = million km 2 a 1 = 50.9 million km 2 /ppbv a 2 = million km 2 /K A = 0 for EESC = ppbv  = residual area r = (r 2 =.943) Area is a function of Effective Equivalent Stratospheric chlorine (EESC) and temperature EESC = 0.8 G(CCl y ) + G(CBr y ) G = Age Spectrum (6 year mean age, 3 year width) CCly and CBry from WMO (2003 )

46 Recovery Predictions

47 Ozone Hole Area vs. Year (1)

48 Ozone Hole Area vs. Year (2) Temperature effect is removed

49 Ozone Hole Area vs. Year (3) Black line represents the fit of area to EESC Area residual  = 1.8 M km 2  (92) Unexplained residual for 1992 ~ 3 m km 2

50 Ozone Hole Area vs. Year (4) Using WMO (2003) Cly and Bry projections, we use our fit to project the ozone hole area

51 Ozone Hole Area vs. Year (5)

52 Add uncertainty to fits EESC: We assume mean age = 5.5 years and the spectrum width = 2.75 years = EESC 0 –Monte Carlo mean age (  = 0.5 years) and width (  = 0.5 years) to generate new EESC time series = EESC 1 –Add 80 pptv of “noise” to EESC 1 = EESC 2 Area: Use original area fit (A 0 ) + added noise re-sampled from area residuals = A 1 Refit new Area (A 1 ) as a function of EESC 2 Project forward using EESC 1 for calculating new recovery dates

53 Ozone Hole Area vs. Year (6) The ozone hole area peaked in 2001 from the area fit to EESC The ozone hole area will remain large (and relatively unchanged for 20 years ( ) Area will start decreasing in approximately 2017 The area will have decreased 1-  by 2018 and 2-  by 2027 Based upon our boot-strap statistics, recovery will first be detected in 2024 The area will be zero in 2070

54 Uncertainties

55 Uncertainties Are the chlorine and bromine levels over Antarctica well represented by using WMO (2003) and an age-spectrum for the period? How good is WMO (2003)? New revisions (A1) increased recovery to 2070 from Is a 5.5 year mean age and a 2.75 width appropriate for the age spectrum? How do we represent interannual variability in age, Cly and Bry estimates? Will climate change impact H 2 O levels and the initial conditions for the ozone hole?

56 Full Recovery vs. mean age-of-air The recovery dates are proportional to our estimate of the mean age-of-air inside the Antarctic vortex: Age sensitivity=9.0 yr/yr Critical to improve our understanding of age in the vortex and to understand age variation in future climate scenarios

57 Climate change effect on ozone hole recovery

58 How will climate change impact the ozone hole? K/decade cooling CMIP2 data from IPCC (2001) Peak size 2011 (2004) Area will start decreasing in approximately 2018 (2017) The area will have decreased 1-  by 2025 (2024) and 2-  by 2031 (2029) Based upon our boot-strap statistics, recovery will first be detected in 2028 (2027) The area will be zero in 2079 (2075) magenta - no T-trend No trend

59 Summary The area of the ozone hole is well represented by T and Cl and Br - We can use this to predict future size and minimum values of Antarctic ozone. Based upon our parametric model: The ozone hole will remain large for a least another decade with no evidence of improvement Actual decreases will begin in about 2017, but can not be detected until 2023 The full recovery will not occur until 2070 GHG change will have small impact on recovery Recovery is strongly dependent on age-of-air and future CFC scenarios Current coupled models are still inadequate for recovery predictions

60 END Jan. 10, local noon, Kiruna, Sweden