Influence of solar spectral
variability on middle atmospheric
ozone
Aimee Merkel, Jerald Harder, Daniel Marsh,
Anne Smith, Thomas Woods
Laboratory for Atmospheric and Space Physics (LASP)
University of Colorado
National Center for Atmospheric Research (NCAR ACD )
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 1
Outline
• Study of solar cycle variations in middle atmospheric ozone.
(lower mesosphere, upper stratosphere)
• MUV (200-300nm) influences photochemistry in this altitude region.
• Ozone observations – SABER, MLS, SME
• Can we explain observations with a modelling study using NCAR’s
WACCM (Whole Atmosphere Community Climate Model).
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 2
SSI Solar Forcing and Earth Atmospheric Response
Many recent studies on atmospheric response to solar cycle 23-24
Haigh et al. 2010 – IC2D model
Cahalan et al. 2010 –GISS ModelE
Merkel et al. 2011 – WACCM
Ineson et al. 2011 – HadGEM3
Dhomse et al. 2011 – 3D chemical model
Beig et al. 2012 - HAMMONIA
Oberlander et al. 2012 –EMAC-FUB
Swartz et al. 2012 – GEOSCCM
Wang et al. 2012- JPL MLS OH
Study focus - SSI implications on:
- Photochemistry
- Radiative response
- Circulation
- “Top down” vs “Bottom up”
Our focus and contribution to
this body of work
aaa
Middle Atmospheric Ozone
response to SSI
Top down = direct impact from the UV through the stratosphere
Bottom up = visible-IR feedback from ocean uptake
and sea surface temperatures
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 3
Our Focus: Ozone Photochemistry
Herzberg
Continuum
Hartley
bands
242 nm
• Ozone concentration in the upper and middle atmosphere is influenced by
the solar radiation in the 200-300nm band. (production and loss)
- Herzberg Continuum (200-242nm) – J2 - O2 Photolysis – O3 production
- Hartley Bands (200-300nm)– J3 - O3 Photolysis – O3 loss
• Any changes in the UV irradiance will modify the photo-dissociation rates and
influence ozone concentrations.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 4
Mesospheric ozone observed over SC 23-24
Possible Datasets: Criteria for
lower mesospheric ozone studies
AURA/MLS
(Microwave Limb Sounder)
 Limb profiles up through the mesosphere
 Cover the solar cycle
in question
TIMED/SABER

Consistent
sampling
(Sounding
of the
Atmosphere using Broadband Emission
 Global coverage Radiometry)
 Local time sampling
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 5
Aura MLS Ozone: Haigh et al. 2010
Approx. Altitude (km)
Pressure Level (hPa)
MLS Ozone
Annual Mean VMR
Max-Min % difference
Max= Avg(2008 & 2009)
Min = Avg(2005 & 2006)
• Conclusions from MLS analysis
• Mesospheric ozone change of -1.7% (out-of-phase with solar cycle)
• Stratospheric ozone change of +4.5% (in-phase with solar cycle)
• Criticism: not enough data to determine solar cycle (Garcia 2010)
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 6
TIMED/SABER Ozone
Measurements






O3 emission measurements from the 9.6μm channel
Limb profiles, both stratosphere and mesosphere sampled
Spans ~11 years (2002-2012)
All local times observed – so can separate into day and night
Very good consistent sampling
Global coverage
Data validated by Rong et al. 2009 and recently by A. Smith 2012.
Known systematic bias compared to other ozone measurements.
Bias is constant over time, so does not influence differences.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 7
SABER Analysis – Annual Zonal mean trends
Published Merkel et al. 2011
Annual zonal mean trends
Day
Lower Mesosphere
 Out-of-phase with solar cycle
 Trend in day, absent at night
 photolysis ceases at night
Sept. 19, 2012 SORCE Science Meeting
Night
Stratosphere
 In-phase with solar cycle
 Trend similar for day and night
 Less driven by photochemistry
 Very little diurnal variation
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 8
SABER Analysis – Annual Zonal mean trends
Annual zonal mean trends
Day
Lower Mesosphere
 Out-of-phase with solar cycle
 Trend in day, absent at night
 photolysis ceases at night
Sept. 19, 2012 SORCE Science Meeting
Night
Stratosphere
 In-phase with solar cycle
 Trend similar for day and night
 Less driven by photochemistry
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 9
SABER Analysis – Annual Zonal mean trends
Annual zonal mean trends
Day
Night
Monthly zonal mean trends
Regression analysis with 95% confidence.
Day
Lower Mesosphere
Solar Trend = +2.8% ± 1.4%
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Stratosphere
Solar Trend = -2.6% ± 0.63%
Merkel - 10
Equatorial Mean: % Difference from 2008/2009
Equatorial Mean
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 11
Equatorial Mean: % Difference from 2008/2009
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 12
SABER Compared to HALOE
HALOE
Halogen Occultation Experiment
(Remsberg et al., 2008)
SABER
1991-2005
O3 % (Max-min)
➨SABER stratospheric ozone is consistent with previous measurements.
➨SABER able to resolve lower mesospheric ozone response to solar forcing.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 13
Ozone measurements in previous solar cycles
Question: Why have we not seen this mesospheric ozone behavior in
previous solar cycles?
Reason
Historical ozone data comprised
of occultation data.
The solar signal is “washed out”
due to the mixing of source and
loss mechanisms of ozone due to
local time of observations.
(sunrise and sunset)
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 14
Analyze SABER as if Occultation Experiment
Day
Night
Partial Day-Night
Annual Mean Difference
25S-25N
-
When SABER is analyzed with only measurements taken at “occultation” local times:
Solar signal is washed out and response is more similar to night results
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 15
Analyze WACCM at Occultation local times
Modeled ozone results
Day
Night
Partial Day-Night
When WACCM is analyzed with only measurements taken at “occultation” local times:
Solar signal is washed out and response is more similar to night results.
Confirms results from SABER.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 16
Conclusions so far
• So far we have only observed mesospheric out-of-phase SC response in
ozone in the current solar cycle.
• The response has been observed in two independent measurements from
MLS and SABER.
• The SC response in mesospheric ozone is very dependent on local time.
Only a trend in daytime measurements.
• Not observed in previous solar cycles. Occultation datasets were used to
determine SC response of mesospheric ozone. Signal shows no trend like
nighttime data.
• Can we find the signal in daytime ozone data from another solar cycle?
- Solar Mesosphere Explorer (SME)
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 17
Solar Mesosphere Explorer
• Daytime (3pm) limb profiles of ozone
with good global coverage.
• Two channels
- UV ozone Spectrometer (UV ozone)
- 1.27μm Spectrometer (airglow ozone)
• Descending phase of Solar cycle 21
- Good ozone measurements
between 1982-1986
• SME measurements can be analyzed
consistent with SABER analysis.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 18
Solar Mesosphere Explorer compared to SABER
Max – Min Percent Difference
SABER 9.6 channel (2003-2009)
SME Airglow Ozone (1982 – 1986)
SME UV Ozone (1982 – 1986)
• Very good comparison with SABER.
• First time the out-of-phase
mesospheric solar cycle signal in
ozone is observed in a different solar
cycle.
• Could imply that there is more UV
variability in SC 21.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 19
Modeling Study
• Modelling study using NCAR’s WACCM (Whole Atmosphere Community
Climate Model).
• Response of the photochemistry in the middle atmosphere when
forced by different distributions of solar spectral irradiance (SSI).
• Compare modelled ozone response to observations. Can SSI variability
explain ozone observations in the mesosphere?
•
Model sensitivity studies are important to help to constrain the uncertainties
in the solar measurements. Ie. how good does the solar measurement
need to be to adequately model this part of the atmosphere?
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 20
Experimental Setup
WACCM Model Simulations
WACCM is a fully coupled chemistry, radiation and dynamics global model
extending from the surface to the thermosphere
Approach: Constrain model so atmospheric response is from solar forcing.
•
•
•
•
•
•
Green house gases held constant – Model spin up
Time slice experiment – not transient simulation
TSI is conserved between active cases
Perpetual year simulations per case study
Scaled NRLSSI spectra to SORCE variability
Climatological SST, “Top Down” Focus
Case studies:
NRLSSI
SORCE 1 – SOLSTICE < 242 nm
SIM > 242 nm
SORCE 2 – SIM > 210 nm
2.5% increase in intensity in 200-242nm band
between SORCE 1 and SORCE 2.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 21
Ozone response Model Simulations
Active 2004 – Quiet 2007 (% Difference)
NRLSSI
SORCE 1
SORCE 2
O3 Difference (%)
O3 Difference (%)
O3 Difference (%)
Latitude
SORCE 1
SOLSTICE < 242 nm
SIM > 242 nm
Sept. 19, 2012 SORCE Science Meeting
25S – 25N
O3 Difference (%)
Latitude
SORCE 2
SIM > 210 nm
2.5% more intensity in
Herzberg Continuum
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 22
Annual Mean Differences
(25N-25S)
Mesosphere (less O3 at solar active) Why?
Ozone is depleted through several catalytic processes with OH and H
=
More UV
More O3 photolysis
J3 rate O3=O(1D) + O2
=
More O(1D)
O(1D)
=
H2O
More OH
=
OH
Less O3
O3
SORCE 1 – SOLSTICE < 242 nm
SIM > 242 nm
SORCE 2 – SIM > 210 nm
2.5% increase in intensity
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 23
Annual Mean Differences
Stratosphere -more O3 at solar active
J2 rate o =2o photolysis
2
(25N-25S)
J3 rate O =O( D) + O
3
1
More UV radiation is transmitted to lower levels
• Greater O2 photolysis and thus more O3
• 2.5% increase in intensity in 210-242nm band
• Increased J2 rate from 4.5% to 7%
• Increased J3 rate from 3% to 3.5%
• Altitude of sign change seems to depend on both J2 and J3 rate
Sept. 19, 2012 SORCE Science Meeting
O3
2
SORCE 1 – SOLSTICE < 242 nm
SIM > 242 nm
SORCE 2 – SIM > 210 nm
2.5% increase in intensity
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 24
Model Ozone Compared to SABER
Annual Mean Difference 25S-25N
Upper Mesosphere:
Both SORCE model simulations
reproduce the observations
Lower Mesosphere
More modeling work is needed
Stratosphere
More UV between 200-240nm
helps reproduce the O3
In-phase solar response
SORCE 1 – SOLSTICE < 242 nm
SIM > 242 nm
SORCE 2 – SIM > 210 nm
2.5% increase in intensity
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 25
Conclusions
 Mesospheric out-of-phase SC response in ozone observed in two independent
measurements from MLS and SABER.
 The SC response in mesospheric ozone is dependent on local time.
Solar cycle variability only in daytime measurements.
 First time out-of-phase response observed in a different SC.
Proof that occultation measurements should not be used for solar cycle analysis
in the mesosphere.
 Very important to measure the solar cycle variability of the SSI. Has implications
on the photochemistry in Earth’s atmosphere.
 An increase in UV variability as observed by SORCE (both SIM and SOLSTICE) help
to resolve differences between modeled ozone and observations in the
mesosphere.
 Important to continue modeling studies with observed solar data. Implications
are significant and evidence exists that SORCE data helps reproduce
observational data.
Sept. 19, 2012 SORCE Science Meeting
Influence of Solar Spectral Variability on Middle Atmospheric Ozone
Merkel - 26
“Corona Arch”
Moab, UT
Thank You