Xiong Liu ([email protected]) Harvard-Smithsonian Center for Astrophysics Collaborators: Kelly Chance, Christopher Sioris, Robert Spurr, Thomas Kurosu (CFA) Randall Martin (Dalhousie Univ., Canada) Mike Newchurch (University of Alabama in Huntsville) May Fu, Jennifer Logan, Daniel Jacob, Paul Palmer (Harvard University) PK Bhartia (Atmospheric Chemistry and Dynamics Branch, NASA GSFC ) Rob Chatfield (Atmospheric Chemistry and Dynamics Branch, NASA AMES ) First Directly-Retrieved Global Distribution of Tropospheric Column Ozone from GOME
38
Embed
Xiong Liu ( [email protected]) [email protected] Harvard-Smithsonian Center for Astrophysics Collaborators: Kelly Chance, Christopher Sioris, Robert.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Introduction to Ozone and Tropospheric Ozone Satellite-based Tropospheric Ozone Retrievals Algorithm Description Intercomparison with TOMS, Dobson, and
Ozonesonde observations Examples of Daily Retrievals Global distribution of Tropospheric Column Ozone
and comparison with the 3D GEOS-CHEM model Summary and Future work Atmospheric Measurements and Studies at Harvard-
Smithsonian CFA (Time allowed)
3
Ozone
Stratopause
Ozonelayer
Tropopause
Courtesy of Randall Martin
First discovered by Schönbein [1840], a reactive oxidant in the atmosphere.
First quantitative observation: early 20th century in Europe (e.g., Dobson, Götz)
First discovery of ozone hole by Farman et al. [1985]
First attempts to understand ozone in the 1930s. Until the last 30-40 years, the stratospheric ozone chemistry (NOx, HOx, ClOx) is well understood.
Noble chemistry prizes were awarded to Paul Crutzen, Mario Molina, and Sherwood Rowland in 1995.
4
Tropospheric Ozone
CO, VOCs, NOx
HO2OH
NO2
H2O2
O3
hv, H2O
HNO3
OH
CO, VOCs
NOx
HOx
Simplified Tropospheric O3 Chemistry
Courtesy of Randall Martin
Key species in climate, air quality, and tropospheric chemistry Major Greenhouse gas, 15-20% of climate radiative forcing Primary constituents of photochemical smog Largely controls tropospheric oxidizing capacity
NO
5
Residual-based Satellite Tropospheric Ozone Retrievals Why satellite observations: global coverage
Challenge: only 10% of Total column Ozone (TO)
Residual-based approaches: TO – Strat. Column Ozone (SCO)
Tropospheric ozone residual: TOMS minus SAGE/SBUV/HALOE/MLS [Fishman et al., 1990, 2003, Ziemke et al., 1998, Chandra et al., 2003]
Cloud/clear difference techniques: [Ziemke et al., 1998; Newchurch et al., 2003, Valks et al., 2003]
Modified residual method [Kim et al., 1996; Hudson et al., 1998]
Topographic contrast method [Jiang et al., 1996, Kim and Newchurch et al., 1996, Newchurch et al., 2003]
Scan-angle method (Special) [Kim et al., 2001]
Limits: poor spatiotemporal resolution and large spatiotemporal variability of SCO mostly climatological or tropics
Limb/Nadir matching: TO/SCO from same instrument or satellite
Atmospheric trace gas absorptions detected in satellite spectra
8
Physical Principles of Ozone Profile Retrieval (UV/Vis.)
Wavelength-dependent O3 absorption & dependence of Rayleigh scattering provide discrimination of O3 at different altitudes from backscattered measurements.
Temperature-dependent ozone absorption in the Huggins bands provides additional tropospheric ozone information.
Direct tropospheric ozone retrieval: daily global distribution of tropospheric ozone without other SCO measurements or deriving SCO
Several groups [Munro et al., 1998; Hoogen et al., 1999; Hasekamp et al., 2001; van der A et al, 2002; Muller et al., 2003] have developed ozone profile retrieval algorithms from GOME: each of them demonstrates that limited tropospheric ozone information can be detected.
However, global distribution of tropospheric column ozone has not been published from these algorithms Require accurate and consistent calibrations. Need to fit the Huggins bands to high precision. Tropospheric column ozone is only ~10% of total column ozone Limited Vertical Resolution
K : Weighting function matrix, sensitivity of radiances to ozone
Sa: A priori covariance matrix
Sy: Measurement error covariance matrix
2 2
2
2 2
{ ( - ) -[ ( )]} ( - ) a
1 1- -2 2
y i i+1 i i i+1 aS K X X Y - R X S X X
( ) { [ - ( )] - ( )} a a
T -1 -1 -1 T -1 -1i+1 i i y i i y i i aX X K S K S K S Y R X S X - X
11
Algorithm Description — Radiative Transfer Simulation
RcRo
IB,o IB,c
Pc
Rs
d
Radiative transfer model: LIDORT [Spurr et al., 2001] Model Ring effect with a first-order single-scattering model Radiance polarization correction with a look-up table
Forward model inputs SAGE strat. [Bauman et al., 2003] & GOCART trop. aerosols [Chin et al., 2002]Daily ECMWF T profiles and NCEP PsClouds: Lambertian surfacesCloud-top pressure from GOMECAT [Kurosu et al., 1999] Cloud fraction derived at 370.2 nm with surface albedo database [Kolemeijer et al.,2003]Wavelength dependent albedo (2-order polynomial) from 326-339 nm to take account of residual aerosol and cloud effects
12
Algorithm Description Perform external wavelength and radiometric calibrations
Derive variable slit widths and shifts between radiances/irradiances Co-add adjacent pixels from 289-307 nm to reduce noise Perform undersampling correction with a high-resolution solar reference Fit degradation for 289-307 nm on line in the retrieval
Optimize fitting windows: 289-307 nm, 326-339 nm
Latitude/monthly dependent TOMS V8 climatology
Retrieval Grid: 11 layers, use daily NCEP tropopause to divide the troposphere and stratosphere, 2-3 tropospheric layers
Tropospheric column ozone: sums of tropospheric partial columns
A Priori influence in TCO: 15% in the tropics, 50% at high-latitudes
15
Retrieval Errors
Precision: 2-8% (< 2DU) in the strat., <12%(5DU) in the troposphereSmoothing: 10% at 20-40 km, 15% at > 40 km, and 30% at <10 km
TO: <2 DU(0.5); 3 DU (1.0%)
SCO: <2 DU(1%); 2-5 DU (1-2%)
TCO: 1.5-3 DU(6-12%); 3-6 DU(12-25%)
16
An Orbit of Retrieved Ozone Profiles
Ozone Hole (120 DU)Biomass burning
17
Validation GOME data are collocated at 33 WOUDC ozonesonde stations during 96-99. Validate retrievals against TOMS V8, Dobson/Brewer total ozone, and ozonesonde TCO. Data mostly from WOUDC Collocation criteria:
Within ~8 hours, 1.5° latitude and ~500 km in longitude Average all TOMS points within GOME footprint
Number of comparisons: 4711, 1871, and 1989 with TOMS, Dobson, and ozonesonde, respectively.
Large positive bias of 5-15 DU at some northern tropical and subtropical regions: central America, tropical North Africa, Southeast Asia, Middle East
Usually >0.6.
Poor correlation: central America, equatorial remote Pacific, tropical North Africa and Atlantic, North high latitudes
26
GOME/GEOS-CHEM vs. MOZAIC (Central America)
20
30
40
50
MOZAIC: www.aero.obs-mip.fr/mozaicData: 1994-2004 (vary from location to location)Evaluate GOME/GEOS-CHEM TCO in seasonality
27
GOME/GEOS-CHEM vs. MOZAIC (Southeast Asia)
A Priori Retrieval GOES-CHEM MOZAIC
28
GOME/GEOS-CHEM vs. MOZAIC (Accra)
A Priori Retrieval GOES-CHEM MOZAIC
29
GOME/GEOS-CHEM vs. MOZAIC (Middle East)
A Priori Retrieval GOES-CHEM MOZAIC
30
Summary Ozone profiles and Tropospheric Column Ozone (TCO) are retrieved
from GOME using the optimal estimation approach. Retrieved TO and TCO compare very well with TOMS,
Dobson/Brewer, and ozonesonde measurements. The retrievals clearly show signals due to convection, biomass
burning, stratospheric influence, pollution, and transport, and are capable of capturing the spatiotemporal evolution of TCO in response to regional or short time-scale events.
The overall structures between GOME and GEOS-CHEM are similar, but some significant positive biases occur at some northern tropical and subtropical regions.
The GOME retrievals usually agree well with the MOZOAIC measurements, to within the monthly variability and some biases can be explained by the reduced sensitivity to lower tropospheric ozone, spatiotemporal variation, and the large spatial resolution of GOME retrievals.
31
Future Work
Improve retrieval algorithms (Chappuis bands, external degradation correction) and complete more than 8-year GOME data record.
Apply the algorithm to SCIMACHY, OMI, GOME-2, OMPS, or future geostationary satellite measurements.
Integrate with the GEOS-CHEM model and other in-situ data, improve our understanding of global/regional budget of tropospheric ozone
Tropospheric ozone radiative forcing
Acknowledgements This study is supported by the NASA and by the Smithsonian Institution. Thank all collaborators. We thank WOUDC and its data providers (e.g., SHADOZ, CMDL), TOMS, MOZAIC for providing correlative measurements. We are grateful to NCEP/NCAR ECMWF reanalysis projects. We appreciate the ongoing cooperation of the European Space Agency and the Germany Aerospace Center in the GOME program.
32
Variable Slit Widths and Shifts
33
Fitting Residuals
34
Aerosol Effects
35
Comparison with SAGE-II (>15 km)
Comparison with SAGE-II V6.2 ozone profiles above 15 km during 1996-1997 (5732)
Means Biases: <15%
1: <10% for the top seven layers and <15% for the bottom layer
Systematic biases in the retrievals but not in the a priori, suggesting residual measurement errors in the GOME level-1 data