Scenarios: Target Mode Observations of fixed ground target

1. O2 A-band CO2 2.06 m CO2 1.61m Polarization Effects on Column CO2 Retrievals from Non-Nadir Satellite Measurements in the Short-Wave Infrared Vijay Natraj1, Hartmut Bösch2, Robert J.D. Spurr3, Yuk L. Yung1 1Department of Planetary Sciences, California Institute of Technology, Pasadena, CA, USA 2Department of Physics and Astronomy, University of Leicester, Leicester, UK 3RT Solutions, Inc., Cambridge, MA, USA Contact: Vijay Natraj, Phone: +1-626-395-6962, Email: vijay@gps.caltech.edu A51A-0091 Introduction • Greenhouse Gases Observation Satellite (GOSAT): successful • Orbiting Carbon Observatory (OCO): to be rebuilt (?) • Quantify sources and sinks of CO2 using precise column abundance measurements • Reflected sunlight at the top of the atmosphere (TOA) • Spectrometers sensitive to atmospheric polarization. • Need to consider polarization in modeling of atmospheric radiative transfer (RT). XCO2 Errors: Target Mode • Scenarios: Target Mode • Observations of fixed ground target • Validations of space-based observations with coincident ground-based column measurements • Aerosol/Cirrus: Same as for glint • Scattering angle: 85°-150° XCO2 Errors: Glint Mode Figure 6: XCO2 Errors for (left) scalar model (right) R-2OS model • Conclusions • Scalar • Largest errors for low AOD/COD • Low order scattering is highly polarized • 2OS • 1-2 orders of magnitude smaller XCO2 errors than scalar model • Thin cirrus modeled well • Largest errors for optically thick aerosol scenarios • Polarized multiple scattering causes large errors • XCO2 errors • Interplay between CO2 Jacobians and forward model errors • Forward model error compensated by changing CO2 • CO2 Jacobians larger for aerosol than cirrus • Aerosol extinction decreases significantly with wavelength • Cirrus extinction remains more or less constant with wavelength • Forward model error larger for aerosol than cirrus • Aerosol particles small and scatter less => low order (> 2) scattering • Cirrus larger and scatters more => multiple scattering • Error compensation when all variables retrieved simultaneously Figure 1: OCO Spectral Regions • R-2OS Model • Fast polarization correction algorithm • Assume that only two scattering events (two orders of scattering, 2OS) contribute to polarization • 2OS model used in conjunction with scalar RT model Radiant (R) to simulate OCO backscatter measurements • Isca, Icor: intensity with polarization neglected, scalar-vector intensity correction • I, Q, U: Stokes parameters Figure 2: XCO2 Errors for (left) scalar model (right) R-2OS model Figure 4: (top) Forward Model Errors (bottom) CO2 Jacobians Black: AOD = 0.3, Red: AOD = 0.3 (high altitude); Blue: COD = 0.3 SZA = 40° Figure 3: XCO2 Errors for CO2-only retrievals References [1] D. Crisp, et al., Adv. Space Res., 34(4), 700-709, 2004. [2] V. Natraj and R.J.D. Spurr, J. Quant. Spectrosc. Radiat. Transfer, 107(2), 263-293, 2007. [3] V. Natraj, et al., J. Geophys. Res., 113, D11212, 2008. [4] S. Chandrasekhar, Radiative Transfer, 1960. [5] C. D. Rodgers, Inverse Methods for Atmospheric Sounding, 2000. State Vector: 1-19: CO2 20: H2O scaling 21: Surface pressure 22-40: Temperature 41-59: Aerosol 60: Wind speed • Scenarios: Glint Mode • Sunglint over ocean • High signal to noise ratio (SNR) over ocean • SZA: 20°, 30°, 40°, 50°, 60°, 65°, 70°, 75° • Aerosol/Cirrus: AOD = 0.05, 0.3, 0.3 (high altitude); COD = 0.05, 0.3, 0.3 (low altitude); AOD = 0.05, COD = 0.25; AOD = 0.25, COD = 0.05 Surface pressure and wind speed are correlated Figure 5: Correlation Coefficients (SZA = 40°)