Optical effects of clouds on trace-gas absorption

1. Optical effects of clouds on trace-gas absorption Joanna Joiner, Alexander Vasilkov, P.K. Bhartia, Robert Spurr, Nick Krotkov, Jerry Ziemke, Sushil Chandra, Pieternel Levelt, Graeme Stephens

2. Parameters that affect atmospheric absorption • Cloud fraction (geometric) • Cloud optical depth • Surface albedo • Cloud vertical structure Radiative cloud fraction Radiative cloud pressure Joiner, Aura Sci Team Meeting

3. Radiative Cloud Fraction • Definition: Fraction of measured radiation that is scattered by clouds. Where Ic and Im are the cloudy and measured (total) radiances, respectively (normalized by solar irradiance) and fg is the geometrical cloud fraction It can be shown approximately that • If cloud is brighter than the surrounding atmosphere then fR>fg. • In a non-absorbing atmosphere, fR increases with l (Im decreases, IC doesn’t change). • Can be computed with simple cloud models Joiner, Aura Sci Team Meeting

4. Simple cloud models • Lambertian-equivalent reflectivity (LER): Surface (cloud/ground) is an opaque and isotropic scatterer with reflectivity R • Mixed LER: Pixel is composed of weighted clear and cloudy components; In OMI algorithms, clouds are assumed to have R=0.8; Proper weighting via fr accounts for light scattered from beneath the cloud. • Plane-Parallel Cloud (PPC): Use Mie scattering theory with a horizontally infinite and vertically homogeneous cloud with an effective optical depth (related to fr). • Mixed PPC: Similar to Mixed LER but uses PPC model for cloud Joiner, Aura Sci Team Meeting

5. How well can we predict Rayleigh scattering with simple cloud models? • Rayleigh scattering can be described well by simple MLER model with 1 parameter (not significantly affected by cloud vertical structure) • Implication: May not need subpixel imager on future instrument From Ahmad et al. JGR (2004) Joiner, Aura Sci Team Meeting

6. How well does Mixed LER (radiative cloud fraction concept) work for trace-gas absorption? LambertianCloud No Cloud Mie Cloud Mixed LER (gives correct column ozone) Joiner, Aura Sci Team Meeting

7. What do real clouds look like to OMI?Cloudsat (A-train) helps us to answer MODIS: sensitive to cloud-top (not appropriate for UV-VIS trace-gas retrievals) OMI radiative cloud pressure from Raman scattering: UV/visible light penetrates deeper OMI simulated from Cloudsat Cloudsat optical depth profiles Cloudsat radar reflectivity thick upper deck Optical depth peaks in liquid water portion of cloud Optical depth peaks in ice portion of cloud 2 cloud decks: thin upper deck- OMI retrieve pressure of lower deck Joiner, Aura Sci Team Meeting

8. O3, O2-O2 cloudy Jacobians: significant photon penetration inside clouds IR cloud-top Joiner, Aura Sci Team Meeting

9. Application: Cloud-slicing From Ziemke et al., in prep. (2007) – see poster Joiner, Aura Sci Team Meeting

10. OMI Cloud-Slicing: Slope Mixing Ratio; Intercept  Stratospheric Column Pacific Africa Provides indirect validation of retrieved cloud-pressures From Ziemke et al., in prep. (2007) Joiner, Aura Sci Team Meeting

11. A-train cloud synergy Passive Sensors Active Sensors (UV/VIS) OMI POLDER MODIS (IR) MODIS AIRS CALIPSO MLS (micro-wave) AMSR AMSU Cloudsat Joiner, Aura Sci Team Meeting

12. Summary • OMI DOAS retrievals use cloud pressures from O2-O2 • OMI TOMS currently uses cloud climatology from thermal IR (e.g. can have significant errors over bright clouds); next version (and TOMS reprocessing) will use OMI-derived climatology • Can estimate radiative cloud fraction from UV/VIS reflectances or radiance ratio, cloud pressures from Raman scattering, O2-O2, or O2 absorption. • Radiative cloud pressure is distinct from (IR) cloud-top; Significant photon penetration inside clouds • Can use these concepts e.g. to retrieve in-cloud mixing ratios • A-train provides unique opportunity to combine cloud information to derive information about cloud vertical extent from different passive sensors and validate with active sensors. Joiner, Aura Sci Team Meeting