Preliminary comparison of OMI and OMPS rotational-Raman cloud optical centroid pressures Alexander Vasilkov1, Joanna Joiner2, and Colin Seftor1 1. Science Systems and Applications, Inc. 2. NASA Goddard Space Flight Center OMI Science Team Meeting 18 March 11-13, 2014 1 Rotational-Raman scattering cloud algorithm basics • Cloud pressures (a.k.a. OCP) are retrieved from the highfrequency structure of TOA radiance caused by RRS. Mixed Lambertian Equivalent Reflectivity (MLER) model Fitting window 345.5-354.5nm. Reflectivity/Effective cloud fraction at 354.1nm Fixed solar irradiance of Day1 to calculate normalized radiances Lookup tables generated with a synthetic high-resolution solar spectrum (OMPS solar data did not work well) – Spline interpolation of TOA normalized radiance over the lookup table wavelength grid. Linear interpolation did not work at all (OMPS less well sampled than OMI) – – – – • Soft calibration is important: Over Antarctic Plateau the scene pressure is assumed be equal to the surface pressure. – Compute spectral residuals (observed - calculated radiances) at each swath position – Use calculated residuals (determined from Dec. 2012) to correct TOA radiances A. Vasilkov, J. Joiner, C. Seftor, “First results from the Suomi National Polar-orbiting Partnership (NPP) Ozone Mapping Profile Spectrometer (OMPS) nadir mapper 2 rotational-Raman scattering cloud algorithm”, submitted to AMTD. OMPS solar flux Expected solar flux based on convolution of high resolution spectrum with OMPS band passes Actual measured OMPS solar flux (Day 1) Dots are for different cross track positions (total 36). Each cross track position has a different set of wavelengths and a different band pass The allowed pre-launch specification in radiance and irradiance calibrations are 8% and 7% respectively. The normalized radiance errors are generally less than 2%, well within instrument specification (Seftor et al. 2014) 3 Rotational-Raman scattering effect on TOA radiances OMPS less sensitive to effects of rotational-Raman scattering (RRS) than OMI owing to its lower spectral resolution Good for SO2 & O3 retrievals, not so good for RRS cloud pressure retrievals Percentage effect of RRS on OMPS radiances at the total ozone algorithm wavelengths Wavelength (nm) 317.62 331.34 360.15 filling-in (%) - 0.29 0.42 - 0.79 4 Soft calibration of TOA radiance Corrections to TOA radiances are within ± 1% which is quite similar to OMI. Depend on wavelength and cross-track position Spectral corrections that are applied to measured TOA OMPS radiances as derived from data over Antarctica. Each curve is for a particular OMPS cross-track position. 5 OMPS Comparison of OMPS and OMI Cloud pressure retrievals of Jan 07, 2013 (ECF>0.05) OMI • Most cloud OCP patterns are same (northern Pacific, Mexico, northern Atlantic, northern China) • OMI OCP retrievals are somewhat lower than OMPS particularly in the tropics 6 PDFs of effective cloud fraction Comparison of PDFs calculated for different latitude & ECF bins. A direct comparison cannot be accomplished because of (1) different spatial resolution, (2) observations not at the same time. PDFs are practically same. Not necessarily expected because of different sizes of the OMI and OMPS footprint. 7 Comparison of PDFs of cloud pressure Southern mid-latitudes Tropics Northern mid-latitudes •In general, OMI retrieves somewhat lower cloud OCPs than OMPS does. Differences most pronounced in the tropics. •Not understood yet. The use of experimental OMPS stray light correction does not improve the agreement. 8 Comparison of PDFs of cloud pressure, OMI O2-O2 added Southern mid-latitudes Tropics Northern mid-latitudes Differences between OMI RRS and OMPS cloud pressures appear to be similar to differences between OMI RRS and OMI O2-O2 except for the differences in the tropics. 9 Total column ozone: OCP climatology vs retrieved OCP Ozone differences: - mostly positive - up to 5% - well correlated with bright clouds Caused by deviation of the actual OCP from the monthly climatology. Reflectivity 10 OCP climatology Comparisons with OMI Retrieved OCP Retrieved OCPs slightly improve OMPS – OMI comparison: the width of the OMPS-OMI difference distribution becomes slightly less 11 Conclusions • OMPS less sensitive to rotational-Raman scattering than OMI owing to lower spectral resolution • OMPS retrieves somewhat higher cloud OCPs than OMI, particularly in the tropics (not understood) • PDFs of ECF retrieved from OMI and OMPS are practically same, good confidence in the OMPS ECF • Monthly gridded OMI OCP climatology captures much of the variability in cloud OCP • The use of actual cloud OCPs improves total column O3 estimates 12 Backup 13 Wavelength shift Wavelength shift as a function of OMPS cross-track position Wavelength shift as a function of OMPS orbit scan position - Typical OMI wavelength shift is much smaller: it is about 0.002 nm. OMPS wavelength shift agrees well with that reported by NOAA. 14 Surface pressure – Retrieved cloud pressure: cross track dependence OMPS OMI Bad cross track position #36 (seen on the OMPS cloud pressure map). 15 Tropical total ozone: Comparison of the use of retrieved OCP (OMI), OCP climatology (NASA OMPS), & CTP (1st 1.5 years IDPS) The use of cloud-top pressures (red curve) leads to unphysical increase of total ozone for bright deep convective clouds Monthly gridded OMI OCP climatology captures much of the variability in cloud OCP 16 Stray light correction Comparison of PDFs of cloud pressures retrieved with and without the stray light correction shows insignificant differences only. 17
© Copyright 2024 ExpyDoc