Aerosol Optical Depth Measurements by Airborne Sun Photometer in SOLVE II: Comparisons to SAGE III, POAM III and Airborne Spectrometer Measurements* P. Russell1, J. Livingston2, B. Schmid3, J. Eilers1, R. Kolyer1, J. Redemann3, S. Ramirez3, J-H. Yee4, W. Swartz4, R. Shetter5, C. Trepte6, A. Risley, Jr.7, B. Wenny7, J. Zawodny6, W. Chu6, M. Pitts6, J. Lumpe8, M. Fromm8, C. Randall9, R. Bevilacqua10 1NASA Ames Research Center, Moffett Field, CA 2SRI International, Menlo Park, CA 3Bay Area Environmental Research Institute, Sonoma, CA 4Applied Physics Laboratory, Johns Hopkins University, Laurel, MD 5National Center for Atmospheric Research, Boulder, CO 6NASA Langley Research Center, Hampton, VA 7SAIC, NASA Langley Research Center, Hampton, VA 8Computational Physics, Inc., Springfield, VA 9Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO 10Naval Research Laboratory, Washington, DC SAGE: AATS: SZA=90o SAGE: AATS: SZA=90o SAGE: AATS: SZA=90o DC-8 flight altitudes Pixel Group: 21 SAGE l: 449.9 nm AATS l: 452.6 nm Pixel Group: 22 SAGE l: 520.3 nm AATS l: 519.4 nm Pixel Group: 35 SAGE l: 675.6 nm AATS l: 675.1 nm Pixel Group: 51 SAGE l: 869.3 nm AATS l: 864.5 nm Pixel Group: 81 SAGE l: 1019.3 nm AATS l: 1019.1 nm Altitude 449.9 nm [km] , SZAmeas , SZAmeas SAGE: AATS: SZAmeas SZA=90o , SZAmeas , SZAmeas 4. Converting AATS Transmissions to 90o Refracted SZA 1. Introduction The 14-channel NASA Ames Airborne Tracking Sunphotometer (AATS-14) measured solar-beam transmission on the NASA DC-8 during the Second SAGE III Ozone Loss and Validation Experiment (SOLVE II). The companion poster by Livingston et al. describes the AATS-14 instrument, measurement procedures, and results for ozone. This poster presents AATS-14 results for multiwavelength aerosol optical depth (AOD), including its spatial structure and its relationship to results from two satellite sensors and another DC-8 instrument. These are the Stratospheric Aerosol and Gas Experiment III (SAGE III), the Polar Ozone and Aerosol Measurement III (POAM III) and the Direct beam Irradiance Airborne Spectrometer (DIAS). 5. AATS-SAGE Comparisons For 5 AATS-SAGE near-coincidences--on 19, 24, 29, and 31 Jan, and the SAGE III spectral survey case of 2 Feb 2003--the method shown in Section 4 was used to convert AATS transmission T and LOSOT from SZAmeas to 90o (refracted), for components and totals. Results for T , LOSOT, and vertical OD were then compared to SAGE III results (Figs. J, K, L, M). Converting transmission from SZAmeas to 90o needs to be done by components, since different components have different vertical profile shapes, and hence different dependences of airmass on SZA. SAGE III Spectral Survey Case AATS-14 provides aerosol results at 12 wavelengths l between 354 and 1556 nm, spanning the full range of SAGE III and POAM III aerosol wavelengths. Because most AATS measurements in SOLVE II were made at solar zenith angles (SZA) near 90o, retrieved AODs are strongly affected by uncertainties in the relative optical airmass (here called airmass for brevity) of the aerosols and other constituents along the (generally refracted) line of sight (LOS) between instrument and sun. For any given constituent and/or wavelength, airmass is defined as the ratio of LOS optical thickness (OT) to vertical OT (i.e., OD). For SZA near 90o, airmass is sensitive to the vertical profile of the associated attenuator. Uncertainties in such vertical profiles therefore produce corresponding uncertainties in the associated airmass. To reduce dependence of the AATS-satellite comparisons on airmass, we perform the comparisons in LOS transmission and LOS OT as well as in vertical OT (or optical depth). 2. AATS-14 and SAGE III Transmission Measurements Fig. E. Conditions and results for the DC-8 flight of 19 January 2003. Top left: DC-8 track (black line) on map of forecast MPV. Top right: Aerosol extinction profiles retrieved from SAGE III occultation at UT, lat, and lon shown. Middle: DC-8 altitude, distance from SAGE III 10-km tangent point, and true and apparent (refracted) SZA, with AOD retrieved from AATS-14 at two wavelengths. Red vertical dotted line marked by red arrow and “SAGE” shows time of SAGE III tangent at 10 km altitude. Bottom: Airmasses for Rayleigh, ozone, and aerosol (755 nm), at SZAmeas and apparent SZA=90o, computed from the SAGE III profiles at top right using the Yee algorithm Fig. J. Comparison of SAGE and AATS transmission, LOSOT, and OD, by components and totals, for 19 January 2003 at 10.43 km. All AATS results for T and LOSOT have been converted from SZAmeas to apparent SZA 90o. Fig. I. AATS transmission, LOSOT, and OD, for components and totals, for 19 January 2003 at 10.43 km. Top: Transmission at SZAmeas and converted to 90o. Middle: LOSOT at SZAmeas and converted to 90o. Bottom: OD. Shown for comparison are SAGE total transmission, LOSOT, and OD, plus DIAS LOS AOT at 400 nm and SZAmeas . Fig. L. Comparison of SAGE and AATS transmission, LOSOT, and OD, by components and totals, for the 2 February 2003 SAGE spectral survey case, at 9.7 km. The AATS period is 12.43-12.80 UT. All AATS results for T and LOSOT have been converted from SZAmeas to apparent SZA 90o. 6. AATS-DIAS Comparisons Fig. MM. DIAS LOS aerosol optical thickness (AOT) at SZAmeas compared to AATS LOS AOT at SZAmeas.for 19, 24, 29, and 31 January 2003. Also shown are AATS results for other components and total LOSOT, both at SZAmeas and converted to 90o, along with SAGE total LOSOT. Fig. K. Comparison of AATS and SAGE LOSOT (total and components) for 19, 24, 29, and 31 January 2003. AATS values have been converted to apparent SZA= 90o, using the procedure shown in Sect. 4 and illustrated in Fig. I. Fig. M. As in Fig. K, but for vertical OD. Fig. A. SAGE III vertical profiles of transmission at representative wavelengths for the occultation events near the DC-8 on 19, 24, 29, and 31 January, with near-coincident AATS transmission values at the AATS measurement SZA, SZAmeas, and converted to apparent SZA=90o. Fig. F. As in Fig. E, but for 24 January 2003. 7. AATS-POAM Comparisons 8. Summary and Conclusions 3. Airmass dependence on SZA, extinction profiles, wavelength, and DC-8 altitude • AATS-14 measurements on the DC-8 in SOLVE II provide AOD spectra for wavelengths 384-1550 nm, covering the full SAGE III and POAM III wavelength range. The AATS-14 results show AOD spatial structure along the DC-8 flight path and provide AOD spectra for comparison to SAGE III and POAM III at their tangent points. • A new airmass algorithm [Yee, 2003; Magistre and Yee, 2002] validates the Thomason et al.  algorithm to within 2% for SZA<90o, and in addition provides results for SZA>=90o. • The five AATS-SAGE comparisons shown in Figs. A and I-M have several features in common: Fig. N. Conditions and results for the DC-8 flight of 19 January 2003. Top left: DC-8 track (black line) on map of forecast MPV. Top right: Aerosol extinction profiles retrieved from POAM III occultation at UT, lat, and lon shown. Middle: DC-8 altitude, distance from POAM III 10-km tangent point, and true and apparent (refracted) SZA, with AOD retrieved from AATS-14 at two wavelengths. Red vertical dotted line marked by red arrow and “POAM” shows time of POAM III tangent at 10 km altitude. Bottom: Airmasses for Rayleigh, ozone, and aerosol (755 nm), at SZAmeas and apparent SZA=90o, computed from the POAM III profiles at top right using the Yee algorithm. v4 AATS POAM III v4 • AOD differences were ≤0.0036 for all l>400 nm. RMS differences were ≤0.0023. Mean differences (AATS-SAGE) were ≤0.0021. • RMS percentage differences in AOD ([AATS-SAGE]/AATS) were ≤31% for l <~755 nm and 56% for 1020 nm. • SAGE and AATS total LOSOT values (both at apparent SZA=90o) agree to within 10% of the AATS value for l<~755 nm but differ by as much as 41% of the AATS value at 1020 nm and more at 1545 nm. Mean and rms differences, (SAGE-AATS)/AATS, are <8% for l <~755 nm but increase to 36% at 1020 nm, and larger at 1545 nm. • - SAGE and AATS LOS total transmissions T (both at apparent SZA=90o) agree to within 22% at all wavelengths l, 450-1545 nm. Mean and root-mean-square (rms) differences, (AATS-SAGE)/SAGE, are ≤15% at each l. Fig. O. Comparison of AATS-14 and POAM III AOD spectra for the 19 January 2003 occultation shown in Fig. N. Fig. B. Left frames: Representative vertical profiles of aerosol extinction and of ozone and molecular number density. Aerosol and ozone profiles are from SAGE III and POAM III retrievals near DC-8 flights on the dates shown. Aerosol profiles are for one SAGE III wavelength (775.4 nm) and one POAM III wavelength (779.4 nm). The molecular density profile is the MODTRAN subarctic winter model. Right frame: Corresponding results for airmass calculated for altitude 10 km using the method of DeMajistre and Yee (2002) and the constituent profiles in the left frames v3 (within 6% of v4) AATS: o oo POAM III v4: D DD Fig. G. As in Fig. E, but for 24 January 2003. Fig. D. Absolute and percent differences between the airmass results from the methods of DeMajistre and Yee (2002, labeled Yee) and Thomason et al. (1983, labeled UA) shown in Fig. 3 • The case with largest AATS-SAGE difference at short l, 24 January, is also the case with largest scan-to-scan differences in SAGE short-l transmission (see Fig. A). • *The case with best AATS-SAGE agreement is the SAGE spectral survey of 2 Feb 2003 (Fig. L). What causes this? • AATS-POAM AOD differences were ≤0.004 for all l>400 nm. RMS differences were ≤0.0026. Mean differences (AATS-POAM) were ≤0.0015. • * RMS percentage differences in AOD ([AATS-POAM]/AATS) were ≤31% for all l between 450 and 1020 nm. • * AATS-DIAS differences in LOS AOT at 400 nm and SZAmeas were ≤11%. AATS & DIAS Rayleigh LOS Transmission and OT tracked very well at all l, 353-778 nm.See “Late-Breaking AATS-DIAS Comparisons” • Four tests for frost on the AATS-14 window were all negative for all cases shown here (see Appendix of ms*). Frost deposition was prevented by a window heater and extensive purging with dry nitrogen. v4 v4 v4 Fig. Q. Comparison of AATS-14 and POAM III AOD spectra for the 21 January 2003 occultations shown in Fig. P. References DeMajistre, R., and J.-H. Yee, Atmospheric remote sensing using a combined extinctive and refractive stellar occultation technique: 2. Inversion method for extinction measurements, J. Geophys. Res., 107, 10.102/2001JD000795, 2002. Thomason, L. W., B. M. Herman, and J. A. Reagan, The effect of atmospheric attenuators with structured vertical distributions on air mass determinations and Langley plot analyses, J. Atmos. Sci., 40, 1851-1854, 1983. Yee, J.-H., et al., Atmospheric remote sensing using a combined extinctive and refractive stellar occultation technique: 1. Overview and proof-of-concept observations, J. Geophys. Res., 107, 10.1029/2001JD000794, 2002. v3 (within 6% of v4) Fig. C. Comparison of airmass results from the methods of DeMajistre and Yee (2002, labeled Yee) and Thomason et al. (19xx, labeled UA), for different wavelengths of aerosol extinction and for molecular, ozone, and water vapor number density. Profiles in the left frames yield the airmasses in the right frames for two altitudes: 8.5 km (top row) and 10.5 km (bottom row). Fig. P. As in Fig. N, but for 21 January 2003. *Manuscript in preparation for Atmos. Chem. Phys. Posted at ftp://science.arc.nasa.gov/pub/aats/pub/phil/ Fig. H. As in Fig. E, but for 24 January 2003.