Auxiliary Material for 2008GL035022 Satellite remote sounding of mid-tropospheric CO2 M. T. Chahine,1 Luke Chen,1 Paul Dimotakis,2 Xun Jiang,1 Qinbin Li,1 Edward T. Olsen,1 Thomas Pagano,1 James Randerson,3 and Yuk L.Yung2 1Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA. 2California Institute of Technology, Pasadena, California, USA. 3Department of Earth System Science, University of California, Irvine, California, USA. The Auxiliary Material includes: This text file Figs. S1, S2_A, S2_B, S2_C, S2_D and S3 Animation Movie Global five-day running mean sequence of July, 2003 AIRS mid-tropospheric CO2 Chahine, M. T., L. Chen, P. Dimotakis, X. Jiang, Q. Li, E. T. Olsen, T. Pagano, J. Randerson, and Y. L.Yung (2008), Satellite remote sounding of mid tropospheric CO2, Geophys. Res. Lett., 35, L17807, doi:10.1029/2008GL035022. The auxiliary material presents additional information and figures on the CO2 retrieval method as well as the results of comparison with in situ aircraft measurements and a sequence of daily frames of the global variations and transport of CO2. I. Notes on AIRS and the VPD retrieval algorithm AIRS is a cross-track scanning grating spectrometer covering the 3.74 um to 15.4 um spectral range with 2378 channels at a nominal spectral resolving power, lambda/delta_lambda = 1200. Profiles of atmospheric temperature T(p), water vapor H2O(p) and ozone O3(p) are retrieved globally in the presence of fractional clouds across a 1650Ękm wide orbital swath at a geospatial resolution of 45 km at nadir [Chahine, et al., S-2006; Divakarla, et al., S-2006; Susskind, et al., S-2006]. The resulting cloud-cleared radiances and profiles are used as input to the CO2 retrieval. After 4 years in space, AIRS has demonstrated an absolute calibration better than 250 mK and a stability of 10 mK/yr with a spectral accuracy of the center frequency of 2 parts per million [Aumann, et al., S-2006]. The vanishing partial derivative method (VPD) was described by Chahine et al. (2005). The method is based on a general property of the total differential of multi-variable functions that requires the first partial derivatives of the function with respect to each unknown to individually vanish at the point of absolute minimum. We determine this point of absolute minimum by means of iteration. The iterative process is initialized with the AIRS cloud-cleared radiances and retrieved geophysical products of T(p), H2O(p) and O3(p). We linearly perturb the temperature, water vapor and ozone and compute the residuals (RMS difference) between the AIRS cloud-cleared radiances and the radiances computed from the perturbed atmospheric state. A solution is reached when the partial derivatives individually become equal to zero. This condition is necessary but in practice may not be sufficient. Therefore, we track the variations of residuals with iteration and only accept solutions whose residuals decrease monotonically with iteration. We impose an additional quality check by requiring that a cluster of four adjacent CO2 retrievals (e.g., a 2x2 array covering a 90x90 km area at nadir) agree to within 2 ppmv in an RMS sense. We accept only clusters satisfying these conditions and report the average of the 2x2 array of CO2 retrievals. Using this approach, we have shown that the retrieved CO2 results are independent of the initial value of CO2 and the error distribution in the retrieved CO2 mixing ratios is nearly Gaussian. Figure S1 shows the tropical, mid-latitude and polar average weighting functions for the 13 selected CO2 channels used in the retrieval, normalized by their peak values. The weighting function peak sensitivities for the tropical, mid-latitude and polar standard atmospheres respectively occur at 390 hPa, 435 hPa and 527 hPa. Their respective widths at half-maximum span the pressure ranges of 640 hPa to 200 hPa, 680 hPa to 180 hPa and 840 hPa to 175 hPa. If elevated topography results in the surface emission contribution to the observed radiance in a channel to exceed 0.05 K, the channel is excised. At least 3 channels must remain for a retrieval to be attempted. II. Comparisons with CONTRAIL in situ measurements In Figure S2 (panels A through D) we compare our retrievals with in situ aircraft flask measurements obtained at cruising altitudes of 9.8Ękm to 11.6Ękm during commercial flights between Australia and Japan [Matsueda et al., 2002], which are now part of the Comprehensive Observation Network for Trace gases by AirLiner (CONTRAIL). These flask measurements have been conducted biweekly since 1995 and have an estimated accuracy <0.5 ppmv. In order to achieve an average of 9 independent AIRS retrievals within +/- 4 hours around each Matsueda point, we require a strip of +/- 1 deg latitude and +/- 15 deg longitude, centered on Matsueda point measurements in order to meet our averaging need. There is good agreement between AIRS retrievals and aircraft observations in the tropics. The standard deviation of the differences between the Matsueda measurements and AIRS retrievals for the data shown in Figure S2 falls within the range 0.4 < std dev < 1.50 ppmv. Figure S2 shows AIRS retrieved CO2 (blue line) and Matsueda biweekly aircraft flask measurements (red dots) for (A) April 2003, (B) July 2003, (C) April 2004 and (D) July 2004. Matsueda flask measurements were made on two commercial flights between Australia and Japan on April 15 and 29, 2003. The AIRS retrieved CO2 are averages of 2 deg latitude by 30 deg longitude grid boxes centered on the commercial flight path longitude (between 145E and 150E) for the month of April, 2003. The purple error bars are the standard deviation of AIRS retrievals in each grid box. The solid grey line depicts the number of AIRS retrievals in each grid box (axis on the right hand side). III. The chemistry transport models For Figure 2 of the paper, we computed the vertical CO2 profiles using three chemistry and transport models (CTMs): the Caltech/JPL 2-D CTM, the GEOS-Chem global 3-D CTM, and the MOZART-2 global 3-D [Shia et al., 2006; Jiang et al., S-2008; Suntharalingam et al., S-2003; Horowitz et al., S-2003]. A summary description of these models is given below, also in Jiang et al. (S-2008). We then convolve the model-simulated profiles of CO2 mixing ratios with the appropriate (tropical, mid-latitude or polar) average weighting function of the AIRS CO2 channels shown in Figure S1. Finally, we compute zonal averages of the convolution result. Transport in the 2-D model is by stream function and horizontal and vertical diffusivities. The stream function is derived from National Center for Climate Prediction (NCEP) reanalysis data [Jiang et al., S-2004], and the 2-D model has been tuned to account for the effect of age of the stratospheric air [Boering et al., 1996; Waugh et al., 2002; Morgan et al., 2004]. The GEOS-Chem model is driven by NASA GMAO/GEOS-4 assimilated meteorological data with a horizontal resolution of 2 deg x 2.5 deg and 30 vertical levels extending to 0.01 hPa. Deep and shallow convections in GEOS-4 are based on Zhang and McFarlane (S-1995) and Hack (S-1994), respectively. Two boundary conditions are applied: (a) a concentration boundary condition in which the CO2 concentrations at the lower boundary are set to the ground-based measurements from GMD7 ([Tans et al., S-1998], and (b) a flux boundary condition with prescribed/parameterized CO2 source and sink terms [Suntharalingam et al., S-2003; Randerson et al., S-1997; Takahashi et al., S-1997; Duncan et al., S-2003]. The MOZART-2 model is driven by NCEP Reanalysis-1 meteorological data with a horizontal resolution of 2.8 deg x 2.8 deg and 28 vertical levels extending up to ~40 km [Horowitz et al., S-2003]. Deep and shallow convections in NCEP Reanalysis-1 are from Pan and Wu (S-1994) and Tiedtke (S-1983), respectively. The boundary condition applied is a concentration boundary condition in which the CO2 concentrations at the lower boundary are set to the ground-based measurements from GMD [GLOBALVIEW-CO2, 2007; Tans et al., S-1998]. IV. Global 5-day sliding average animation of July, 2003 AIRS mid-tropospheric CO2 The animation shows AIRS retrieved CO2 averaged on a scale of 500 km x 500 km and 5-day running averages for the month of July 2003. The transport of the mid-tropospheric CO2 and impact the mid-latitude jet streams in the NH and SH are readily apparent. Figure S3 is frame number 24 of Animation 1, showing the five-day running mean average between July 24-28, 2003. A large plume of CO2 is shown arising from the Sosol facility in South Africa becoming entrained in the southern mid-latitude jet as the NCEP 500 hPa wind speed reaches 28m/s over South Africa (25S-35S, 10E-25E) The CO2 sources supplying this belt are (a) biogenesis activity in South America [Randerson et al., S-1997], (b) fires in both South America and central Africa [Giglio et al., S-2003], which are regions dominated by anticyclonic flows, and (c) clusters of gasification plants in South Africa and Southeastern Australia [Metz et al., S-2005; Marland et al., S-2007], when the mid-latitude jet stream shifts location to be overhead. The dynamic process in the storm track can be observed in a sequence of five-day running mean images in which the CO2 appears to be transported around the globe. The large concentration east of the Andes mountain range is due partly to the orographic uplift of air over nearly 7 km high mountains to the altitude of greater sensitivity by the AIRS CO2 channels. Also visible is the largest stationary source of CO2, located in South Africa at approximate latitude 30?S [IEA GHG, S-2002; Metz et al., S-2005]. Much of the emission volume of this source is associated with the Fischer-Tropsch coal-to-liquid Suid Afrikaanse Steenkool en Olie (Sasol) facility. In the same five-day running mean sequence, one can discern a plume from the South African source as it becomes entrained in the southern mid-latitude jet stream and carried eastward. Auxiliary References All references added in this Auxiliary materials are listed with publication year Preceded with the letter S- to distinguish them from the references in the main paper. Aumann, H. H., S. Broberg, D. Elliott, S. Gaiser, and D. Gregorich (S-2006), Three years of Atmospheric Infrared Sounder radiometric calibration validation using sea surface temperatures, J. Geophys. Res., 111, D16S90, doi:10.1029/2005JD006822. Chahine, M.T. et al. (S-2006), The Atmospheric Infrared Sounder (AIRS): improving weather forecasting and providing new insights into climate. Bulletin of the American Meteorological Society, 87, 911-926, doi: 10.1175/BAMS-87-7-911. Divakarla, M. G., C. D. Barnet, M. D. Goldberg, L. M. McMillin, E. Maddy, W. Wolf, L. Zhou, and X. Liu (S-2006), Validation of Atmospheric Infrared Sounder temperature and water vapor retrievals with matched radiosonde measurements and forecasts, J. Geophys. Res., 111, D09S15, doi:10.1029/2005JD006116. Duncan, B. N., R. V. Martin, A. C. Staudt, R. Yevich, and J. A. Logan (S-2003), Interannual and seasonal variability of biomass burning emissions constrained by satellite observations, J. Geophys. Res., 108(D2), 4100, doi:10.1029/2002JD002378. Giglio, L., J.D. Kendall, and R. Mack (S-2003), A multi-year active fire data set for the tropics derived from the TRMM VIRS, Int. J. Remote Sens., 24, 4505, doi: 10.1080/0143116031000070283. Hack, J. J. (S-1994), Parameterization of moist convection in the National Center for Atmospheric Research community climate model (CCM2), J. Geophys. Res., 99(D3), 5551-5568, doi:10.1029/93JD03478. Horowitz, L. W., et al. (S-2003), A global simulation of tropospheric ozone and related tracers: Description and evaluation of MOZART, version 2, J. Geophys. Res., 108(D24), 4784, doi:10.1029/2002JD002853 IEA GHG (S-2002), Opportunities for early application of CO2 sequestration technology. Report Ph4/10, IEA Greenhouse Gas R&D Programme (Cheltenham, UK). Jiang, X., C. D. Camp, R. Shia, D. Noone, C. Walker, and Y. L. Yung (S-2004), Quasi-biennial oscillation and quasi-biennial oscillation annual beat in the tropical total column ozone: A two-dimensional model simulation, J. Geophys. Res., 109, D16305, doi:10.1029/2003JD004377. Jiang, X. et al. (S-2008), Simulation of upper tropospheric CO2 from chemistry and transport models. accepted Global Biogeochemical Cycles. Marland, G., T. A. Boden, and R. J. Andres (S-2007), Global, Regional, and National CO2 Emissions. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, (Oak Ridge, TN). Metz, B., O. Davidson, H. de Coninck, M. Loos and L. Meyer (eds.), (S-2005), IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change (Cambridge U.P., New York, NY). Pan, H. L. and W. S. Wu (S-1994), Implementing a mass flux convection parameterization package for the NMC medium-range forecast model. NMC office note, No. 409, 40pp. Randerson, J. T., M. V. Thompson, T. J. Conway, I. Y. Fung, and C. B. Field (S-1997), The contribution of terrestrial sources and sinks to trends in the seasonal cycle of atmospheric carbon dioxide. Global Biogeochem. Cycles, 11(4), 535-560. Suntharalingam, P., C. M. Spivakovsky, J. A. Logan, and M. B. McElroy (S-2003), Estimating the distribution of terrestrial CO2 sources and sinks from atmospheric measurements: Sensitivity to configuration of the observation network, J. Geophys. Res., 108(D15), 4452, doi:10.1029/2002JD002207. Susskind, J., C. Barnet, J. Blaisdell, L. Iredell, F. Keita, L. Kouvaris, G. Molnar, and M. Chahine (S-2006), Accuracy of geophysical parameters derived from Atmospheric Infrared Sounder/Advanced Microwave Sounding Unit as a function of fractional cloud cover, J. Geophys. Res., 111, D09S17, doi:10.1029/2005JD006272. Takahashi, T., et al. (S-1997), Global air-sea flux of CO2: An estimate based on measurements of sea-air pCO2 difference. Proc. Natl. Acad. Sci. U. S. A., 94, 8929. Tans, P.P., et al. (Eds.), (S-1998), Carbon Cycle, in Climate Monitoring and Diagnostics Laboratory No. 24 Summary Report 1996-1997, edited by D.J. Hoffmann et al., chap. 2, pp. 30-51, NOAA Environ. Res. Lab., Boulder Colo. (Data available from http://www.cmdl.noaa.gov/). Tiedtke, M. (S-1983), The sensitivity of the time-mean large-scale flow to cumulus convection in the ECMWF model. ECMWF Workshop on Convection in Large-Scale Models, 297-316 (Reading, England). Zhang, G. J. and N. A. McFarlane (S-1995), Sensitivity of climate simulations to the parameterization of cumulus convection in the Canadian climate centre general circulation model. Atmos. Ocean, 33, 407-446.