MEASUREMENT OF AIR POLLUTION FROM SATELLITES (MAPS) 1994 CORRELATIVE ATMOSPHERIC CARBON MONOXIDE MIXING RATIOS Last Revision: December 1998 CONTRIBUTORS: Paul Novelli NOAA Climate Monitoring and Diagnostics Laboratory (CMDL) 325 Broadway Boulder, Colorado 80303 Ken Masarie Cooperative Institute for Research in Environmental Sciences (CIRES) University of Colorado Boulder, Colorado 80309 COMPILED BY: Linda J. Allison (lja@ornl.gov) and Tom Boden (tab@ornl.gov) Carbon Dioxide Information Analysis Center World Data Center-A for Atmospheric Trace Gases Oak Ridge National Laboratory U.S. Department of Energy Oak Ridge, Tennessee 37831-6335 ===================================================================================== This descriptive file (db1020.txt) contains the following sections: * Introduction * File Descriptions * SAS and Fortran Codes to Read the Data Files * Data Checks and Processing Performed by CDIAC * Instructions for Obtaining the Files * References ===================================================================================== INTRODUCTION Carbon monoxide (CO) mixing ratios in the middle troposphere have been examined for short periods of time using the Measurement of Air Pollution from Satellites (MAPS) instrument. MAPS measures CO from a space platform using gas filter correlation radiometry (Reichle et al. 1986, 1990). During the 1981 and 1984 flights of MAPS, measurement validation was attempted by comparing space-based measurements of CO to those made in the middle troposphere from aircraft. These correlative measurements were made by two laboratories using different methods and calibration scales. The effectiveness of the validation programs was limited by the sparsity of and inconsistencies in the correlative measurements. The latter were traced to differences in the calibration scales used by the participating laboratories (Reichle et al. 1990). Before the 1994 MAPS flights aboard the space shuttle Endeavour (Reichle et al., in press), a correlative measurement team was assembled. This team agreed to provide the MAPS team with results of their field measurement programs during the April and October missions. To maximize the usefulness of the correlative data, all team members agreed to participate in an intercomparison of CO measurements. Eleven laboratories (Table 1) participated in the intercomparison experiment. The principal investigators, their analytical methods and calibration schemes, and the locations of their field measurements are provided in Novelli et al. (1998). The correlative data, which are presented in this database, provide an internally consistent, ground-based picture of CO in the lower atmosphere during Spring and Fall 1994. The data show the regional importance of two CO sources: fossil-fuel burning in urbanized areas and biomass burning in various regions of the Southern Hemisphere (Novelli et al. 1998). ------------------------------------------------------------------------------------ Table 1. Laboratories, Principal Investigators (PIs), and Sampling Locations Constituting the 1994 MAPS Correlative Measurement Team ------------------------------------------------------------------------------------ No. Laboratory PI Sampling Locations --- ---------- -- ------------------ 1. Climate Monitoring and Paul Novelli Pt. Barrow, Alaska Diagnostics Laboratory Mauna Loa, Hawaii (NOAA/CMDL) Niwot Ridge, Colorado NOAA/CMDL Cooperative Air Sampling Network 2. Council for Scientific Ernst Brunke Cape Point, South Africa and Industrial Research (CSIR) 3. Commonwealth Scientific Paul Steele Cape Grim, Tasmania and Industrial CSIRO Air Sampling Research Organisation Network (CSIRO) 4. Fraunhofer Institute for H. Eckhart Scheel Zugspitze, Germany Atmospheric Environmental Research (IFU) 5. Hong Kong Polytechnic K.S. Lam Cape D'Aguilar, University (HKP) Hong Kong 6. Instituto Nacional de Volker Kirchhoff Cuiaba, Brazil Pesquisas Espaciais (INPE) 7. Japan Meteorological Takayo Matsuo Ryori, Japan Agency (JMA) Minamitorishima, Japan 8. NASA-DC8 (NASA-Langley Bruce Anderson Aircraft transects over Research Center) North America 9. National Institute of Carl Baring Head, New Zealand Water and Atmospheric Brenninkmeijer* Research (NIWA) 10. NOAA Aeronomy Laboratory David Parrish Sable Island, Canada (NOAA_AL) 11. University of Maryland Bruce Doddridge Heimeay, Iceland (UMD) Mace Head, Ireland Ragged Point, Barbados Chesapeake Bay, Maryland ----------------------------- *Now at Max Planck Institute, Mainz, Germany ------------------------------------------------------------------------------------ ====================================================================================== FILE DESCRIPTIONS The MAPS correlative database consists of 155 files: * this descriptive file; * the original MAPS correlative data files submitted by CMDL to CDIAC: 136 individual site data files (69 locations, April and October sampling dates; not all stations were sampled both months); * one single merged file containing all 1994 MAPS correlative CO mixing ratio data; * five subsets based on sampling date (April and October) and sampling technique (Flask, In Situ, and Aircraft); * 10 calibration summary files; and * Fortran and SAS programs for reading the data (.dat) files. The CMDL flask data files contain only "edited" and "selected" data, while CMDL in situ files were "edited" but not "selected." CMDL flask samples were collected only at times when winds are from the background oceanic or continental sectors. The "editing" process identifies samples with known instrument problems. The "selection" process screens samples that may not represent "background" conditions. The CMDL flask data were "selected" using a statistical routine which fits a smoothed curve to the data points and flags those points which fall outside a 3 sigma threshold. This step is repeated until no samples are flagged. The second step of the CMDL "selection" process uses an automated procedure to compare the results from each flask of a pair and flag pairs with differences greater than 3 ppb. Each participating laboratory used its own screening and selection processes. Further details on the data processing and handling for the individual sites may be obtained from the PIs listed in Table 1. Format and Content of the Data (.dat) Files CO mixing ratios are found in the .dat files. The original MAPS correlative data files submitted to CDIAC by CMDL are identified by thirteen-character codes ending with the extension .dat (e.g.,maq_02_199410.dat). The first three characters of this code identify the sampling site (see Table 2.). The next three characters identify the sampling technique, except for the shipboard sampling: _00 indicates the data were collected from an aircraft, _01 indicates land sites having in situ measurements, _02 indicates the data result from flask samples. For the shipboard data sets (POC and SCS), the second three characters of the code convey the sampling latitude location (e.g., N05). Digits 8-13 of each data file name indicate the sampling period year and month as yyyymm. ------------------------------------------------------------------------------ Table 2. Sampling Sites in the MAPS 1994 Correlative Sampling Network ------------------------------------------------------------------------------ Code Site Latitude Longitude ---- ---- -------- --------- ALT Alert, N.W.T., Canada 82 27 N 62 31 W ASC Ascension Island 07 55 S 14 25 E BAL Baltic Sea, Poland 55 30 N 16 40 E BHO Baring Head, New Zealand 41 24 S 174 52 E BME St. Davids Head, Bermuda 32 22 N 64 39 W BMW Southhampton, Bermuda 32 16 N 65 53 W BRW Barrow, Alaska 71 19 N 156 36 W CBA Cold Bay, Alaska 55 12 N 162 43 W CFA Cape Ferguson, Australia 19 17 S 147 04 E CGA Cape Grim, Tasmania (CSIRO) 40 41 S 144 41 E CGO Cape Grim, Tasmania (NOAA) 40 41 S 144 41 E CHR Christmas Island 01 42 N 157 10 W CMO Cape Meares, Oregon 45 29 N 123 58 W CPT Cape Point, South Africa 34 21 S 18 29 E CUI Cuiaba, Brazil 15 22 S 56 04 W DC8 N. America aircraft transects* 25-56 N 122-167 W GMI Guam, Mariana Islands 13 26 N 144 47 E GOZ Dwejra Point, Gozo, Malta 36 03 N 14 11 E HKP D'Aguilar, Hong Kong 22 12 N 114 18 E HUN Hegyhatsal, Hungary 46 58 N 16 23 E ICE Heimaey, Iceland 63 15 N 20 09 W ITN Grifton, North Carolina 35 21 N 77 23 W IZO Tenerife, Canary Islands 28 18 N 16 29 W KEY Key West, Florida 25 40 N 80 12 W KUM Cape Kumahkahi, Hawaii 19 31 N 154 49 W MAQ Macquarie Island, Australia 54 29 S 158 58 E MAW Mawson, Antarctica 67 37 S 62 01 E MBC Mold Bay, Canada 76 15 N 119 21 W MCI Minamitorishima, Japan 24 18 N 153 58 E MHT Mace Head, Ireland 53 20 N 09 54 W MID Sand Island, Midway 28 13 N 177 22 W MLO Mauna Loa, Hawaii 19 32 N 155 35 W NWR Niwot Ridge, Colorado 40 03 W 105 35 W POC Pacific Ocean Ship Transects POC000 00 00 N 163 00 W POCN05 05 00 N 158 00 W POCN10 10 00 N 152 00 W POCN20 20 00 N 140 00 W POCN25 25 00 N 134 00 W POCN30 30 00 N 126 00 W POCS05 05 00 S 168 00 W POCS10 10 00 S 174 00 W POCS15 15 00 S 178 00 W POCS20 20 00 S 175 00 E POCS25 25 00 S 168 00 E POCS30 30 00 S 160 00 E PSA Palmer Station, Antarctica 64 55 S 64 00 W PTC Point Charles, Australia 12 25 S 130 19 E QPC Qinghai Province, China 36 16 N 100 55 E RAM Cape Rama, India 15 05 N 73 50 E RPB Ragged Point, Barbados 13 10 N 59 26 W RYO Ryori, Japan 39 02 N 141 50 E SBL Sable Island, Canada 43 56 N 60 01 W SCS South China Sea Ship Transects SCSN03 03 00 N 105 00 E SCSN06 06 00 N 107 00 E SCSN09 09 00 N 109 00 E SCSN12 12 00 N 111 00 E SCSN15 15 00 N 113 00 E SCSN18 18 00 N 113 30 E SCSN21 21 00 N 114 00 E SEY Mahe Island, Seychelles 04 40 S 55 10 E SMO Tutuila, American Samoa 14 15 S 170 34 W SPO South Pole, Antarctica 89 59 S 24 48 W SYO Syowa, Antarctica 69 00 S 39 35 E TAP Tae-ahn Peninsula, S. Korea 36 44 N 126 08 E TDF Tierra del Fuego, Argentina 54 52 S 68 29 W UMD Univ. of Maryland 38 54 N 76 04 W UTA Wendover, Utah 39 54 N 113 43 W UUM Ulaan Uul, Mongolia 44 27 N 111 06 E ZEP Spitsbergen, Norway/Sweden 78 54 N 11 53 E ZUG Zugspitze, Germany 47 15 N 10 58 E --------------------------------------------- *Multiple latitude and longitude locations ------------------------------------------------------------------------------ Examples of data set names for the individual site data sets are: cgo_00_199404.dat = Aircraft data from April 1994 flights above Cape Grim, Australia zug_01_199404.dat = In situ measurements made at Zugspitze, Germany during April 1994 asc_02_199410.dat = Flask samples collected at Ascension Island in October 1994 scsn18_02_199410.dat = Flask samples collected in the South China Sea at latitude 18 deg N in October 1994 Data set names for the data sets compiled by CDIAC are: Temporal Data Sets: april.dat, october.dat Merged Data Set: correlat.dat Sampling Technique Data Sets: flask.dat, in_situ.dat, and aircraft.dat. Each data record in the site.dat files, with the exception of the DC8 data sets, is presented in the following format: code year doy hr mn sc alt lat lon co(csiro) co(cmdl) CFA_02 1994 101 6 54 -9 0.00 -19.280 147.057 50.0 50.8 The DC8 data sets include ozone (o3) measurements and are presented in the following format: code year doy hr mn sc alt lat lon co(nasa) co(cmdl) o3(nasa) DC8_00 1994 100 12 56 59 -999.99 -999.999 0.000 -999.9 -999.9 28.5 Data sets compiled by CDIAC include a column for ozone data: code year doy hr mn sc alt lat lon co co(cmdl) o3 ALT_02 1994 97 15 2 -9 0.21 82.450 -62.520 195.50 195.50 -999.90 ALT_02 1994 97 19 17 -9 0.21 82.450 -62.520 190.90 190.90 -999.90 The variable formats are described in Table 3. --------------------------------------------------------------------------------- Table 3. Variable formats for the MAPS correlative CO mixing ratio database --------------------------------------------------------------------------------- Variable Variable Variable Starting Ending name type format column column -------- ---------- -------- -------- -------- code Character 6 1 6 year Numeric 4 9 12 doy Numeric 3 15 17 hr Numeric 2 20 21 mn Numeric 2 24 25 sc Numeric 2 28 29 alt Numeric 8.2 32 39 lat Numeric 8.3 42 49 lon Numeric 8.3 53 60 co Numeric 8.2 66 73 co_cmdl Numeric 8.2 77 84 o3 Numeric 8.1 86 93 ---------------------------------------------------------------------------------- Where CODE is the 6-character site code; YEAR is the sampling year (1994); DOY is the day of year when the sample was collected; HR is the hour of the sampling time in GMT (missing values are denoted by -9); MN is the minute of the sampling time in GMT (missing values are denoted by -9); SC is the second of the sampling time in GMT (missing values are denoted by -9); ALT is the altitude of the sampling location. For aircraft sampling locations, the altitude is expressed in km. For all other sampling locations, the altitude is expressed in meters above sea level (masl). Site elevations are believed to be surface values for the location of the site, rather than the true sampling height. For the CMDL flask program, the true sampling height is the surface plus three meters. Many observatories sample from intake lines 10-20 m above the surface; however, the intake heights for the non-CMDL programs are not reflected in the values given here. (missing values are denoted by -999.99)*; LAT is the station latitude in decimal degrees (negative values indicate the Southern Hemisphere; missing values are denoted by -999.999)*; LON is the station longitude in decimal degrees (negative values indicate the Western Hemisphere; missing values are denoted by -999.999)*; CO is the CO mixing ratio provided by the laboratory that made the measurements [name in parentheses, e.g., (csiro)]. Mixing ratios are expressed in parts per billion (ppb), relative to the standards used by that laboratory, and missing values are denoted by -999.90; CO_CMDL is the CO mixing ratio from column 10 (CO), normalized to the CMDL scale using results from the intercomparison of 4 CO-in-air standards ranging from 50 to 200 ppb. Calibration results of each lab were plotted against the CO values. CMDL assigned the standards and a second order distance orthogonal regression was then used to relate the two data sets. The curve coefficients were then applied to the original field data. The uncertainty assigned each mixing ratio used in the regression was equal to the square root of the standard error. Statistics from the fits are provided in the .sum files (described below). Details of the data handling procedures are given in Novelli et al. (1998). (missing values are denoted by -999.90). The Pacific and South China Sea shipboard results represent the average of mixing ratios combined into 5-degree latitude bands. They do not account for occasional changes in longitude from cruise to cruise; longitudinal differences are averaged to generate mean values. The Pacific cruises have different northward and southward cruise tracks (see Novelli et al. 1998); again, the results are in 5-degree latitude bands without concern for the longitude. O3 is the ozone concentration expressed in ppb. Ozone values are provided only for the North America Aircraft Transects (DC8) data set for April 1994. Ozone measurements were made on the April flights because NASA was involved in another program that required O3 just before the April deployment. These ozone measurements provide a convenient discriminator between the troposphere and the stratosphere (missing values are denoted by -999.9). --------------------------- * Missing and negative altitude values and missing latitude and longitude coordinates in the DC8 aircraft data resulted from the loss of signal from the aircraft data system. This occurred near the beginning or end of the flights when the operator was either bringing the system up or shutting it down. -------------------------------------------------------------------------------- ================================================================================== Calibration and Reprocessing Summary Files (.sum) Statistics resulting from using the second order distance orthogonal regression are provided in the .sum files. Results of the intercomparison of 4 CO-in-air standards were used to normalize all field measurements to the CMDL CO reference scale. The information used for the normalizations is provided in the lab.sum files. For example, information on the normalization of the IFU field data is provided in the file ifu.sum (see below). Included in this file are the calibration results for both IFU and CMDL that were used to reprocess the IFU field data. These files are available from the individual investigators (see Table 1). Also included are the coefficients derived from the orthogonal regression used to relate the results of the two laboratories. Contents of ifu.sum: co (lab) stddev weight co (cmdl) stddev weight 51.90 1.800 0.31 51.60 0.340 8.65 96.90 2.300 0.19 97.50 0.630 2.52 149.00 2.300 0.19 147.80 0.660 2.30 200.40 2.100 0.23 197.60 1.260 0.63 # coef stddev 0 -1.779512 1.719400 1 1.043337 0.033063 2 -0.000245 0.000132 Examples of data set names for the statistics data sets are: csir.sum noaa_al.sum ====================================================================================== SAS AND FORTRAN CODES TO READ THE DATA FILES The following SAS and Fortran 90 codes can be used to read and print the data files. SAS code: *************************************************** *DAT.SAS: SAS Code to read any of the .dat files.** ***************************************************; filename alt4 'alt_02_199404.dat'; data a; infile alt4 lrecl=93 firstobs=16 missover; input @1 code $6. @9 year @15 doy @20 hr @24 mn @28 sc @32 alt @42 lat @53 lon @66 co @77 co_cmdl @86 o3; data b; set a ; format lat lon 8.3 co co_cmdl o3 8.2; proc print data = b; run; Fortran 90 codes: ********************************************************** *DAT.FOR: Fortran 90 code to read any of the .dat files * ********************************************************** CHARACTER COLUMNS*82, SITE*3, DTYPE*3 INTEGER YEAR, DOY, HR, MN, SC REAL ALT, LAT, LON, CO, CO_CMDL OPEN(UNIT=5,FILE='umd_00_199404.dat') OPEN(UNIT=7,FILE='output.file') READ(5,10) COLUMNS 10 FORMAT(//////////////,A82) WRITE(7,15) 15 FORMAT(1X,'SITE',2X,'DATA',29X,'ALTITUDE',6X,'LATITUDE',5X, 1 'LONGITUDE',3X,'CO-LAB',3X,'CO-CMDL',/,1X,'CODE', 2 2X,'TYPE',2X,'YEAR',2X,'DOY',2X,'HR',2X,'MIN',2X,'SEC',2X, 3 '(m asl or km)',2X,'(dec. deg.)',2X,'(dec. deg.)', 4 3X,'(ppb)',4X,'(ppb)',/) 20 CONTINUE READ(5,25,END=99) SITE, DTYPE, YEAR, DOY, HR, MN, SC, 1 ALT, LAT, LON, CO, CO_CMDL 25 FORMAT(A3,A3,2X,I4,2X,I3,2X,I2,2X,I2,2X,I2,2X,F7.2,2X, 1 F8.3,3X,F8.3,6X,F6.1,6X,F6.1) WRITE(7,30) SITE, DTYPE, YEAR, DOY, HR, MN, SC, ALT, 1 LAT, LON, CO, CO_CMDL 30 FORMAT(2X,A3,3X,A3,2X,I4,2X,I3,2X,I2,3X,I2,3X,I2,3X,F7.2, 1 7X,F8.3,6X,F8.3,4X,F6.1,3X,F6.1) GOTO 20 99 CONTINUE CLOSE(UNIT=5) CLOSE(UNIT=7) STOP END ************************************************************* *READ03.FOR: Fortran 90 code to read any of the .dat files * *containing ozone measurements * ************************************************************* CHARACTER COLUMNS*82, SITE*3, DTYPE*3 INTEGER YEAR, DOY, HR, MN, SC REAL ALT, LAT, LON, CO, CO_CMDL OPEN(UNIT=5,FILE='dc8_00_199404.dat') OPEN(UNIT=7,FILE='output.file') READ(5,10) COLUMNS 10 FORMAT(//////////////,A82) WRITE(7,15) 15 FORMAT(1X,'SITE',2X,'DATA',29X,'ALTITUDE',6X,'LATITUDE',5X, 1 'LONGITUDE',3X,'CO-LAB',3X,'CO-CMDL',5X,'OZONE',/,1X,'CODE', 2 2X,'TYPE',2X,'YEAR',2X,'DOY',2X,'HR',2X,'MIN',2X,'SEC',2X, 3 '(m asl or km)', 2X,'(dec. deg.)',2X,'(dec. deg.)',2X, 4 '(ppb)',5X,'(ppb)',6X,'(ppb)',/) 20 CONTINUE READ(5,25,END=99) SITE, DTYPE, YEAR, DOY, HR, MN, SC, 1 ALT, LAT, LON, CO, CO_CMDL, O3 25 FORMAT(A3,A3,2X,I4,2X,I3,2X,I2,2X,I2,2X,I2,2X,F7.2,3X, 1 F8.3,3X,F8.3,5X,F6.1,5X,F6.1,5X,F6.1) WRITE(7,30) SITE, DTYPE, YEAR, DOY, HR, MN, SC, ALT, 1 LAT, LON, CO, CO_CMDL,O3 30 FORMAT(2X,A3,3X,A3,2X,I4,2X,I3,2X,I2,3X,I2,3X,I2,3X,F7.2, 1 7X,F8.3,6X,F8.3,4X,F6.1,3X,F6.1,5X,F6.1) GOTO 20 99 CONTINUE CLOSE(UNIT=5) CLOSE(UNIT=7) STOP END ===================================================================================== DATA CHECKS AND PROCESSING PERFORMED BY CDIAC The Carbon Dioxide Information Analysis Center (CDIAC) quality assures (QAs) data before distribution. Data received by CDIAC are rarely in a condition that would permit immediate distribution, regardless of the source. To guarantee data of the highest possible quality, CDIAC conducts QA reviews that involve examining the data for completeness, reasonableness, and accuracy. Although the MAPS correlative data were in excellent condition, CDIAC still conducted QA checks on the data and found a few minor discrepancies. These discrepancies have all been resolved with the assistance of the MAPS correlative team investigators, and thus the files available from CDIAC may differ slightly from those previously available from the CMDL ftp site. The following information summarizes the data processing and QA checks performed by CDIAC. 1. All files were checked for formatting consistency, and exceptions were addressed. 2. All data sets were input to one merged SAS data set, and various SAS procedures were run to check for errors and inconsistencies. Missing values codes were standardized for each variable. Data were plotted for most stations to check for consistency and potential outliers. 3. Station and flight coordinates were mapped using the GMT-SYTEM software to check for latitude or longitude outliers. 4. Six new files were created. ==================================================================================== INSTRUCTIONS FOR OBTAINING THE FILES The MAPS 1994 Correlative Data Base is available free of charge from CDIAC. The data and documentation files are available from CDIAC's anonymous FTP (File Transfer Protocol) area via the Internet. Commands used to obtain the data base are shown below. For addditional information, contact CDIAC. >ftp cdiac.esd.ornl.gov or >ftp 128.219.24.36 Login: anonymous or ftp Password: YOU@your internet address (e.g., tab@ornl.gov) Guest login ok, access restrictions apply. ftp> cd pub/maps-co ftp> dir ftp> get db1020.txt ftp> cd data (change to the directory with the .dat data files) ftp> get aircraft.dat (example for an individual file) or mget filenames (for several files) ftp> quit The complete data and documentation can also be obtained from CDIAC's Web site at the following URL: http://cdiac.esd.ornl.gov/db1020.html. For non-FTP data acquisitions (e.g., floppy diskette, 8-mm tape, etc), users may order through CDIAC's online ordering system (http://cdiac.esd.ornl.gov/pns/how_order.html) or contact CDIAC directly. Address: Carbon Dioxide Information Analysis Center Environmental Sciences Division Oak Ridge National Laboratory U.S. Department of Energy Post Office Box 2008 Oak Ridge, TN 37831-6335, U.S.A. Telephone: (423) 574-3645 (Voice) (423) 574-2232 (Fax) Electronic Mail: cdiac@ornl.gov URL: http://cdiac.esd.ornl.gov/ ==================================================================================== REFERENCES Novelli, P. C., V. S. Connors, H. G. Reichle, Jr., B. E. Anderson, C. A. M. Brenninkmeijer, E. G. Brunke, B. G. Doddridge, V. W. J. H. Kirchhoff, K. S. Lam, K. A. Masarie, T. Matsuo, D. D. Parrish, H. E. Scheel, and L. P. Steele. 1998. An internally consistent set of globally distributed atmospheric carbon monoxide mixing ratios developed using results from an intercomparison of measurements. J. Geophys. Res. 103:19,285-19,294. Reichle, H. G., Jr., B. E. Anderson, V. S. Connors, T. S. Denkins, B. B. Gormsen, R. L. Langenfelds, D. O. Neil, S. R. Nolf, P. C. Novelli, N. S. Pougatchev, M. M. Roell, and L. P. Steele. Space shuttle based global CO measurements during April and October 1994, MAPS instrument, data reduction, and data validation. J. Geophys. Res. (in press). Reichle, H. G., Jr., V. S. Connors, J. A. Holland, W. D. Hypes, H. A. Wallio, J. C. Casas, B. B. Gormsen, M. A. Saylor, and W. D. Hesketh. 1986. Middle and upper tropospheric carbon monoxide mixing ratios as measured by a satellite-borne remote sensor during November 1981. J. Geophys. Res. 91:10865-10887. Reichle, H. G., Jr., V. S. Connors, J. A. Holland, R. T. Sherrill, H. A. Wallio, J. C. Casas, E. P. Condon, B. B. Gormsen, and W. Seiler. 1990. The distribution of middle tropospheric carbon monoxide during early October 1984. J. Geophys. Res. 95:9845-9856. ======================================================================================= =======================================================================================