CARBON-13 ISOTOPIC ABUNDANCE AND CONCENTRATION OF ATMOSPHERIC METHANE FOR BACKGROUND AIR IN THE SOUTHERN AND NORTHERN HEMISPHERES FROM 1978 TO 1989. CDIAC NDP-049 (1995) C.M. Stevens, Chemical Technology Division, Argonne National Laboratory, Argonne, Illinois, U.S.A. CONTENTS I. Acknowledgments II. Background and Source Information III. Methodology IV. Limitations and Restrictions V. File Descriptions "ndp049.dat" (atmospheric methane concentrations and carbon-13 isotopic abundances at globally distributed sites from 1978 through 1989) "ndp049.for" (FORTRAN 77 data retrieval program to read and write the atmospheric methane concentration and carbon-13 isotopic abundance data) "ndp049.sas" (SAS data retrieval program to read and write the atmospheric methane concentration and carbon-13 isotopic abundance data) VI. Data Checks Performed by CDIAC VII. How to Obtain the Data and Documentation VIII. References I. ACKNOWLEDGMENTS Measurements were carried out in the Chemistry Division, Argonne National Laboratory, Argonne, IL and supported by the Office of Basic Energy Sciences, Division of Mathematical and Geosciences, U.S. Department of Energy, Contract W-31-109-Eng-38 and the Interdisciplinary Research Program in Earth Sciences, National Aeronautics and Space Administration, Order No. W-16188. II. BACKGROUND AND SOURCE INFORMATION Atmospheric methane (CH4) may become an increasingly important contributor to global warming in future years. Its atmospheric concentration has risen, doubling over the past several hundred years, and additional methane is thought to have a much greater effect on climate, on a per molecule basis, than additional CO2 at present day concentrations (Shine et al., 1990). The causes of the increase of atmospheric CH4 have been difficult to ascertain because of a lack of quantitative knowledge of the fluxes (i.e., net emissions) from the numerous anthropogenic and natural sources. The goal of CH4 isotopic studies is to provide a constraint (and so reduce the uncertainties) in estimating the relative fluxes from the various isotopically distinct sources, whose combined fluxes must result in the measured atmospheric isotopic composition, after the fractionating effect of the atmospheric removal process is considered. In addition, knowledge of the spatial and temporal changes in the isotopic composition of atmospheric CH4, along with estimates of the fluxes from some of the major sources, makes it possible to calculate growth rates for sources whose temporal emissions trends would be difficult to measure directly. A detailed discussion of the use of carbon isotopic data to elucidate CH4 source fluxes and growth rates is given in Stevens (1993), a reprint of which is included as Appendix B of the hard copy documentation that accompanies this data package. The atmospheric CH4 concentration and carbon-13 isotopic abundance data presented in this package are derived from air samples collected over the period 1978-1989 at globally distributed clean-air sites. The data set comprises 201 records, 166 from the Northern Hemisphere and 35 from the Southern Hemisphere. The air samples were collected mostly in remote rural or marine locations, far from large sources of CH4, and are considered representative of tropospheric background conditions. Sampling locations in the Northern Hemisphere included 32 Pacific Ocean sites and 11 land-based sites. In the Southern Hemisphere, locations included 18 Pacific Ocean sites and 4 land-based sites. At many locations, air samples were collected by C.M. Stevens and his colleagues. For some Southern Hemisphere and Pacific Ocean measurements, and all measurements at Cape Meares, Oregon, samples were obtained from stored air cylinders contributed by R.A. Rasmussen from the sample bank at the Oregon Graduate Institute of Science and Technology (formerly the Oregon Graduate Center), Portland, Oregon, U.S.A. These data represent one of the earliest records of carbon-13 isotopic measurements for atmospheric methane at globally distributed clean-air sites. Additional information, including measurements of the isotopic composition of CH4 sources, estimates of the atmospheric lifetime of CH4, measurements of carbon-14 abundances in atmospheric CH4, and some CH4 source flux estimates derived from emissions inventories, are provided in Stevens (1993). Together with these additional data, the atmospheric 13-CH4 data presented in this package have been used to refine estimates of the global CH4 fluxes from combined rice and cattle production and from biomass burning, to calculate growth rates in CH4 fluxes from biomass burning in the Southern Hemisphere and from natural wetlands in the Northern Hemisphere, and to provide evidence of changes in the rate of the atmospheric removal process (Stevens, 1993). III. METHODOLOGY Sampling Techniques and Conditions The sampling in both hemispheres occurred in two main phases: Phase 1: Samples collected during 1978-83 were, with few exceptions, obtained from the air sample bank of the Oregon Graduate Center in collaboration with R.A. Rasmussen. The samples in the bank had been stored from 1 to 5 years in cleaned and treated stainless steel cylinders of various sizes (approximately 2 to 30 liters) and at various high pressures (up to 25 atmospheres). These cylinders were shipped to Argonne National Laboratory and then transferred by expansion to atmospheric pressure into 33-liter evacuated stainless steel cylinders (untreated WWII-surplus oxygen cylinders) of accurately known volume and analyzed within 1 to 2 days. The air in these stored samples had been collected either at sea or at land stations upwind of any urban or anthropogenic sources of CH4. During this period, some of the Northern Hemisphere samples were collected at the Cape Meares station of the Oregon Graduate Center; the others were taken at sea in the Pacific. Beginning in 1983, samples were also collected in Northern Illinois. The average delta 13-C values of 36 samples collected in Northern Illinois during 1983 were compared with the average of seven samples collected at Cape Meares during the same year and were found to agree within 0.01 per mil. Phase 2: From 1984 through 1989 all samples were collected in Argonne 33-liter evacuated stainless steel cylinders (untreated WWII-surplus oxygen cylinders), either at rural sites in Northern Illinois or, in the case of the Southern Hemisphere samples, at rural sites upwind of Canberra (Australia), or at the shore near the NOAA observatory in American Samoa. A special set of analyses was carried out in 1988 on four samples which had been collected in 1978 and analyzed in 1983, in order to compare results after another five years of storage. The results of these reanalyses are found in the data set (see data for samples collected at Cape Meares, dated 4-4-78, 4-7-78, 10-5-78, and 10-5-78). The Illinois samples were collected mostly in the afternoon on days with winds in excess of 10 mph. Collections in the morning were avoided because of overnight temperature inversions. Typically, only one flask sample was collected at each site. An exception occurred on 9-8-88, when six different samples in six different flasks were collected simultaneously and analyzed at intervals over a six week period in order to ascertain the integrity of samples stored in Argonne cylinders. Multiple samples were combined only when it was necessary to combine air stored in several small cylinders (obtained from the Oregon Graduate Institute) in order to make enough for the 33 liters required for isotopic analysis. Multiple analyses from a single flask occurred in several cases, as in the samples with collection times denoted 11-00-81 (Tasmania) and 05-00-82 (Pacific). (Days denoted as 00, as in 11-00-81, indicate that the exact day of sampling is not known.) Analytical Methods Air samples were processed by the oxidation of the CH4 in the air to CO2 after quantitative removal of atmospheric CO2 and CO2 from oxidation of atmospheric CO by Schutze reagent. (The experimental apparatus used to carry out this procedure is shown schematically in Figure 1 of the hard copy documentation that accompanies this package.) The CO2 from CH4 oxidation was quantitatively separated from the air by distillation and analyzed isotopically. Details of the experimental procedures are as follows. The air pressure in the sampling flask was first measured, then the flask was connected to the air intake line and air was passed through a low-efficiency, high-flow liquid N2 trap to remove water, most of the atmospheric CO2, and non-methane hydrocarbons. Next, the samples were passed through molecular sieve 13X and another liquid N2 trap of high efficiency to further remove any CO2 present. The air stream was then passed through a granular Schutze reagent made up of I2O5 on silica gel, by which the atmospheric CO was oxidized to CO2. Another liquid N2 trap then removed this CO2. The gas was then passed through an electrically heated quartz combustion tube containing platinized silica, which quantitatively oxidizes the CH4 to CO2 and H2O, which are then trapped in high-efficiency, liquid N2 traps. The air in the original cylinder was passed through this treatment until the pressure was reduced to approximately 5% of atmospheric pressure. The flow rate was maintained at a constant flow of approximately 0.5 liters/min by adjustment of the valve on the cylinder until the valve was wide open and the pressure was reduced to approximately 25% of atmospheric pressure; the flow then gradually decreases until terminated at 5% of atmospheric pressure. The line beyond the CH4 combustion section was pumped to a good vacuum and then isolated. The trap with the CO2 and H2O from CH4 oxidation was warmed with a hot water bath, and the CO2 and H2O were transferred with a controlled helium flow through a trap cooled by dry ice/methanol to remove the H2O. The CO2 was condensed in a trap in the micromanometer section, which was then pumped to a good vacuum and isolated. The trap containing the CO2 from CH4 oxidation was then warmed, and the CO2 pressure was measured with a precision of +-0.3%. Finally, the CO2 was distilled to a sample tube for isotopic analysis on a Consolidated-Nier isotope-ratio mass spectrometer. The concentration of CH4 was calculated stoichiometrically with a precision of +-0.5%. The delta 13-C values of the CO2 samples were measured with a precision of +-0.05 per mil. A correction of -0.11 per mil was made for an impurity of N2O produced in the combustion train. The overall uncertainties in the analyzed values of CH4 concentration and isotopic composition are estimated to be 0.02 parts per million (ppm) for the concentration values, 0.2 per mil for isotopic values prior to 1985, and 0.1 per mil for isotopic values after 1985. IV. LIMITATIONS AND RESTRICTIONS The isotopic results for the samples of stored air from 1978 through 1983 were all analyzed during 1982-83 in random order and using a laboratory isotopic standard which had been accurately measured at -24.8 per mil. This standard had been used for many years and was replenished without fractionation 2-3 times per year. The order of processing and analysis of samples was random for both the hemisphere of origin and the chronology of collection; therefore, it is unlikely that there are any systematic errors in the set of samples from this period. The same cannot be said for the samples from 1984 through 1989. During this period, new isotopic working standards were started, having a value nearly the same as the isotopic composition of the CH4 samples in order to reduce the mass spectrometer error resulting from measuring large differences, sometimes called the "memory" effect. There may have been some drift of the value of the isotopic standard due to fractionation while it was consumed during the analysis process, as these standards were not prepared in large amounts as had been the case for the standard used for the earlier samples. The repeat analyses done in 1988 on second aliquots of four of the stored air samples showed an average difference of +0.3 per mil compared to the first analysis done five years earlier using the -24.8 per mil isotopic standard. This difference is attributed to background CH4 contamination from the cylinder during storage. The analyses of all stored air samples were therefore corrected by -0.067 per mil per year of storage time between the collection and analysis dates (see column 9 of the data set). Another check of the last isotopic standard used for the 1988-89 samples indicated a possible drift of about +0.5 per mil. The most important feature of the isotopic data is the measured interhemispheric trend, which is independent of the individual hemispheric trends because the analyses were done with the same standard at more or less the same time period. Mathematical analysis of the hemispheric trends shows that the difference of the hemispheric trends is the most important term in the calculation of the trends of the fluxes in each hemisphere. The average trend of each hemisphere is of secondary importance in the analysis. Hence, systematic errors such as fractionation of the isotopic standard would have only a small effect on the final application of the results. Additional uncertainties may affect the ability to compare isotopic ratios in atmospheric CH4 from clean-air sites with isotopic ratios from the various source types for the purpose of resolving uncertainties in the global methane budget. Perhaps the largest uncertainty is the determination by inventory estimates of the globally averaged isotopic ratio of individual sources such as biomass burning or methane from ruminants. This uncertainty arises because the CH4 from these sources is made up of contributions of uncertain relative amounts globally from metabolically distinct plant types (i.e., C-3 and C-4 plants), which have large isotopic variations. The globally averaged isotopic composition of CH4 sources from coal mining and natural gas losses are subject to large uncertainties of +-4 per mil because of large variations among the many individual sources. V. FILE DESCRIPTIONS Filename: ndp049.dat This 21.5 kB file presents 201 records (166 records from Northern Hemisphere sites and 35 records from Southern Hemisphere sites) of atmospheric methane concentrations and carbon-13 isotopic abundances from background (clean-air) samples collected at globally distributed sites from 1978 through 1989. Each data record presents the following variables: sample collection date; number of samples combined for analysis; sampling location (site name, latitude, and longitude); analysis date; CH4 concentration; carbon-13 isotopic abundance (uncorrected and corrected); and flag codes to indicate outliers, repeated analyses of a sample, and other information. The file can be read by using the following FORTRAN code: C FORTRAN data retrieval routine to read and write the file named C "ndp049.dat". C C Unit 1 is input. C CHARACTER NSAMP, LOCN*27, LAT*9, LON*11, AFLAG, OFLAG, AVCODE*2 INTEGER CMONTH, CDAY, CYEAR, AMONTH, ADAY, AYEAR REAL CONC, C13MEAS, C13CORR OPEN (UNIT=1, FILE='ndp049.dat') READ (1,10) CMONTH, CDAY, CYEAR, NSAMP, LOCN, LAT, LON, AMONTH, 2 ADAY, AYEAR, AFLAG, CONC, C13MEAS, OFLAG, C13CORR, 3 AVCODE 10 FORMAT (//////I2,1X,I2,1X,I2,2X,A1,2X,A26,2X,A9,2X,A11,2X,I2,1X, 2 I2,1X,I2,1X,A1,2X,F5.3,2X,F6.2,1X,A1,2X,F6.2,1X,A2) 20 READ (1,30,END=99) CMONTH, CDAY, CYEAR, NSAMP, LOCN, LAT, LON, 2 AMONTH, ADAY, AYEAR, AFLAG, CONC, C13MEAS, 3 OFLAG, C13CORR, AVCODE 30 FORMAT (I2,1X,I2,1X,I2,2X,A1,2X,A26,2X,A9,2X,A11,2X,I2,1X,I2,1X, 2 I2,1X,A1,2X,F5.3,2X,F6.2,1X,A1,2X,F6.2,1X,A2) GO TO 20 99 STOP END The data can also be read by using the following SAS code: * SAS data retrieval routine to read and write the file named "ndp049.dat"; *; DATA NDP049; INFILE 'ndp049.dat'; IF _N_=1 THEN INPUT ////// CMONTH 1-2 CDAY 4-5 CYEAR 7-8 NSAMP $ 11 @14 LOCN $CHAR26. @42 LAT $CHAR9. @53 LON $CHAR11. AMONTH 66-67 ADAY 69-70 AYEAR 72-73 AFLAG $ 75 @78 CONC 5.3 @85 C13MEAS 6.2 OFLAG $ 92 @95 C13CORR 6.2 AVCODE $ 102-103; ELSE INPUT CMONTH 1-2 CDAY 4-5 CYEAR 7-8 NSAMP $ 11 @14 LOCN $CHAR26. @42 LAT $CHAR9. @53 LON $CHAR11. AMONTH 66-67 ADAY 69-70 AYEAR 72-73 AFLAG $ 75 @78 CONC 5.3 @85 C13MEAS 6.2 OFLAG $ 92 @95 C13CORR 6.2 AVCODE $ 102-103; RUN; where CMONTH is the numeric month of the year in which the air sample was collected; CDAY is the numeric day of the month on which the air sample was collected; CYEAR is the final two digits of the year (since 1900) in which the air sample was collected; NSAMP is the number of samples combined from different locations for a single analysis; LOCN is a descriptive character string consisting of (1) the location of the sampling site and, in some cases, (2) the sample number(s), denoted as one or more numbers following a "#" sign and referring to the identity of the stored air sample(s) contributed from the sample bank at the Oregon Graduate Institute of Science and Technology, Portland, Oregon, U.S.A., by R.A. Rasmussen; LAT is the estimated latitude (or range of latitudes) of the sampling site(s), given in decimal degrees; LON is the estimated longitude (or range of longitudes) of the sampling site(s), given in decimal degrees; AMONTH is the numeric month of the year in which the air sample was analyzed; ADAY is the numeric day of the month on which the air sample was analyzed; AYEAR is the final two digits of the year (since 1900) in which the sample was analyzed; AFLAG is a one-character flag code providing additional information about the sample analysis: 'R' -- entry represents one of two or more repeated analyses carried out on the same air sample; 'C' -- identifies samples where the actual date of analysis is not given but was within one week of the date of collection; CONC is the CH4 concentration in the air sample, given in parts per million (1 * 10**6) by volume; C13MEAS is the measured del C-13 (per mil), corrected by +0.10 per mil for N2O contamination but uncorrected for water vapor contamination; OFLAG is a one-character flag code denoting values of C13MEAS that are considered as outliers and not included in any subsequent averages; the symbol for the flag code is "*"; C13CORR is a corrected value of del C-13 (per mil), calculated only for samples stored for a considerable time before analysis; values represent an addition of -0.067 per mil per year for background contamination (from the cylinder walls) that accumulated between the collection and analysis dates; AVCODE is a two-character code denoting an entry whose C13CORR value is an average of those of the previous 2 to 5 entries [i.e., the immediately preceding entries containing AFLAG values of 'R' (excluding outliers, denoted by OFLAG="*")]: A2, A3, and A5 signify entries representing averages of 2, 3, and 5 previous entries, respectively; Stated in tabular form, the contents include the following. ____________________________________________________________________________ Variable Variable Starting Ending Variable* type width** column column ____________________________________________________________________________ CMONTH Numeric I2 1 2 CDAY Numeric I2 4 5 CYEAR Numeric I2 7 8 NSAMP Character A1 11 11 LOCN Character A26 14 39 LAT Character A9 42 50 LON Character A11 53 63 AMONTH Numeric I2 66 67 ADAY Numeric I2 69 70 AYEAR Numeric I2 72 73 AFLAG Character A1 75 75 CONC Numeric F5.3 78 82 C13MEAS Numeric F6.2 85 90 OFLAG Character A1 92 92 C13CORR Numeric F6.2 95 100 AVCODE Character A2 102 103 ____________________________________________________________________________ * Missing values for numeric variables are represented as follows -- CMONTH, CDAY, CYEAR, AMONTH, ADAY, AYEAR: 00; CONC: 9.999; C13MEAS, C13CORR: 999.99. Missing values for character variables (NSAMP, AFLAG, OFLAG, AVCODE) are represented as blanks. ** Values for variable width are entered as FORTRAN 77 format codes. Filename: ndp049.for This file contains a FORTRAN 77 data retrieval routine to read and write the file "ndp049.dat". The following is a listing of this program. For additional information regarding variable definitions and format statements, please see the file description for "ndp049.dat". C FORTRAN data retrieval routine to read and write the file named C "ndp049.dat". C C Unit 1 is input. C Unit 2 is output. C CHARACTER NSAMP, LOCN*27, LAT*9, LON*11, AFLAG, OFLAG, AVCODE*2 INTEGER CMONTH, CDAY, CYEAR, AMONTH, ADAY, AYEAR REAL CONC, C13MEAS, C13CORR OPEN (UNIT=1, FILE='ndp049.dat') OPEN (UNIT=2, FILE='output') READ (1,10) CMONTH, CDAY, CYEAR, NSAMP, LOCN, LAT, LON, AMONTH, 2 ADAY, AYEAR, AFLAG, CONC, C13MEAS, OFLAG, C13CORR, 3 AVCODE 10 FORMAT (//////I2,1X,I2,1X,I2,2X,A1,2X,A26,2X,A9,2X,A11,2X,I2,1X, 2 I2,1X,I2,1X,A1,2X,F5.3,2X,F6.2,1X,A1,2X,F6.2,1X,A2) 20 WRITE (2,30) CMONTH, CDAY, CYEAR, NSAMP, LOCN, LAT, LON, AMONTH, 2 ADAY, AYEAR, AFLAG, CONC, C13MEAS, OFLAG, C13CORR, 3 AVCODE 30 FORMAT (I2.2,1X,I2.2,1X,I2.2,2X,A1,2X,A26,2X,A9,2X,A11,2X,I2.2, 2 1X,I2.2,1X,I2.2,1X,A1,2X,F5.3,2X,F6.2,1X,A1,2X,F6.2,1X, 3 A2) READ (1,40,END=99) CMONTH, CDAY, CYEAR, NSAMP, LOCN, LAT, LON, 2 AMONTH, ADAY, AYEAR, AFLAG, CONC, C13MEAS, 3 OFLAG, C13CORR, AVCODE 40 FORMAT (I2,1X,I2,1X,I2,2X,A1,2X,A26,2X,A9,2X,A11,2X,I2,1X,I2,1X, 2 I2,1X,A1,2X,F5.3,2X,F6.2,1X,A1,2X,F6.2,1X,A2) GO TO 20 99 CLOSE (1) CLOSE (2) STOP END Filename: ndp049.sas This file contains a SAS data retrieval routine to read and write the file "ndp049.dat". The following is a listing of this program. For additional information regarding variable definitions and format statements, please see the file description for "ndp049.dat". * SAS data retrieval routine to read the file named "ndp049.dat"; *; DATA NDP049; INFILE 'ndp049.dat'; IF _N_=1 THEN INPUT ////// CMONTH $ 1-2 CDAY $ 4-5 CYEAR $ 7-8 NSAMP $ 11 @14 LOCN $CHAR26. @42 LAT $CHAR9. @53 LON $CHAR11. AMONTH $ 66-67 ADAY $ 69-70 AYEAR $ 72-73 AFLAG $ 75 @78 CONC 5.3 @85 C13MEAS 6.2 OFLAG $ 92 @95 C13CORR 6.2 AVCODE $ 102-103; ELSE INPUT CMONTH $ 1-2 CDAY $ 4-5 CYEAR $ 7-8 NSAMP $ 11 @14 LOCN $CHAR26. @42 LAT $CHAR9. @53 LON $CHAR11. AMONTH $ 66-67 ADAY $ 69-70 AYEAR $ 72-73 AFLAG $ 75 @78 CONC 5.3 @85 C13MEAS 6.2 OFLAG $ 92 @95 C13CORR 6.2 AVCODE $ 102-103; FILE 'output'; PUT CMONTH 1-2 CDAY 4-5 CYEAR 7-8 NSAMP 11 @14 LOCN $CHAR26. @42 LAT $CHAR9. @53 LON $CHAR11. AMONTH 66-67 ADAY 69-70 AYEAR 72-73 AFLAG 75 @78 CONC 5.3 @85 C13MEAS 6.2 OFLAG 92 @95 C13CORR 6.2 AVCODE 102-103; RUN; VI. DATA CHECKS PERFORMED BY CDIAC The Carbon Dioxide Information Analysis Center (CDIAC) endeavors to provide quality assurance (QA) of all data before their distribution. To ensure the highest possible quality in the data, CDIAC conducts extensive reviews for reasonableness, accuracy, completeness, and consistency of form. While having common objectives, the specific form of these reviews must be tailored to each data set; this tailoring process may involve considerable programming efforts. The entire QA process is an important part of CDIAC's effort to assure accurate, usable data for researchers. For the atmospheric methane concentration and carbon-13 isotopic abundance data, the QA procedure consisted of the following: 1. The format of all information was checked to ensure consistency throughout each data record. 2. Data values were examined for reasonableness, absence of typographical errors, and absence of outliers. No errors or inconsistencies of the types described above were found in the atmospheric methane concentration and carbon-13 isotopic abundance data received by CDIAC. All data values in the file distributed by CDIAC are identical to those received from C.M. Stevens. However, in order to enhance their value and their ease of use, the data records were reformatted and appended in the following way: 1. The latitude and longitude of each sampling site, which were absent for most locations in the original data set, were obtained from C.M. Stevens and appended to each data record. 2. The conventions for missing values and repeated analyses were altered in order to create a consistent format for all data records. New flag codes were created to identify any of the following: (1) uncertainties in the analysis date, (2) records that represent repeated measurements, and (3) records that represent averages of two or more previous measurements. VII. HOW TO OBTAIN THE DATA AND DOCUMENTATION The data and hard copy documentation described herein are available from: Carbon Dioxide Information Analysis Center Oak Ridge National Laboratory Post Office Box 2008 Oak Ridge, TN 37831-6335, U.S.A. Telephone (615) 241-4851 The following citation should be used for referencing this archive and/or the documentation report: Stevens, C.M. 1995. Carbon-13 Isotopic Abundance and Concentration of Atmospheric Methane in Background Air in the Southern and Northern Hemispheres from 1978 to 1989. ORNL/CDIAC-80, NDP-049. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge, Tennessee, U.S.A. VIII. REFERENCES Rust, F., and C.M. Stevens. 1980. Carbon kinetic isotope effect in the oxidation of methane by hydroxyl. International Journal of Chemical Kinetics 12:371-77. Shine, K.P., R.G. Derwent, D.J. Wuebbles, and J-J. Morcrette. 1990. Radiative forcing of climate. pp. 41-68. In J.T. Houghton, G.J. Jenkins, and J.J. Ephraums (eds.), Climate Change: the IPCC Scientific Assessment. Cambridge University Press, Cambridge. Stevens, C.M. 1993. Isotopic abundances in the atmosphere and sources. pp. 62-88. In M.A.K. Khalil (ed), Atmospheric Methane: Sources, Sinks, and Role in Global Change, Proceedings of the NATO Advanced Research Workshop on the Atmospheric Methane Cycle: Sources, Sinks, Distributions, and Role in Global Change, held at Mt. Hood near Portland, OR, U.S.A., October 7-11, 1991. NATO ASI Series I: Global Environmental Change, Vol. 13. Springer-Verlag, Berlin and Heidelberg. Stevens, C.M. 1988. Atmospheric methane. Chemical Geology 71:11-21. Stevens, C.M., and A. Engelkemeir. 1988. Stable carbon isotopic composition of methane from some natural and anthropogenic sources. Journal of Geophysical Research 93(D1): 725-33. Stevens, C.M., and A. Engelkemeir. 1985. Causes of increasing methane fluxes based on carbon isotopic studies. Special Environmental Report No. 16, WMO 647. WMO Technical Conference on Observation and Measurement of Atmospheric Contaminants, World Meteorological Organization, Geneva.