JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 98, NO. D8, PAGES 14,753-14,770, AUGUST 20, 1993 Atmospheric Methane at Cape Meares: Analysis of a High-Resolution Data Base and Its EnvironmentalImplications M. A. IC KHALIL, R. A. RASMUSSEN, AND F. MORAES Global Change Research Center, Department of Environmental Science and Engineering, Oregon Graduate Institute of Science and Technology, Portland, Oregon Between 1979 and 1992we took some 120,000measurements of atmospheric methane at Cape Meares on the Oregon coast. The site is representative of methane concentrations in the northern latitudes (from 30øNto 90øN). The average concentration duringthe experiment was 1698 partsper billion by volume (ppbv). Methane concentration increased by 190 ppbv(or 11.9%) during the 13-year span of the experiment.The rate of increase wasabout 20 ___ 4 ppbv/yr in the first 2 yearsand 10 ___ 2 ppbv/yr in the last2 years of the experiment suggesting a substantial decline in the trendat northern middle and high latitudes (-1 ppbv/yr2). Prominent seasonal cycles were observed. During theyear, the concentration stays moreor less constant until May and thenstarts falling, reaching lowest levels in July and August, then risesrapidly to nearlymaximum concentrations in October. The average amplitude of this cycle is about30 __. 7 ppbvand hasincreased duringthe course of the experiment.Interannual variations with small amplitudes of 2-3 ppbv occurwith periods of 1.4 and 6.5 years. The residual concentrations, after accounting for the trends and cycles, have a standard deviation of 6 ppbv for monthly averaged data and 12 ppbv for the daily data. Mass balance calculations showthat to explain the observed seasonality of concentrations, the emissions must peak in late summer and early fall (August-September). No increases in regional annualemissions are requiredover the last decadeto explainthe data. For further research, readersmay obtain the completeand averaged data from the archives. 1. INTRODUCTION Atmospheric methaneis of considerable scientific interest because its concentration is increasing and it hasan important role in global atmospheric chemistry and the greenhouse effect. In the troposphere, methaneaffects the oxidizing capacity of the atmosphere, it may createozone in the presence of NOx, and it causes global warming. In the stratosphere, methane scavenges chlorine atoms preventingthe destruction of the ozone layer from chlorofluorocarbons and other chlorine- containinggases. It also produces water vapor in the stratosphere, and possibly high clouds, which may have the opposite effect on ozone (for reviews see World Meteorological Organization [1985, 1988, 1989, 1991], Intergovernmental Panel on Climate Change [1990], and Wuebbles and Tamaresis [19931). More than a decade ago, automatedmeasurements from Cape Meares established that methane was increasing in the atmosphere at a rapidrate [Rasmussen andKhalil, 1981]. The Cape Meares data have been used in many subsequent analyses of the trends and budgets of methane (see, for example, KhalilandRasmussen [1983,1990]). The subject of this paper is the present record,which spans the 13 years between 1979 and 1992 containing some 120,000 individual measurements linkedto a single absolute calibration standard. We will discuss the nature of the data in section 2. Section 3 ison the most significant patterns in the dataset,namely the trends andseasonal cycles; bothhave beenchanging during the pastdecade. The validity of the measurements is discussed in section 4. The Cape Meares data provideinformation on the nature and seasonality of sources and sinksas discussed in section 5. We have tabulated the monthly data for useby the readers; the complete data are available from the archives as explained at the end. 2. MEASUREMENTS AND THE NATURE OF THE DATA The measurements weretakenusing an automated sampling and measurement system. At the heart of the system is a gas chromatograph with a flame ionizationdetector (GC/FID). Air isdrawn with a pump and dried to a dewpointof-30øC by a Nation© Dryer. The sample is injected into the gas chromatograph and the analysis cycle begins.Eachanalysis of an ambientair sample is followed by an analysis of a precisely calibrated laboratory standard. The chromatographic peaks representing methane (and CO and CO 2 in the earlypart of the experiment) are integrated using a electronic integrator. At the start of the experiment a Carle 211-MS Gas Chromatograph was installed. In mid-1985 a Hewlett-Packard GC was also installed so that the Carle instrument could be replaced.The two instruments were operated simultaneously until the beginning of 1987, after which the Carle instrument was removed. The overlap period is used to adjust for the small differencebetweenthe instruments. The overlap also allowed time to fine tune the Hewlett-Packard GC to obtain a high degree of precision before the established instrument was discontinued. Details of the analytical system and experimental methodology are discussed by Rasmussen and gha•i• [• 98•]. The ambient concentration C(t) is calculated as Copyright 1993 by the AmericanGeophysical Union. Cmeas (t) = Aa(t)CJ{1/2[A•(t-• ) + As(t+•)]} (1) Papernumber 93JD01197. 0148-0227/93/93JD-01197 $0.5.00 whereC a is the concentration in the standard, Aa is the peak area for methane in the ambient sample, and A•(t-•) and 14,753