L.N. Yurganov (1,2, 4 )

Phone (416) 978-0341 Fax (416) 978-8905 E-mail: leonid@atmosp.physics.utoronto.ca

D.A. Jaffe (2)

E-mail: djaffe@gi.alaska.edu

E. Pullman (2)

P.C. Novelli (3)

E-mail:pnovelli@cmdl.noaa.gov

Published in the Journal of Geophysical Research: August, 1998.

Vol. 103 , No. D15 , p. 19,337-19,346

1) Arctic and Antarctic Research Institute, St.-Petersburg, 199397, Russia.

2) Geophysical Institute, University of Alaska, P.O. Box 757320, Fairbanks, AK 99775

3)NOAA , Climate Monitoring and Diagnostic Laboratory, Boulder, CO 80303

4) Department of Physics, University of Toronto, Toronto, Ontario, M5S 1A7, Canada

Abstract.

The results of correlated investigations of atmospheric carbon monoxide in Alaska during the spring-summer of 1995 using three different techniques are presented. CO total column abundance was measured in Fairbanks using IR spectroscopy with the sun as a light source. A new computer retrieval code was developed and compared with the previously used technique. Surface mixing ratios were determined in-situ by gas filter correlation (GFC) and by gas chromatography (GC) with a mercuric oxide reduction detector. Surface measurements were made at two uncontaminated sites: Poker Flat Research Range in interior Alaska and the NOAA Pt. Barrow observatory. In spring, the measurements revealed considerably more CO in the surface layer as compared with the tropospheric mean values determined by spectroscopy. This suggests an accumulation of anthropogenic CO in the boundary atmospheric layer over vast areas of the Northern Hemisphere during the winter. Beginning in mid April, the CO concentration in the troposphere decreases, but the rate of decrease in the surface layer was 2 - 2.5 times greater than that for the troposphere as a whole. By June the surface mixing ratios and mean tropospheric values nearly converged and CO mixing ratio seemed to be almost constant with altitude. The July measurements revealed days with enhanced CO total column burden; these are most likely associated with lifted layers of air, polluted by forest fires.

Introduction

Carbon monoxide (CO) in the surface layer has been measured by many researchers (e.g. Seiler, [1974]; Novelli et al., [1992] and references therein), and in the total atmospheric column using IR spectroscopy and the sun as a light source [Migeotte, 1949; Dianov-Klokov et al., 1989; Zander et al., 1989; Pougatchev and Rinsland, 1995; Yonemura and Iwagami, 1996]. Direct atmospheric observations revealed growing CO concentrations on a hemispheric scale since the early 1950s, with a rate of 0.8-1.5 % per yr or 1.0-1.8 ppb/yr [Dvoryashina et al., 1984; Khalil and Rasmussen, 1988; Zander et al., 1989]. After 1983, however, a stabilization [Yurganov et. al, 1995] or even a decrease of CO [Khalil and Rasmussen, 1994, Novelli et al., 1994] was detected. The influence of UV attenuation by volcanic stratospheric aerosol on OH, the main CO scavenger, has been proposed to explain the behavior of CO mixing ratio after 1982 [Dlugokencky et al., 1996; Yurganov et al., 1997]. Changes in atmospheric CO concentration between 1800 and 1950 has been assessed from the air bubbles enclosed in polar ice cores [Haan et al., 1996]. The annual cycle of carbon monoxide is driven by both natural and artificial processes [Logan et al.,1981; Mueller and Brasseur, 1995]. Surface emissions account for 62% of CO's global source. Fuel combustion and automobile exhaust are concentrated between 30 and 60 N and vary very little during the year. Biomass burning, another surface source, occurs both in the tropics and in forested areas of the Northern Hemisphere (NH) during dry periods of the year. Photochemical oxidation of methane, which is uniformly distributed in the troposphere, strongly depends on solar UV radiation.

CO loss is primarily due to reaction with OH. CO global average lifetime is about two months [Mueller and Brasseur, 1995], but this is dependent on latitude and season. In fall-winter a reduced OH results in CO accumulation in the northern hemisphere. In spring CO mixing ratios begin to drop due to increasing solar radiation, and [OH] and reaches its minimum in summer. Because of its relatively long lifetime, especially in winter, long-range transport of CO is an important mechanism for its global distribution and temporal variations.

This investigation combines the medium resolution solar spectroscopy and in-situ techniques to investigate CO abundance over Alaska in spring - summer, 1995. This approach makes it possible to obtain some information on the vertical distribution of CO in the troposphere and its evolution during the year. The results can not be applied directly for a validation of the MAPS measurements of 1994, but the approach, as well as general conclusions on CO in high latitudes, appear to be helpful for interpretation of satellite data. Our spectrometer was similar to that used earlier (e.g., Dianov-Klokov et al. [1989]), but a new computer-based retrieval procedure has been worked out and a correlation between the two retrieval techniques was tested. For the boundary atmospheric layer we used measurements at the NOAA/CMDL network station Barrow [Novelli et al., 1992] and in-situ continuous measurements made in central Alaska.

NEXT