2084 IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 39, NO. 9, SEPTEMBER 2001 Communications______________________________________________________________________ First Water Vapor Measurements at 183 GHz From the High Alpine Station Jungfraujoch Andreas Siegenthaler, Olivier Lezeaux, Dietrich G. Feist, and Niklaus Kämpfer Abstract—During six months in 1999, we observed the water vapor emis- sion line at 183.31 GHz with a microwave radiometer at the high alpine site Jungfraujoch in Switzerland. We retrieved statistics on the atmospheric transmission and profiles of stratospheric water vapor on selected days. Our site seems well suited for observations of this spectral line. Index Terms—Microwave radiometry, stratospheric water vapor, tropo- spheric transmittance. I. INTRODUCTION Water vapor is a very important trace gas in the stratosphere. On the one hand, it is a useful tracer for atmospheric transport processes due to its long photochemical lifetime. On the other hand, it plays a con- siderable role in chemical processes since it is the major component of the hydrogen budget in the middle atmosphere and a source gas for the highly reactive OH radical [1]. It is also the strongest greenhouse gas. Ground-based millimeter-wave measurements of stratospheric water vapor are possible at 22 GHz due to a high tropospheric transmittance in this frequency range [2], [3]. However, this line is very weak, and rather long integration times are necessary to gain sufficient SNR. The inten- sity of the water vapor line of the transition at 183.310 09 GHz is approximately 180 times stronger than the one at 22 GHz. But most of the time, the atmosphere is too opaque near this line due to the strong absorption by tropospheric water vapor. Therefore, it is difficult to detect the stratospheric signal from the ground at 183 GHz. Ground-based millimeter-wave measurements of middle atmo- spheric water vapor from the emission line at 183.31 GHz were first reported by Pardo et al. [4] and Hartogh et al. [5]. Pardo et al. observed with a 30-m radio telescope located in Pico Veleta, Spain (altitude 2870 m). They conclude that measurements of middle atmospheric H O at 183 GHz are only possible from high altitude observation sites during the winter season under very dry conditions. They already suggested the Jungfraujoch station in the Swiss Alps (46.55 N, 7.98 E, 3580 m altitude) as a possible observation site. For a feasibility study of such measurements, we installed the mi- crowave radiometer Airborne Millimeter and Submillimeter Wave Ob- serving System (AMSOS) at the Jungfraujoch and measured almost continuously from July to December 1999. AMSOS is usually used to observe the water vapor line at 183.31 GHz from an aircraft [6]. II. INSTRUMENTATION AMSOS consisted of a hetorodyne receiver with an uncooled, sub- harmonically pumped Schottky diode mixer, which converted the at- Manuscript received October 5, 2000; revised May 4, 2001. This work was supported by NF-Grant20.55270.98 and EC-Grant ENV4-CT97-0515 (WAVE), respectively, BBW-No. 97.0074, EC-Grant EVK2-CT-1999-00047 (THESEO 2000), and BBW-no. 99.0218-1. The authors are with the Institute of Applied Physics, University of Berne, Berne, Switzerland (e-mail: andreas.siegenthaler@mw.iap.unibe.ch). Publisher Item Identifier S 0196-2892(01)08120-7. mospheric signal to an intermediate frequency (IF) of 3.7 GHz. The IF signal was amplified by a low noise amplifier and a power amplifier and spectrally analyzed with an acousto-optical spectrometer (AOS) with 1725 equally spaced channels over a bandwidth of 1 GHz. Each channel had a frequency resolution of 1 MHz. A Martin-Puplett in- terferometer suppressed the image sideband by more than 25 dB. The single sideband receiver noise was below 4100 K over the whole band- width. The atmospheric signal entered the instrument from a zenith angle of 50 through a styrofoam window. The construction of the ob- servation building did not allow observation at smaller zenith angles. The instrument was calibrated in a total power mode with two black- bodies at ambient and at liquid nitrogen temperature. III. MEASUREMENTS AND RESULTS AMSOS measured continuously on Jungfraujoch from July to De- cember 1999, except for some short shutdowns for maintenance. From the measured spectra we estimated the tropospheric transmittance. On days with high transmittance, we were able to retrieve vertical profiles of the volume mixing ratio of stratospheric water vapor. A. Tropospheric Transmittance Assuming that the troposphere is a horizontally stratified isothermal layer with a mean temperature , the measured brightness temper- ature at a certain frequency can be written as (1) is the brightness temperature that one would measure above the troposphere with the same observation angle. This includes emis- sions from the stratosphere and mesosphere as well as the cosmic back- ground, where (2) is the tropospheric opacity, and is the path along the beam through the troposphere, with the limits at the observation site and at the tropopause. The absorption coefficient describes the absorption along the path. If , , and are known, (1) can be solved for the tropospheric transmittance (3) One has to remember that is the transmittance for the observation angle and not for zenith direction. The tropospheric mean temperature can be estimated by (4) where is the air temperature at the observation site. According to a model by Ingold et al. [7], has a value of 11.5 K ( 2K root mean square error [RMSE]) for the altitude of the Jungfraujoch, a zenith angle of 50 , and a frequency of 183 GHz. For the brightness temperature , we took hourly means of the measured spectra that were averaged over the whole bandwidth. We calculated the stratospheric brightness temperatures with the Bernese Atmospheric Model (BEAM) [8] and estimated an error of less than 2 K. From these values, we derived hourly averaged tropo- 0196–2892/01$10.00 © 2001 IEEE