The accuracy of tropospheric hydroxyl radical measurements by long-path absorption spectroscopy is ultimately limited by the uncertainty Or the effective OH absorption cross sections. The latter were determined from calculated spectra for the and Q(1)(2), Q(1)(3), and P-1(1) rotational lines of the OH A(2) Sigma(+), (upsilon'=0) <-- X(2)II, (v''=0) transition at 308 nm. The calculations took into account Doppler broadening, measured data of the collisional broadening of OH by air molecules, and instrumental line broadening effects. The calculated spectra were compared with OH absorption spectra measured in a flow-reactor near room temperature at 1013 hPa. Excellent agreement between calculated and measured OH spectra was achieved using the collision-broadening parameters determined recently by Leonard (1990). The effective absorption cross sections at the line center (peak absorption cross sections) calculated for the Voigt line shape at 300 K and 1013 hPa are: sigma(eff)(($) over tilde upsilon(0))[Q(1)(2)] = (1.67 +/- 0.1) x 10(-16) cm(2), sigma(eff)(($) over tilde upsilon(0))[Q(1)(3)] = (1.4 +/- 0.1) x 10(-16) cm(2), sigma(eff)(($) over tilde upsilon(0))[P-1(1)] = (1.41 +/- 0.1)x 10(-16) cm(2). The full width at half maximum of the spectral lines is 2.5 pm, and the shape factor a is 1.41. Calculations of the effective absorption cross section under different atmospheric conditions demonstrate a strong pressure dependence for all lines. At 2800 m altitude (720 hPa) the peak absorption cross section of the OH Voigt lines is about 35% higher than at sea level. The temperature dependence in an interval of +/- 20 K around room temperature (at 1013 hPa) is only small (+/- 3% for the Q lines and +/- 6% for the P-1(1) line). The estimated total uncertainty of the effective absorption cross section is 8%.
