Quantification of a dry bias in radiosonde humidity profiles over Antarctica: Implications for calculations of outgoing longwave radiation

Upper tropospheric humidity (UTH) plays a large role in determining the amount of outgoing longwave radiation over the tropics. UTH is routinely measured by radiosondes, even though these instruments are known to have significant biases and uncertainties, particularly for measurements in the upper troposphere. Corrections have been devised to account for many of the known biases, but less is known about the dry bias experienced when the humidity sensor is illuminated with solar radiation (solar-radiation dry bias). In the austral summer of 2003/2004, radiosoundings at Dome C, Antarctica, measured atmospheric conditions similar to those in the tropical upper troposphere in terms of both temperature and humidity. The sonde profiles were corrected for known biases and were then used as input into a radiative transfer model to calculate the clear-sky downwelling spectral infrared-radiance. These calculations are compared to spectral infrared-radiance measurements made at Dome C by the Polar Atmospheric Emitted Radiance Interferometer (P-AERI). In all cases, the calculations have lower radiance from water-vapor emission than do the PAERI measurements, indicating that all of the sondes have a dry bias. To determine the magnitude of this dry bias, the humidity profiles were scaled to obtain agreement between the radiance calculations and measurements. Our results indicate that the dry bias depends on solar zenith angle with larger corrections (20 to 35%) necessary in the afternoon (zenith angles about 35 degrees) than in the evening (0 to 20 %; zenith angles less than 10 degrees). Preliminary estimates indicate that a dry bias of 20% in the tropical UTH (6.8 – 13 km) results in overestimating outgoing longwave flux by about 1.4 W m^-2 , and a dry bias of 30% results in overestimating flux by about 2 W m^-2.