A Comparison of Water Vapor Measurements Made by Raman Lidar and Radiosondes

Abstract

This paper examines the calibration characteristics of the NASA/GSFC Raman water vapor lidar during three field experiments that occurred between 1991 and 1993. The lidar water vapor profiles are calibrated using relative humidity profiles measured by AIR and Vaisala radiosondes. The lidar calibration computed using the AIR radiosonde, which uses a carbon hygristor to measure relative humidity, was 3%?5% higher than that computed using the Vaisala radiosonde, which uses a thin film capacitive element. These systematic differences were obtained for relative humidities above 30% and so cannot be explained by the known poor low relative humidity measurements associated with the carbon hygristor. The lidar calibration coefficient was found to vary by less than 1% over this period when determined using the Vaisala humidity data and by less than 5% when using the AIR humidity data. The differences between the lidar relative humidity profiles and those measured by these radiosondes are also examined. These lidar?radiosonde comparisons are used in combination with a numerical model of the lidar system to assess the altitude range of the GSFC lidar. The model results as well as the radiosonde comparisons indicate that for a lidar located at sea level measuring a typical midlatitude water vapor profile, the absolute error in relative humidity for a 10-min, 75-m resolution profile is less than 10% for altitudes below 8.5 km. Model results show that this maximum altitude can be extended to 10 km by increasing the averaging time and/or reducing the range resolution.