Remote sensing of boundary layer temperature profiles by a scanning 5-mm microwave radiometer and RASS: Comparison experiments

Two techniques for deriving low-altitude temperature profiles were evaluated in an experiment conducted from November 1996 to January 1997 at the Boulder Atmospheric Observatory (BAO). The first used a scanning, single wavelength, 5-mm (60 GHz) microwave radiometer to measure vertical temperature profiles. Two radiometers were operated simultaneously; one used a discrete scan, the other scanned continuously. The second technique was a Radio Acoustic Sounding System (RASS) that operated at 915 MHz. Typically, radiometric profiles were produced every 15 min; those from RASS were 5-min segments taken every hour. Ground truth for the experiment was available from in situ measurements at the BAG. The BAO has an instrumented 300-m tower with 5-min measurements of temperature and relative humidity available at the surface and at altitudes of 10, 50, 100, 200, and 300 m. The tower measurements were occasionally supplemented with radiosonde releases and with hand-held meteorological measurements taken on the tower elevator. The differences between the radiometers and the tower sensors were about 1 degrees C rms. The accuracy using an in situ temperature measurement at the radiometer height as a predictor was also evaluated; at the 200- and 300-m levels, only about 4 degrees C rms accuracies resulted. During the experiment, the RASS occasionally experienced radio frequency interference; to eliminate these effects, a quality-control algorithm for the RASS system was developed and evaluated. In addition, an experiment was herd in September 1996 at the Department of Energy's Atmospheric Radiation Program Southern Great Plains site in north central Oklahoma. For this experiment, evaluations of a scanning 5-mm radiometer relative to 3-hourly radiosondes are presented. Quality control on the radiosondes was provided by comparisons with independent in situ surface and 60-m tower observations. The agreement between the radiometric profiles and the quality-controlled radiosondes was better than 1 degrees C up to 800 m. Plans for future deployments of these instruments are discussed. In addition to the in situ comparisons, theoretical analyses of the scanning radiometer systems were also conducted. The effects of angular resolution of the current system, noise level, prediction from in situ measurements, and vertical resolution were examined.