Standards for Evaluating Radiosonde Measurements

Radiosonde measurement errors occur for various reasons: calibration, mishandling of sensors, poor software, radiation, etc. It is important that these errors be found and corrected, or at the least, corrections determined that might be sensibly applied. Radiation errors apparently are the most serious since the lack of radiative equilibrium between the thermistor and its surrounding environment can not be corrected for just a single thermistor without serious intervention. However, errors may be determined using the Accurate Temperature Measuring (ATM) radiosonde. The ATM radiosonde development was initiated in the mid-1980's using three thermistors; five thermistors are presently incorporated in the ATM radiosonde. Test flights at different locations indicated that the radiative effect on the thermistor varies because of the different environment of each location and, can be corrected. When comparisons between different thermistors (radiosondes) are required the ATM is a valuable tool. Investigation of relative humidity measurements is an ongoing issue because of large discrepancies in observed data. The first chilled mirror (SNOW WHITE) radiosonde was flown from Wallops Flight Facility in 1997. Tests and analyses show the chilled mirror radiosonde present very acceptable data up to the altitude of the tropopause and in some measurements, to 100 hPA. Improvement and better interpretation of the measurements are important. Discussion concentrates on new aspects of the ATM radiosonde application and the utility of the chilled mirror radiosonde.