Area averaged comparison of global radiosonde and radio occultation data, 2007-2016

Atmospheric observation using global positioning system (GPS) radio occultation has become an integral part of weather prediction and research activities since its initial implementation in the mid-1990s. Data from radiosondes is used worldwide for numerical weather prediction and research, so it is important that the data obtained from them is reliable and accurate. With increases in radio occultation data and widespread accessibility, it is becoming clear that radio occultation provides a beneficial impact on atmospheric data availability. Radio occultation is not affected by weather conditions and is traceable with high vertical resolution. This makes it an ideal platform for assessing and calibrating other atmospheric sounding equipment. Post-processed COSMIC-1 radio occultation and radiosonde datasets were obtained from the University Corporation for Atmospheric Research (UCAR) COSMIC Data Analysis and Archive Center (CDAAC), where data is freely distributed. Two data sets were used in this analysis – wetprf and sonprf. The wetprf neutral atmosphere data set is an atmospheric radio occultation profile that includes moisture information interpolated to 100 m height levels. Gridded analysis or short-term forecast is used to separate the pressure, temperature, and moisture contributions from refractivity. The sonprf data set contains radiosonde data in netCDF file format from the CDAAC sonprf data set which originates from the National Center for Atmospheric Research (NCAR) mass store, data set 353.4. Temperature, pressure, and moisture profiles were collocated with radio occultation profiles for comparison spanning from the 1st of January 2009 to the 31st of December 2016. To match the radiosonde and radio occultation data a buffer was developed to limit the temporal and spatial difference between data points. A buffer of 3 hours and 200 km was used to collocate observations. Buffers need to allow sufficient data for analysis while encompassing expected radiosonde flight duration without exceeding expected distances for similar weather conditions. A study determining the impacts of changing temporal and spatial buffers demonstrated that there is no significant difference between using a broader buffer for up to 3 hours and 300 km (Zhang et al., 2011). Standard pressure levels (30, 50, 100, 150, 200, 250, 300, 400, 500, 700, and 850 hPa) are used in this data set as these are the common levels used in radiosonde observations. Tropopause values have been determined from the lapse rate. Area averaging was conducted over a 10° x 10° global grid for temperature, moisture, refractivity and observation height, at each pressure level.