Using GPSRO data to examine radiation induced biases in global radiosonde data

Radiosonde temperature measurements, particularly in the upper troposphere and stratosphere, have long been known to suffer from solar and infrared radiation induced biases, and corrections are required on radiosonde instrument data to eliminate the biases. Soundings transmitted through the Global Telecommunication System (GTS) may contain radiation corrections applied at each site using schemes provided by radiosonde vendors or the originating country. Because these corrections were derived using limited data and possibly non-optimal methods, forecasting agencies such as the National Centers for Environmental Prediction (NCEP) apply further corrections in the process of data assimilation for numerical weather prediction. The NCEP radiation correction was developed based on the comparison of radiosonde data with NCEP's Global Data Assimilation System 6-hr forecast and the use of Vaisala RS-80 as the reference sonde.

Due to the complexity of radiation corrections, radiation biases may still be present in all radiosonde data as noticed by previous studies. This work employs three years (2008-2010) of highly precise and accurate GPS Radio Occultation (GPSRO) data as a reference for assessing the radiation biases for sonde types flown in the global operational radiosonde network. Collocated radiosonde and GPSRO data are collected with the NOAA satellite Products Validation System (NPROVS) operated at the NOAA/NESDIS Center for Satellite Application and Research (STAR). This evaluation will be conducted on the original GTS and NCEP-adjusted soundings as well as using the NCEP model background for additional tests. We will examine how the radiosonde biases vary with solar angle, geographic region, cloud cover and season. These results are applicable to the improvement of radiation correction schemes and the better use of in situ sounding data in satellite retrieval calibration/validation.