Vaisala has recently introduced a software version 3.64 for the Digicora III radiosonde processing system. This software version includes several important correction algorithms, which lead to an improved humidity product of the Vaisala RS92 radiosonde. It has been well documented that the Vaisala RS92 radiosonde processed under previous software versions suffered from a daytime dry bias induced by solar heating of the humidity sensor. In addition, like all other polymer based RH sensors, the RS92 response to changes in humidity slows down at cold temperatures leading to a noticeable time lag in the upper troposphere.
The new software version 3.64 includes correction algorithms, which largely remove the solar heating induced daytime dry bias and reduces the effect of a decreased sensor response time. Thus humidity measurements using the Vaisala RS92 radiosonde and Digicora software version 3.64 are significantly better than humidity measurements using the same radiosonde, but processed with previous software versions.
This improvement, however, may possibly lead to a problem for data assimilation centers correcting the solar dry bias and for long term time series of humidity using the Vaisala RS92. Radiosonde data transmitted operationally either in TEMP format or in BUFR format, may not include the information which software version or which correction algorithm was used to process the humidity measurements. Therefore, the user of these data, either operational data assimilation centers using these data for numerical weather forecasting, or climate data centers storing these data for long term, may not be able to distinguish, whether the humidity data coming from the Vaisala RS92 have been corrected for day time dry bias or not.
Although the improvement of instruments used in atmospheric observations is desired and manufacturers of meteorological equipment are encouraged to implement important updates in their instrumentation, this improvement may lead to inconsistent data sets, if the changes in instrumentation or processing cannot be tracked and transmitted in the meta data that must accompany the observations.
The GCOS Reference Upper Air Network (GRUAN) bases its processing of Vaisala RS92 data on the raw data stored in the dc3db files, which also contain the software version number. Thus GRUAN is not impacted by this change in software.
However, Vaisala data received at operational data centers may be impacted by the missing software version, if the software version or the presence of a new correction algorithm for humidity measurements are not or cannot be included in the TEMP or BUFR formatted data transmitted through the Global Telecommunication System (GTS).
Further understanding is needed to avoid potential issues resulting from an improvement in observational techniques.