U.S. NWS radiosonde stations now have the capability to encode and disseminate in near real time one-second resolution data including radiosonde drift times and positions. This new capability resulted from a software upgrade that as of the writing of this abstract has been fielded at 18 sites (mostly in the eastern U.S.) and is scheduled to be fielded at nearly all NWS sites by this autumn. High-resolution radiosonde data have been available from most European stations for several years and are currently also available for Australia, New Zealand, Canada, Mexico, South Africa, Chile, South Korea, and French holdings across the globe, along with other locations such as Hong Kong, Singapore, and Israel. In addition, high-resolution data are now or soon will be available from other sources, such as U.S. fire-weather soundings and dropsondes from U.S. NWS Hurricane Hunter aircraft.
These data require processing that differs in some respects from conventional radiosonde data that have been decoded from TEMP messages. The availability of thousands of levels in a sounding means that thinning to a few hundred levels must be performed before the data are presented to a data assimilation system. The presence of both TEMP and BUFR messages requires some degree of duplicate checking, whether it is by message or by pressure level. And, the types of errors that must be trapped by quality control are different in these mostly self-consistent BUFR messages. The U.S. Navy’s global NWP system NAVGEM has been utilizing high-resolution BUFR radiosonde data since September 2015.
The presentation will discuss the current availability of near real time high-resolution BUFR in the U.S. and elsewhere, issues encountered in processing these data for use in the U.S. Navy’s global data assimilation system, and results from assimilating these data, including forecast sensitivity observation impacts based on the adjoint of the global system.