Spatial representativeness of GCOS Reference Upper-Air Network observations

As the GCOS Reference Upper-Air Network (GRUAN) moves from concept to a functioning, global climate observing network, several questions have been raised pertaining to the spatial representativeness of radiosonde observations from GRUAN sites. These questions include:

How far apart can measurement systems (e.g., radiosondes and ground-based radiometers) be and yet be considered effectively to sample the same atmospheric column?

How far apart can GRUAN sub-sites be and still be considered a single GRUAN site?

How close must a satellite overpass be to a GRUAN site for the GRUAN observation to be useful for calibrating a satellite observation?

How far from the GRUAN station is a reference radiosonde likely to drift before falling back to ground level for possible retrieval?

We address these questions by analysis of horizontal displacements of radiosondes (i.e., balloon drift) from the ground launch location, using archived global radiosonde data. Using two years of data from a network of about 250 stations, we assemble a statistical climatology of such displacements, considering the effects of station location, season, and height. This global climatological study is supplemented by enhanced targeted analysis of accepted and candidate GRUAN sites. We provide quantitative answers to the questions above to help refine the design and operation of GRUAN to optimize its value as a reference climate observing network.