During the past 25 years of NOAA operational polar satellites, it has become evident that a growing problem concerning their utilization in Climate and also Numerical Weather Prediction (NWP) applications are the systematic errors and uncertainties inherent in the satellite measurements. Similar arguments can be made for global radiosonde observations. These uncertainties are often larger than the sensitive signals and processes that satellite and radiosonde measurements are designed to reveal, particularly in the realm of climate. Possible strategies to quantify and compensate for these problems include the analysis of satellite overlap data and/or available collocations of satellite, radiosonde and numerical weather prediction (NWP) observations. However, overlap observations are typically not available except in extreme polar regions, and current strategies for analyzing collocated radiosonde and satellite (and NWP) observations are insufficient, further compounding the inherent uncertainties in these respective data platforms.

A Satellite Upper Air Network is proposed to provide reference radiosonde launches coincident with operational polar satellite(s) overpass. The SUAN consist of 36 global radiosonde stations sub-sampled from the Global Upper Air Network (GUAN), and is designed to provide a robust and reliable sample of collocated global radiosonde and satellite observations conducive to the monitoring and validation of satellite and radiosonde observations. The routine operation of such a network in conjunction with operational polar satellites would provide a long-term record of performance for these critical observations, of particular importance for climate.

The integration and impact of SUAN with respect to global retrieval algorithms for deriving atmospheric products, particularly products for climate is also discussed. Such products require that systematic errors and uncertainties in the measurements be compensated for in the retrieval solution. The lack of a program such as SUAN has resulted in a poor historical record of such parameters and degraded product value for climate. This along with possible retrieval approaches suitable for climate using past, current and future observations are explored.