Satellite Upper Air Network (SUAN)

During the past 20 years it has become evident that a significant problem inhibiting the use of polar satellite observations in the Numerical Weather Prediction (NWP) and in particular the Climate applications for which they are intended are the systematic uncertainties (bias) inherent in the satellite measurements. These uncertainties are often larger than the sensitive signals normally associated with these applications. At one time it was thought that such errors could be compensated, for example, through analysis of satellite overlap data and/or available collocations of satellite and ground truth (radiosonde) observations. However, overlap observations are typically of too short duration (except in polar regions), and available collocations of satellite and radiosonde observations are not adequate and include systematic uncertainties inherent in the ground-truth radiosonde data. As a result, time consuming and costly efforts to quantify and incorporate such uncertainties have led to a number of different approaches and conflicting results.

A Satellite Upper Air Network (SUAN) consisting of 43 global stations that would routinely provide radiosonde observations coincident with operational polar satellite overpass is discussed. SUAN would not only address the problem of satellite data monitoring and validation, but would also provide useful data for monitoring radiosondes and underlying scientific algorithms, for example, the radiative transfer model(s), all shared items of concern among the satellite product, climate and NWP areas. SUAN represents a first step toward establishing a long-term baseline data set for monitoring bias and uncertainty parameters associated with the critical observations for weather and climate applications in the decades to come. Ongoing programs to try and compile historical data sets of collocated radiosonde and TOVS observations (1979 onward) to "correct" the past are also discussed. An important lesson of the past 20+ years is that validation ultimately determines data usefulness, and without programs like SUAN past problems are destined to be repeated, undermining our research goals and the promise of new millennium technologies.