An Evaluation of the Geophysical Variability in Stratospheric Aerosol and Gas Experiment II Water Vapor Measurements

The Stratospheric Aerosol and Gas Experiment (SAGE) II instrument has collected vertical profiles of stratospheric ozone, nitrogen dioxide, water vapor, and aerosol extinction at four wavelengths with high resolution since the program's inception in October, 1984. The latest version (v6.1) of SAGE II water vapor measurements in particular exhibit clear biases in comparisons with other instruments but were of sufficient quality to provide evidence of long-term stratospheric increases, in agreement with independent measurements. This bias has often been attributed to the inability to completely clear the effects of aerosols. The SAGE II processing team has made great strides in characterizing stratospheric water vapor. Recent advancements in the retrieval algorithm has significantly reduced this bias except in the proximity of heavy aerosol levels. These advancements include a single wavelength shift for 1986-present and the combination of 2 aerosol models. There exists a clear requirement for evaluating the water vapor's regular long-term behavior, and for extraction of information pertinent to broader studies of atmospheric radiation, dynamics, and chemistry. This data set becomes evermore valuable as time goes on due to its usefulness for global change studies. We present results characterizing the geophysical variability in the new v6.2 water vapor retrievals. Analysis techniques include performing singular value decomposition on the covariances between the water vapor and QBO time series to characterize the inter-annual variability in the data and empirical orthogonal function analysis to reproduce the tropical tape recorder. These results will be compared to HALOE v19 and SAGE II v6.1 derived patterns to objectively assess the algorithm improvements.