Understanding and Forecasting Upper Atmosphere Nitric Oxide through Data Mining Analysis of TIMED/SABER Data

Storm time energy input to the upper atmosphere is countered by infrared radiative emissions from nitric oxide (NO). The temporal profile of these energy sources and losses strongly control thermospheric density profiles, which in turn affect the drag experienced by low Earth orbiting satellites. Storm time processes create NO. In some extreme cases an overabundance of NO emissions unexpectedly decreases atmospheric temperature and density to lower than pre-storm values. Quantifying the spatial and temporal variability of the NO emissions using eigenmodes will increase the understanding of how upper atmospheric NO behaves, and could be used to increase the accuracy of future space weather and climate models. Thirteen years of NO flux data, observed at 100-250 km altitude by the SABER instrument onboard the TIMED satellite, is decomposed into five empirical orthogonal functions (EOFs) and their amplitudes to: 1) determine the strongest modes of variability in the data set, and 2) develop a compact model of NO flux. The first five EOFs account for 85% of the variability in the data, and their uncertainty is verified using cross-validation analysis. Although these linearly independent EOFs are not necessarily independent in a geophysical sense, the first three EOFs correlate strongly with different geophysical processes. The first EOF correlates strongly with Kp and F10.7, suggesting that geomagnetic storms and solar weather account for a large portion of NO flux variability. EOF 2 shows annual variations, and EOF 3 correlates with solar wind parameters. Using these relations, an empirical model of the EOF amplitudes can be derived, which could be used as a predictive tool for future NO emissions. We illustrate the NO model, highlight some of the hemispheric asymmetries, and discuss the geophysical associations of the EOFs.