Atomic oxygen modeling in the upper thermosphere

Empirical models of atomic oxygen in the earth's thermosphere are discussed, and calibration problems in satellite drag and in situ mass spectrometer measurements are reviewed. Models based on drag data and mass spectrometer data are found to agree on average to within 15 percent, suggesting that the absolute values are reasonably well known in the upper thermosphere. Comparison of different models with various data sources show residuals of at least 15 percent which are the results of unmodeled magnetic storm, EUV, and geographical variations and smaller scale variations caused by gravity waves.