Comparison of distributions derived from ERA-Interim with those derived from COSMIC GPS temperature profiles show that the two are similar for the period when ECMWF assimilates COSMIC GPS data. Differences in distributions derived from ECMWF prior to assimilation of GPS data will be discussed.
We discuss the applicability of ray tracing and linear theory for the interpretation of the observed distributions. Although useful, we explore the limits of the theory and show that this theoretical approach suffers from significant errors associated with the lack of scale separation in the problem.
We find significant annual and inter-annual variation in the observed Kelvin wave distributions related to changes in the large-scale circulation, and more specifically to the Walker circulation. The effect of ENSO on Kelvin wave propagation in the TTL is isolated and described. The effect of the QBO on waves in the lower stratosphere is also discussed.