Leroy SS, Dykema JA, Anderson JG (2006) Climate benchmarkingusing GNSS occultation. In: Foelsche et al. (2006b) pp 287–301.doi:10.1007/3-540-34121-8_24
We put climate monitoring in a scientific context, which can be arrived at through a careful implementation of Bayesian inference. What we find is that a good climate monitoring tool must help address the physics of a climate model so as to make it better able to predict future climates. GNSS occultation is ideal because it offers sensitivity to improve the model physics which affects the stratospheric Brewer-Dobson circulation, the tropical tropospheric hydrological cycle, and the poleward migration of the mid-latitude storm track. Also, GNSS occultation is ideal because it can be readily made into a benchmark measurement provided clock calibration is always done by double-differencing, and measurements used to determine precise orbits and information on ionospheric activity are archived as auxiliary information. In doing so, GNSS occultation can be made S.I. traceable.