On the estimation of climatology using harmonics (Invited Presentation)

Ed Epstein introduced a procedure for the calculation of climatology at CPC based on a fit of raw estimates of daily (or 5 day mean) climatology to a set of additive harmonic functions. The procedure has both elements of smoothing and interpolation. This approach has been applied widely since that time. We will discuss how this works, track the roots of this procedure, how it differs from Fourier analysis, how it is applied to both analyses and forecasts to correct systematic error, why some users do NOT like (i.e. list the drawbacks), how one deals with leap-days, and how it can be extended when physically we do not only have additive harmonics but also multiplicative harmonics making up the climatology. The true annual cycle is composed of two components, the first and larger component is due to obliquity (earth's rotation axis is not perpendicular to the ecliptic), the second weaker component to eccentricity of the earth's path around the sun.