FY 1993

Temporal sampling requirements for low frequency temperature variability in the eastern equatorial Pacific Ocean

Hayes, S.P., and M.J. McPhaden

NOAA Tech. Memo. ERL PMEL-96, NTIS: PB93-136422, 17 pp (1992)

The effects of temporal subsampling of the near equatorial thermal field are examined using Monte-Carlo techniques. Approximately 3 years of continuous data from moored temperature sensors on the equator at 110°W and 140°W in the Pacific Ocean were randomly subsampled at frequencies of 1, 2, and 5 samples per month. Spectra of subsampled and original series demonstrate that the errors caused by subsampling are position, depth, and frequency dependent. In general, for a given number of samples errors are larger at the eastern location and in the thermocline. Adequate (spectral agreement to within a factor of two) resolution for periods longer than 3 months can be obtained at most depths with five samples per month. Fewer measurements are required near the surface.