Measurements of deep currents in the central North Pacific

Abstract

Two deep arrays of current meters at heights of 1200 and 100 m above the bottom were set under the North Pacific Subtropical Gyre near 30°30'N, 157°45'W; the maximum record length obtained was 19 months. Autospectra show the following characteristics: strong peaks at the M2, S2 tidal frequencies; a broad peak at the diurnal‐inertial peak; a spectral gap centered at 0.02 c h−1; and a regular increase in energy with decreasing frequency below the gap. Rotary components tend to be counterclockwise below 10−2 c h−1 and clockwise above the diurnal‐inertial frequency. The M2 tidal amplitude ranged between 1.0 and 1.6 cm s−1; phase relationships suggest that a significant part of the motion is due to internal wave motion at this frequency. Low‐pass filtered velocity statistics show the following: evidence of nonstationarity in the variances between the first (9 month) and second (10 month) arrays; significant horizontal variation of mean kinetic energy on scales of 100 km; and time‐space averaged eddy kinetic energy is comparable in magnitude to the lowest values measured under the Subtropical Gyre in the western North Atlantic. Currents at 1200 m tended to be greater than at 100 m above the bottom. Spectra of the 100‐m records indicate the presence of two peaks at low frequency: a dominant and ubiquitous peak in the 1/105–1/175 c d−l frequency band; and a secondary peak, which is shown in three of four moorings, in the 1/58–1/75 c d−1 frequency band. Point estimates of the periods of the peaks give average values of 154 and 67 days. Comparison with subtropical western North Atlantic spectra shows that the ‘temporal mesoscale’ period (∼150 d) is somewhat higher in the Pacific data and it does not show the dominance of meridional motions observed in the Atlantic spectra. Plane wave fits to the 154‐day oscillation show the longest wave which closely fits the data has a wave length of 170 km and propagates toward 197°T. The westward phase propagation and particle orbits, which are transverse to the direction of propagation, are consistent with planetary Rossby wave dynamics. This wave does not fit the dispersion relation for flat‐bottom planetary waves; the observed phase speed is greater than that predicted by the dispersion relation. It is shown that either Doppler shifting of the frequency by the mean flow or topographic effects might account for the discrepancy.