The Impact of Nonsynoptic Sampling on Mesoscale Oceanographic Surveys with Towed Instruments

Abstract

Mesoscale ocean features, such as eddies and frontal meanders, have temporal and spatial scales that are poorly resolved by conventional surveying methods such as moorings and CTD casts. Towed instruments provide a method of measuring oceanographic data with a medium spatiotemporal resolution. The separation of sample points is restricted by the towing speed of the instrument and the survey design. The compromise between imaging large-scale and mesoscale features in towed surveys often means that the mesocale components are only just resolved by the cross-track sample spacing. In the analysis of the data, the evolution of the system during the survey period is often ignored, and it is assumed that the measurements are synoptic. A simple isopycnal model is used to examine the errors that arise as a result of this assumption for a typical towed survey consisting of a series of quasi-parallel tracks measured perpendicular to the mean flow. The temporal changes in the model lead to a 40%?50% increase in the rms errors compared to those due to the spatial separation of the sample points. These errors are unevenly distributed across the region and result in a distortion of the mesoscale features. The sample spacing along track is several times smaller than that across track, which is of the same order of magnitude as the eddy diameter. A significant decrease in the accuracy of the interpolated data is seen when the along-track sample spacing exceeds one-third of the cross-track spacing.