Sampling Strategies in Space and Time: Effects on Representation of Mesoscale Processes

摘要

This thesis examines along-track spacing increments from model output to determine the largest separation of data sampling permissible to achieve a reasonable representation of ocean mesoscale features without serious aliasing. With this objective, three sampling strategies are tested and analyzed by conducting simulated aircraft and shipboard cruises. Due to the limited number of observations acquired during each cruise, analyses of the data fields utilize an objective analysis technique which assesses the spatial correlation and RSM error by comparing the sampled data sets to the assumed perfect PE output fields. Of the schemes tested, the sampling strategy of using finer sampling along the coast and random sampling offshore with either aircraft of less than five-day hydrographic surveys results in the 'best' representation of mesoscale processes in a coastal region.