Statistical Models for the Optimal Estimation of Oceanic Fields

摘要

An investigation is being conducted into the drift and leeway of survivors and survivor crafts as part of the improvement of probability of detection in search and rescue. Part of this effort uses freely drifting buoys that transmit their positions to a shore/ship-based receiver to provide an estimate of the surface current field. The data sets are in the form of drift tracks from each drifter. To analyze the irregular spaced drifter tracks, objective analysis techniques were applied to produce optimal estimations of surface current fields on a regularly spaced grid. Transformation of a data set onto a regularly spaced grid allows use of many other standard computer analysis programs. A computer program using this technique was previously successfully applied to large oceanographic data sets of many drifter tracks. The objective analysis program was converted for use on Hewlett Packard microcomputers and applied to a much more limited data set from 6 drifters, the results being that objective analysis can effectively work with small data sets. The effect of very limited data sets on the outcome of the estimated fields was checked by removing drifters, one at a time, from the data set of 6 drifters. The fields generated by the reduced number of drifters were then compared to the results with the 'best estimate' field determined by all 6 drifters. The results from this exercise are that placement of the buoys or drifters relative to the flow field features greatly influences how well the total flow field is estimated. A few well placed buoys provide better information than many poorly placed buoys. Therefore, use of remotely sensed data will aid in determination of the optimal placement of buoys. Ocean currents; Search and rescue. (EDC)