No abstract

A problem which has challenged oceanographers for decades has been that of determining the factors that control the recruitment of young fish into the stock of a species such as Atlantic cod (Gadus morhua). As a very small number of the eggs laid hatch and survive to market size (typically one in millions), small variations in mortality lead to large variations in recruitment. While there has been some effort obtain sample the population at various stages of development through net tows, etc. they are essentially random samples. However, to sample the Ocean more effectively, we must understand how the system will evolve.Thc challcnge is to combine all thc synoptic data available to describe the current state of the ocean (nowcast) and to predict the changes over h e next few days (forecast).The Occan Production Enhancement Network (OPEN) was funded recently by the Canadian government in 1000 under the National Centres of Excellence program. One of the research projects sponsored by OPEN will attempt to follow discrete patches of cod larvae as they develop from hatching to settlement over the period of a few weeks. Success in following a group of larvae over that time scale makes possible extensive observalions of the physical and biological environment and the changes in genetic characteristics of the group. This information will help to show which factors are most significant in determining which larvae survive to the next stage of life.The operational problem is to successfully follow a group of larvae over that period as they are advected by thc ocean currents. The approach implemented by the physical oceanography group is a data assimlative model of the area. The model makes use of tidal and meteorlogical predictions along with real-time density and current measurements to predict the new location of the patch.The model uses the adjoint assimilation approach to absorb the data which flows in from a shipboard Acoustic Doppler Current Profiler (ADCP), moored telemetering Current meters, surface dnfting buoys, tirogued at an appropriate depth, towed and profiling Conductivity, Temperature, Depth (CTD) probes and the larval counts from net tows. The data is acquired into a database, based on the Empress2 database managcment system running on a Unix workstations. The model searches the data b a e for the data (filed in spatial and time coordinates) required to update its predictions. This papcr will provide an overview of the Ocean Probe system, including a discussion of the data streams, d;ita handling and modelling, based on mals in April 1992. The system will also be of interest to groups concerned with tracking effluent or oil-spills as it proposes solutions to the general problem of tracking patches of water at middle depths on the continental shelf.