The objective of the workshop was to provide participants with a basic understanding of geostatistical concepts and methods, as well as the tools to apply this knowledge to fisheries stock assessment using freeware mapping and geostatistical software.The impetus for the workshop was an unfamiliarity of many Scientific Council members with a Voronoi analysis of yellowtail flounder distribution on the Grand Banks.
Lectures pertaining to geostatistical concepts and methods were presented and followed by hands-on exercises that incorporated survey and fisheries data from the Grand Banks. Data visualization and analysis techniques were presented using the GIS (Geographical Information System) software packages of SPANS, ACON and Ocean Data View. Participants viewed demonstrations of the software and completed hands-on exercises. Geostatistical concepts and methods, particularly related to kriging, were also presented and participants learned the basics of using the “R” software package (freeware) and conducted geostatistical analyses using routines prepared by programmers at the Centre de Géostatistique as well as those developed during the workshop.
As a result of the workshop, Scientific Council members now have a better understanding of basic geostatistical concepts and methods; particularly kriging. In addition, the workshop provided members with freeware software tools and hands-on exercises that can be applied to geo-referenced data from the NAFO region in the future.
Participants discussed how they might apply the knowledge gained at the workshop. A recommendation was made to investigate institutionalization of the mapping and geostatistical analyses presented by accessing them via an internet site, such as the GMBIS website, which already provides a mechanism for mapping any geo-referenced data set. However, it was also noted that the time commitment and programming knowledge required represent potential obstacles to implementation. Multiple participants expressed positive feedback about the workshop, particularly with respect to covering such a complex topic, in a short time span, in an understandable way and with the use of data from the Grand Banks in the vicinity of the NRA. Participants also noted that they felt that the knowledge gained at the workshop would now allow them to apply geostatistical analyses to their stock assessments.
An example of the output produced at the WS shows how kriging can be applied to bottom trawl survey data from the Grand Banks. An experimental variogram (left) was computed for all directions and a variogram model (red line) was calculated, for yellowtail flounder abundance (number per tow) for the spring of 2000. The variogram and model are based on the habitat area. A kriging map (right) based on the variogram model is shown with iso-density contours and the 200 m and 400 m isobaths are shown in green.
The report of the Workshop is now published as Scientific Council Studies, No. 39