Fractal Patterns in Coastal Detections on Approaches to Canada


  1. Liu, M.J
  2. Dobias, P.
  3. Eisler, C.
Corporate Authors
Defence Research and Development Canada, Centre for Operational Research and Analysis, Ottawa ON (CAN)
Understanding spatial and temporal patterns in shipping traffic and ship movements on the approaches to Canada will enhance the ability of analysts to identify deviations from expected behaviour. The present paper expands on an earlier analysis of ship detections in Canadian coastal regions using RADARSAT-2 (RS2) and Satellite Automated Identification System (S-AIS) data [1]. It addresses several limitations of the previous approach, and expands the analysis by looking at the properties of sub-regions, in addition to the overall coastal regions. The S-AIS is dependent on ships’ self-reporting (which is mandated for cargo vessels over 300 gross tonnage and for all passenger-carrying vessels), while RS2 collects images over time using Synthetic Aperture Radar (SAR) thus limiting the need for the cooperation from tracked vessels. Employment of spatial entropy and the fractal dimension, calculated using the differential boxcounting method, as measures of randomness in the spatio-temporal distributions of ship detections are presented in the paper; both quantities suggest a presence of non-random patterns of behaviour with annual changes possibly attributable to known causes (such as seasonal changes in shipping routes and fishing). Further work will include exploring additional measures, and the use of the existing measures to determine likelihood maps for expected behaviour for different vessel types.
RADARSAT-2;automatic identification system;surveillance;geo-spatial analysis;fractal dimension;spatial entropy
Report Number
DRDC-RDDC-2015-P138 — External Literature
Date of publication
10 Dec 2015
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: