Serra-Sogas Norma, O'Hara Patrick D, Canessa Rosaline
Department of Geography, University of Victoria, Victoria, BC V8W 3P5, Canada.
Canadian Wildlife Services, c/o Institute of Oceans Science, Sidney, BC V8L 4B2, Canada; Department of Biology, University of Victoria, Victoria, BC V8W 3P5, Canada.
Mar Pollut Bull. 2014 Oct 15;87(1-2):76-87. doi: 10.1016/j.marpolbul.2014.08.009. Epub 2014 Sep 10.
Oily discharges from vessel operations have been documented in Canada's Pacific region by the National Aerial Surveillance Program (NASP) since the early 1990s. We explored a number of regression methods to explain the distribution and counts per grid cell of oily discharges detected from 1998 to 2007 using independent predictor variables, while trying to address the large number of zeros present in the data. Best-fit models indicate that discharges are generally concentrated close to shore typically in association with small harbours, and with major commercial and tourist centers. Oily discharges were also concentrated in Barkley Sound and at the entrance of Juan de Fuca Strait. The identification of important factors associated with discharge patterns, and predicting discharge rates in areas with surveillance effort can be used to inform future surveillance. Model output can also be used as inputs for risk models for existing conditions and as baseline for future scenarios.
自20世纪90年代初以来，加拿大国家航空监视计划（NASP）已记录了加拿大太平洋地区船舶作业产生的油污排放。我们探索了多种回归方法，以使用独立预测变量来解释1998年至2007年检测到的油污排放的分布和每个网格单元的排放数量，同时试图处理数据中大量的零值。最佳拟合模型表明，排放通常集中在靠近海岸的地方，通常与小港口以及主要商业和旅游中心相关。油污排放还集中在巴克利湾和胡安·德富卡海峡入口处。识别与排放模式相关的重要因素，并预测有监视工作的地区的排放率，可用于为未来的监视提供信息。模型输出还可作为现有条件风险模型的输入以及未来情景的基线。
