Characteristics of Environmental Data Layers for Use in Species Distribution Modelling in the Eastern Canadian Arctic and Sub-Arctic Regions
Species distribution modelling (SDM) is a tool that utilizes the relationship between a species and its environment in known (sampled) locations to predict the species’ distribution in unsampled areas. Environmental data are typically collected at different spatial and temporal scales and often require spatial interpolation between data points to provide a continuous surface required by the modelling application. Here we provide detailed information on 111 environmental data layers collected over different spatial scales and temporal resolutions and interpolated using a geospatial method to provide continuous data surfaces for the Eastern Canadian Arctic and Sub-Arctic. Variables were obtained from a broad range of physical and biological data sources and spatially interpolated using ordinary kriging. For each environmental variable we show the distributional properties of the raw data prior to spatial interpolation, model performance indicators and assessment of model performance, and finally, maps of the prediction standard error and interpolation prediction surfaces. These layers have been archived in a common (raster) format at the Bedford Institute of Oceanography to facilitate future use. A subset of these variables has already been used in a conservation management application to identify deep-water coral and sponge Significant Benthic Areas in the Eastern Canadian Arctic.