The severity of a burn for post-fire ecological effects has been assessed with the composite burn index (CBI) and the differenced Normalized Burn Ratio (dNBR). This study assessed the relationship between these two variables across recently burned areas located in the western Canadian boreal, a region not extensively evaluated in previous studies. Of particular interest was to evaluate the nature of the CBI–dNBR relationship from the perspectives of modelling, the influence of fire behaviour prediction (FBP) fuel type, and how field observations could be incorporated into the burn severity mapping process. A non-linear model form best represented the relationship between these variables for the fires evaluated, and a similar statistical performance was achieved when data from all fires were pooled into a single dataset. Results from this study suggest the potential to develop a single model for application over the western region of the boreal, but further evaluation is necessary. This evaluation could include stratification by FBP fuel type due to study results that document its apparent influence on dNBR values. A new approach for burn severity mapping was introduced by defining severity thresholds through field assessment of CBI, and from which development of new models could be incorporated directly into the mapping process.

The present study undertook a hierarchical analysis of the variability within and among some individual fire events in the boreal ecozones of Canada and Alaska. When stratified by ecozone, differences in the spatial and temporal distribution of wildfires were observed in the Canadian Large Fire Data Base that reflect climatic, terrain and land-use differences across the country. Remote-sensing data collected before and after boreal forest fires permitted a rigorous analysis of the variability in burn severity within individual fire events, and the identification of certain fire-prone and more fire-resistant land-cover types. The occurrence of fire skips or islands was related to the distribution of those cover types, resulting in proportionally more unburned area within the perimeter of a burn for larger fires. Differences in burn severity led to differences in post-burn vegetation response of tree, shrub and moss layers that can persist for decades or even centuries. As a result, there can be considerable variability in the survival, density and distribution of residual biota and organic materials. This variability creates a range of post-fire vegetation patterns and contributes much to the habitat diversity of boreal landscapes.

No abstract

The acceptance of 3D building models is critical to all urban 3D visualization projects. Building models that are identified as unacceptable can increase the cost of the project, delay the delivery, and, in some cases, cancel the acceptance of the entire project. A 3D modelling approach of using representative textures and geometry rather than actual photorealistic textures and geometry was conducted to determine whether participants who frequent the building multiple times a week over a period of a year would be able to identify the visual difference. Three focus groups were established and used to evaluate the visual quality of the 3D building models. Participants were asked to rank the visual quality of the building, as well as identifying any geometry, texture, or overall visual quality problems. The participants from the three focus groups did not identify any texture or geometry mistakes present in the building models. The overall visual quality identified by the participants from the three focus groups was above average, suggesting that the 3D modelling approach is an effective means for modelling buildings with high visitation and significance. RÉSUMÉL'adoption des maquettes d'immeubles 3D est indispensable pour tous les projets de visualisation tridimensionnels urbains. Les maquettes d'immeubles définies comme inacceptables peuvent augmenter les coû ts du projet, retarder sa livraison et même en venir à l'annulation du projet. Une étude d'une approche de modélisation 3D utilisant des textures et des formes représentatives plutôt que photoréalistes, a été menée afin de vérifier si les participants qui fréquentent plusieurs fois par semaine au cours d'une année l'immeuble modélisé sont en mesure de déterminer les différences visuelles. On a demandé à trois groupes d'évaluer la qualité visuelle des maquettes 3D selon un certain classement et d'indiquer tout problème avec les formes, les textures et l'ensemble de la qualité visuelle. Les participants de ces groupes n'ont pu déceler d'erreurs dans les formes ou textures, et ils ont estimé que l'ensemble de la qualité visuelle était supérieur à la moyenne, ce qui suggère que l'approche de modélisation 3D est un moyen efficace de modéliser des immeubles achalandés et de grande importance.