Methodological & Operations Problems
Despite the success of this research in determining an appropriate location for a nuclear power facility in northern British Columbia, there were several shortcomings and difficulties conducting this analysis. In order to get a thorough understanding of the geographical characteristics of northern British Columbia, four regional districts were assessed. The scale of this project worked to examine a relatively large area of the province with the smallest of the four districts being over 50,000km² in area. As a result, storage size issues arose and the spatial resolution of 200 metres was used in order to maintain analysis quality. Consequently, because every subsequent dataset produced was of a similar size, large storage spaces were required and running cartographic models in IDRISI and ArcGIS were time consuming. This was particularly the case running the cost distance analyses comparing proposed nuclear sites with friction surfaces which, at this scale, required more than an hour to complete for each of the four cost surfaces created. In order to simplify the analysis, another approach might have been to reduce the area of examination to focus on one particular region district or a portion of that area. Considering the fact that the two locations of highest suitability were located in the Peace River district, further research could be carried out on this district at a smaller spatial resolution in order to determine similarities and differences between scales.
The data collection strategy employed for this project incorporated both methods of determining a study design and question based on a certain area and determining which forms of criteria analysis would be performed based on the existence of available data sources. As a result of time constraints, the majority of this study focused on creating new information and deriving spatial relationships based on existing secondary data sources, supplemented by the digitizing of key niche factors such as bison habitat zones. Even though major datasets used for this assessment were based on already existing sources, a significant amount of time was spent determining the viability of datasets for this examination and editing data to meet the needs of this research. An example of this is the addition of additional provincial parks to the available dataset to enhance the accuracy and completeness of this criterion.
Errors and module failures were common when performing spatial operations especially in IDRISI. Fortunately, operational problems were not experienced when running complicated, time consuming modules such as the cost distance analyses, however, they were frequent when converting files between ArcGIS and IDRISI and when changing between vector and raster file structures.
One issue that arose during the analysis stage of this project was dealing with the ‘fuzzy’ characteristics of strict boundaries. When clipping national and provincial scale data to include only those entities located within the four regional districts a finite boundary was accepted. However, because of the need to create a raster tessellation to conduct analysis, the geographical coverage area was expanded to include areas outside of the regional district study areas. When creating the water bodies constraint image, a 5 kilometre buffer complemented this by expanding the study area to include regions outside of the regional district limits. A consequence of this can be seen with the high suitability rating of a region to the south of Valemount along Kinbasket Lake in the extreme southeast corner of the regional district of Fraser-Fort George. Unfortunately, this area of suitability is actually within the lake, but is not recognized by the GIS because the Kinbasket Lake is separated into both the Fraser-Fort George and Columbia-Shuswap. This location was, therfore, discarded as a possible nuclear power plant location. Further research must be carried out in the future to evaluate the suitability of southern sections of this lake as a possible site location based on analyses completed outside of the Fraser-Fort George regional district boundary.
Despite committing large amounts of time to planning cartographic models, operation methodology often had to be changed spontaneously and many issues arose unexpectedly requiring more time to resolve. This was particularly true when determining constraints and factors, where the use of certain criteria was switched back and forth from between boolean images and weighted factors. In the case of viable water bodies, all rivers examined were assumed to have enough year round water volume to sufficiently supply a nuclear reactor. In the future, further on site hydrological research would have to be performed on each river on a case-by-case basis to determine viability. Fortunately, each of the four possible station locations utilized lakes for water sources with the Charlie Lake site containing sufficient area to supply the plant.
In order to produce a more complete assessment, future research may choose to incorporate additional criteria such as the inclusion of not only First Nations reserves, but also current land claims or disputed aboriginal land in analyses. This additional information could have been used to augment the third WLC multi-criteria to increase emphasis on more than one dataset dealing with nuclear power plant location from the First Nations perspective.