Evaluating a model's strengths and limitations is essential to ensure its application is effective and representative of a spatial problem. Through extensive review of relevant academic literature, this study reflects an accurate portrayal of the major criteria involved in wolf den site selection. Additionally, it incorporates existing knowledge of wildfire risk through hazardous forest types, making a weighted MCE an ideal model to accomplish the research objectives described. Developing a multi-criteria evaluation for the study promotes the analysis of how variables interact with each other in meaningful ways and this information is utilized to generate visual aids predicting areas of viable habitat with imposed wildfire risks.
A weighted MCE is chosen for the initial spatial analysis. A MCE is an effective method of incorporating many different layers of varying importance. The layers are standardized prior to comparison, strengthening the consistency of the study and ensuring the results accurately demonstrate identified relationships. After standardizing, numerous layers are analyzed and compared to observe the relative importance of each. MCE models are widely used to approach geospatial questions and are thus widely understood across many fields of study. Additionally, MCE models are easy to comprehend, especially for professionals who lack formal training in geographic information systems.
Several limitations affect the resulting MCE model. While a plethora of data exists for Algonquin wolves and Algonquin Park land use, fire data are more difficult to locate. The data layer representing fire risk is generalized data for the identification of potentially hazardous forest types rather than a forest fire risk layer specific to Algonquin Park. This reduces the accuracy of the evaluation of the fire risk. Restrictions are also observed when processing raw downloaded data into suitable data for the study. To make layers compatible and therefore comparable, they are standardized. The data is standardized in a raster format. All files that undergo conversion to raster lose some accuracy in the transition. Lastly, some data is older than others, which presents inconsistency with the ground truth.
For the purposes of this study, the limitations discussed are presumed to not significantly affect the output results. Human error and data limitations are unavoidable and common and are most likely not significant enough to discredit the results of this study. Due to the nature of this project, it is difficult to evaluate the model using an existing model as a reference. This project is very specific, with extensive searches for similar research projects yielding no comparable models that could be used to determine if the results are reasonable. Therefore, Google Maps is used to visually examine suitable den sites given by the model. Knowing that wolves prefer den sites based on proximity to wetlands, water, conifer forests, tertiary roads, and slopes (Benson and Patterson 2015), the study area outlined in Figure 1 can be visually inspected on Google Earth to determine if the conditions surrounding an area adhere to what was determined to be a suitable den site for wolves. While this method is very susceptible to human error, it is a fairly reliable method of visually determining if the model created was successful in applying the outlined parameters within the GIS analysis.
