Lab 8: Non-Boolean Multi-Criteria Evaluation

1. What is the advantage of WLC over Boolean MCE? Provide an example. [3 marks]

WLC is Weighted Linear Combination is an aggregation procedure that allows us to retain the variability from continuous factors and allows for trade-offs to occur within the different factors (Eastman, 2015).

Boolean MCE is the constraints that provide you with areas that are either suitable or unsuitable.

The advantage of WLC over Boolean MCE is that the hard constraints are not used, so instead of the area being either suitable or unsuitable, there is a “fuzzy” distance that buffers the feature to allow for analyzation that shows the risks of building in different areas dependent on the risk (Eastman, 2015). An example of this could be the best location for a park. For Boolean MCE, areas would be seen as either unsuitable or suitable (reclassifying as a 0 or 1), but with WLC, areas would be shown with the best place for the park and then a buffer with decreasing values as you get further from the best locations (reclassifying as 0-1.0). This does not mean that it is unsuitable, rather it just shows it is less suitable and may have more risks or costs associated with building in those areas.

2. What is meant by ‘fuzzy standardization of continuous factors’? Provide an example. [2 marks]

Fuzzy standardization of continuous factors means that there is a continuous buffer that ranges from 0-1.0 that allows for suitability in each are, just some are more suitable than others, instead of hard factors of unsuitable and suitable (Eastman, 2015). A good example of this is finding a house within relative distance to a Skytrain station. Instead of choosing parcels of land that either are unavailable or available, this will create an area outward from the Skytrain stations starting with a 1.0 near the stations and decreasing as you get further away from the stations, to show lower suitability.

3. How does rescaling differ between categorical and continuous data? [2 marks]

Rescaling in categorical data is done with specific values, it is rescaled with a certain number or type (based on whether it is suitable or not) and cannot be continuous. In contrast, rescaling with continuous data allows us to rescale the range to continuous measures between 0 and 1.0. This means that there is a continuous increase or decrease that allows us to search a scale to compare and combine areas based on their degree of suitability.

4. With reference to a dataset you will use in your project, what is a more appropriate method of rescaling: WLC or OWA? Explain why. Refer to the first page of tutorial 2-9 for an understanding of OWA. [3 marks]

WLC allows for full tradeoff among all factors with moderate risk, meaning that we can assign specific weights to factors that will influence the decision making process more than others. The order is not important for this and it is just weighing one factor against another and giving it a specific weight.

OWA is Ordered Weighted Averaging which allow us to control the amount of risk and tradeoff that we wish to include in the result using rank-ordered positions of the factors (Eastman, 2015). This works the same as WLC, but has a second set of weights added to it meaning that we can control the tradeoff between factors and the level of risk associated with the suitability decision. This means that you choose which factor is the most important and rank them accordingly, as well as assigning weights.

For this project I am comparing the distance from industrial centres, proximity to residential areas and roads, and the distance from waterbodies. Therefore, WLC would be more appropriate as there are not ones that are necessarily more important to have rank wise, but they all have factors that can be traded off and compromised for if the location is right. This way I can give them different weights so that they still influence more than others, but are not ranked as definitely more important with no compromise.

5. Produce a professional location analysis product. This will take the form of a one-pager for a client who is looking in Westborough for suitable locations to develop one of the following facilities (select only one): hospital; day care center; helicopter tour station; open pit gravel mine; café.

Conduct a non-Boolean MCE for your selected scenario. Use only the files in this tutorial. You must customize the weights, use different weighting methods, and include a table describing each input data set, including the weighting method, weights, and why you chose them. Finally, reclass your final suitability map into five categories: unsuitable, low, medium, high, excellent.

Using Microsoft PowerPoint, compose a professional-grade, engaging, and concise one-page summary sheet that describes the project, the parameters, and results. [15 marks]

Works Cited

Eastman, Ronald J. TerrSet. Version 18.31. Worcester, MA: Clark University, 1987-2017