Factors must be determined to demonstrate suitable area for the site of a UAV launch port and access road. To determine the best site for the UAV launch port, a multi-criteria evaluation (MCE) is performed using sets of criteria and constraint factors. Constraints employ a Boolean pass or fail function to define area as either suitable or unsuitable. Criteria are factors ranked along a scale that are then assigned weights based on a pairwise comparison matrix to show varying levels of suitability (Latinopoulos & Kechagia, 2015). To find the ideal placement for an access road, a least-cost pathway (LCP) analysis is performed where areas are assigned a cost to traverse based on suitability. The variables of UAV port size, protected areas, wetlands, forested areas, lakes and rivers, urban areas, slope, distance to preexisting infrastructure, distance to Thunder Bay, and distance to remote communities will be employed in both the MCE and the LCP.

Required UAV Port Size

    Based on airport zoning recommendations and size of another small airport in Canada, it is determined the ideal site size should be no less than 688 hectares (Prince Rupert Airport, 2016; Government of Canada, 2018). This will allow for adequate facilities for the storage of goods and maintenance of UAVs while adhering to zoning regulations set out by the Government of Canada for airports and aerodromes.

Slope

A steep slope causes high costs of construction to level the area for an airport or, in this case, a UAV launch station and so slope will be a constraint with areas with slopes higher than 15 degrees set as exclusion zones (Latinopoulos & Kechagia, 2015). As well, traversing a steep slope takes significantly more effort relative to performing the same action over a level-surface (Bell et al., 2002). Thus, slope must be factored into the analysis of the least-cost pathway for a transportation route to the UAV launch station.

Land Cover Factors

Protected Areas

    When setting the UAV launch station, it’s important to avoid disturbing fragile environmental systems. The maintenance of large protected areas as part of naturally functioning ecosystems largely free from anthropogenic influence is identified as an important variable in maintaining biodiversity and ecosystem integrity (Powers et al., 2013). Therefore, any area classified as a protected area will be set as a constraint since it is important to avoid building a site in protected areas as it would cause significant disturbance combined with the required road, resulting in habitat fragmentation, destruction, and degradation (Powers et al., 2013; Ogden, 2012).

Wetlands

Similarly, wetlands provide important ecosystem services that benefit the surrounding environment such as maintaining regional and national diversity, cycling nutrients, attenuating floodwater, and recharging groundwater (Gleason et al., 2011). This type of land cover will be set as a constraint and excluded from suitable area for development of a UAV port and access road since it is important that these areas remain minimally disturbed.

Forested Areas

Forests, even those in urban areas, provide important economic, social, and ecological functions (McWilliam, et al., 2012). As well there are hazards involved with constructing a site for an aviation vehicle in heavily treed areas (Government of Canada, 2018). Considering that forests are prominent throughout Northern Ontario it is unlikely that forests could be simply used as exclusion zones so this land cover classification was considered a criterion. For both the MCE and the LCP analysis, heavily treed areas are given a value that will encourage their avoidance in favor of more suitable land cover for construction and sparsely treed areas are considered highly suitable based on the importance of forest cover for species diversity and ecosystem integrity (Richmond et al., 2015).

Lakes and Rivers

The selection of these factors stems from the fact that you cannot build a launch station in a lake as it is economically unfeasible (Latinpoulos & Kechagia, 2015). Moreover, the site nor its access road should be built in a way that disrupts river connectivity as river management and restoration efforts are becoming increasingly centered on connectivity as an integral component of habitat quality (Blanton & Marcus, 2010). As a result, lakes and rivers were used as constraints and set as exclusion zones for the MCE and given a high cost to traverse in the LCP.

Urban Areas

Due to the noise of the UAV launch site, its preferable to avoid building in urban areas and should account for an 1000m buffer zone as an exclusion zone (Ballis, 2003; Latinopoulos & Kechagia, 2015). As well, roads should be a constraint for the MCE considering that the port cannot be built directly on a road or in a way that interferes with preexisting infrastructure. 

Land cover factors are all combined into a final land cover criteria map.

Proximity Factors

Distance to Preexisting Transportation Infrastructure and Thunder Bay

 Since the purpose of this research is to determine the ideal site for a UAV port, in order to provide greater access to transported goods for remote communities in North-Western Ontario, it is important that the launch site is in close proximity to a city with adequate access to the resources and preexisting transportation infrastructure such as those listed in Table 4.2. As a result, Thunder Bay was chosen as proximity to the Trans-Canada Highway, Canadian National Railway and Port of Thunder Bay all make this city accessible ("RAC Canadian Rail Atlas", 2018; "The Port of Thunder Bay", 2018). Therefore, proximity to roads as well as proximity to Thunder Bay are both important criteria to consider for the ultimate economical transportation of goods to the launch site (Ballis, 2003; Charabi & Gastli, 2011). 

Distance from Remote Communities       

The purpose of this study is to help remote northern communities in Ontario who have poor access to food due to extreme transportation costs (Veerarghavan et al., 2016; Allini, 2017). Thus, proximity to remote communities must be considered as criteria. 

Proximity factors are all displayed on individual criteria maps.