rimary data-gathering involved direct user input and fieldwork.

For the Pukaskwa test site
this primary data consisted of discussions with park staff to gain information about management
policies for controlling access along roads in the park, water access, and the use of coastal hiking
trails.77 For the Yoho site, the process of primary data-gathering followed the same format. Park staff
was questioned about management practices in relation to controlling access along the rivers in the
park and recreational activities on lands managed by the B.C. Ministry of Forests. This Ministry also
provided additional information concerning their ROS-based land classification system for managing
recreation on B.C. Crown lands.78 As well, the Yoho case study consulted the Atlas of Central Rockies
Ecosystems and this book provided information that helped in the determination of encounter norms
for trails and roads. Encounter norms are simply the likelihood of encountering other people along a
trail or within a wilderness area. In wilderness areas the encounter norm was quite low while more
urban regions have a high likelihood of encounter; encounter norms relate to the potential or usage of
the area by tourists.79 As with any type of modelling in GIS there is a subjective element that must be
addressed. In the creation of this Recreation Opportunity Spectrum model for national parks the
creators (the team of stakeholders) involved in the development process had to make make some basic
assumptions about their data. These assumptions made in the development stage affect the output,
however they are necessary to create such a complex and diverse model. Since the models are being
applied to national parks in totally different provinces, Ontario and British Columbia, the researchers
had to generalize some data in order to make the two models more compatible, this involved making
three assumptions about the data.
To make the Yoho model match its counterpart in Ontario (Pukaskwa) developers had to
assume that the terrain found in Yoho National Park in B.C. was relatively flat, however this was not
the case as it is quite rugged especially in the Kicking Horse Pass region of the park. Secondly, in
terms of defining soundsheds within Yoho this process was done by ear. In reality, noise in Yoho
would have effects up the mountainside and this would have the effect of changing potential
opportunities for tourists in the area as sound would travel farther, disrupting the peacefulness of the
80
area. Lastly, assumptions were made for estimates of user densities within Pukaskwa and Yoho
National Parks. The methods of gathering this information differed greatly in both parks with traffic
data being used in Yoho National Park and backcountry permits were used for Pukaskwa National
Park. Neither source of data was optimal according to Angus Carr, an author of the models.81 For
Pukaskwa and Yoho the respective park management plans were digitized, as well, there was a digital
database created for the park and adjacent areas containing roads, watercourses, settlements, lakes, and
park boundaries.82

Original data for the Pukaskwa model came from their existing national park Geographic Information

System. Mapping of varying scales already in existence were also used to provide a useable base for

the models, specifically this mapping included the Ontario Basic Mapping (OBM) at scales of 1:

20,000 and 1: 50,000 along with the national 1: 250,000 NTS map sheets for the regions where these

83

parks were located.