Lorraine Foelske, Bureau County Soil and Water Conservation District
Visitor use in national parks is dynamic and can be challenging to understand. One important aspect of visitor use is to understand where people go. Knowing where visitors travel within a park helps park managers allocate resources more effectively, identify potential environmental impacts such as vegetation trampling and soil erosion, and preserve the experience of other park visitors. Tracking use in remote and expansive wilderness areas is challenging as there are few reliable methods to record detailed use patterns. However, a growing body of research uses Global Positioning System (GPS) technologies to track visitor use in remote and wilderness settings (Gundersen and Andersen 2010, Gundersen et al. 2019, Hallo et al. 2004, Stamberger et al. 2018). In Denali National Park and Preserve, we used GPS units to track backpackers’ locations (Stamberger et al. 2018) as part of a larger project focused on front and backcountry visitor values and pro-environmental behaviors (van Riper et al. 2019). The aim of this paper is to provide a detailed description of the methods of GPS visitor tracking used in Denali in 2016. We point the reader to van Riper and others (2017, 2019, and 2020) and Stamberger and others (2018) for descriptions of the characteristics of backcountry users who participated in GPS tracking.
Backpacking in Denali National Park
Denali National Park and Preserve covers six million acres of subarctic land in the Alaska interior. Denali provides a multitude of recreational experiences for visitors in untrammeled settings within its two million acres of designated wilderness. An especially unique experience for visitors is the fact that most of the park is trailless. In the backcountry, visitors are instructed, in most instances, to avoid using informal (or social) trails and to “Find Your Own Trail.” This slogan is used as a purposeful management strategy to disperse visitor use in fragile tundra ecosystems.
Backcountry visitors to Denali have the freedom to travel within the park. However, the park’s unit quota system affects dispersion of use. Denali is segmented into 87 backcountry units, and the units in the “old park” (Mt McKinley National Park) have specific visitor quotas per night. During peak season (June through August), the quotas are often met and aid in dispersing campsites and foot traffic, especially in popular backcountry units. The 92-mile road into the park also influences travel patterns. The road is often used as a launching point into the backcountry. Due to the immense size and trailless qualities of Denali, understanding the geographic extent of backpackers’ travel is challenging.
Need for Visitor Use Dispersion Tracking
In Denali, managers have expressed concern over a growing network of informal trails and concentrated impacts of camping areas along popular routes. Routes can become popular due to reasons related to topography and vegetation where higher elevations involve less bushwhacking, length of time to get into the heart of the park, recommendations made by staff, and social or other media. The specific motivations for why backpackers choose the route they do is less understood, but general motivations include experiencing pristine nature, quiet and solitude, and adventure (Keller and Toubman 2019). Rising numbers of park visitors over the last decade has increased pressure to both facilitate and preserve quality experiences related to landscapes, species, and sense of place (including wilderness character) that national parks are mandated to protect.
Denali has had an informal trail monitoring program using trail counters, patrols, photographs, and erosion monitoring for over a decade. Additionally, backcountry patrols have a protocol in place to identify and locate impacts within the backcountry, adapted from the problem assessment survey method (Leung and Marion 1999). Currently, park managers primarily rely on this seasonal monitoring effort to understand where backpackers disperse into the backcountry. Also, these occasional backcountry patrols are the only means to enforce dispersion. Thus, little is known about where visitors travel and if dispersed hiking and camping guidelines are being followed. Our study provided the park with season-long spatial data of backpackers’ trip extent, access points, camping locations, and use densities to assist in backcountry patrols and give park managers an overview of common visitor pathways.
Visitor Tracking Technologies
Researchers have used a variety of technologies to track visitor use trends for planning, regulation, and mitigation purposes. Trail counters, for example, are useful for their consistent data collection and, if calibrated regularly, their reliability (D’Antonio et al. 2010). However, trail counters are only as good as an install, and installation sites are determined a priori of where people go. In other words, trail counters do little to capture dispersed and highly variable recreation (Beeco and Hallo 2014) or monitor trends in these contexts. Denali has used passive infrared trail counters for years to establish a broad sense of use on established and known informal trails in the park, but these counters are necessarily rotated among sites and calibrated infrequently.
Starting in the early 2000s, GPS units became widely popular for navigation in wilderness and other recreational settings (Hallo et al. 2004). Human dimensions researchers capitalized on this trend, especially when GPS units shrank to a size that fit into pockets. This tracking technology, unlike passive counters, captures on-the-ground travel patterns and creates detailed and accurate spatial data (Edwards and Griffin 2013, Kidd et al. 2015). More importantly, this technology provides a link between spatial data and social data collected with social science surveys (Stamberger et al. 2018). Due to the richness that spatially linked social data can provide for planners, managers, and practitioners, this study has wide-ranging applicability.
During the peak season of 2016 (June-August), we distributed GPS units to backpackers who agreed to participate in the study. One person in each group was responsible for carrying the GPS unit the entirety of the trip and then dropping it off upon their return. Backpackers were usually responsible for turning on the units themselves. We provided a drop-box outside of the park’s backcountry office to ensure the GPS units could be returned at any hour. If willing, backpackers provided contact information so the researchers could meet them upon their return out of the backcountry to collect the units and administer a follow-up survey. In large parks with dispersed use, this re-intercept method ensures the GPS units are returned. We administered a survey to individuals who took tracking units into the backcountry in order to ascertain attributes such as trip motivation, backpacking experience, and park knowledge. Our sample also included GPS trackers returned from guided hikes in the backcountry (NPS and other educational groups) to compare dispersion, use, and distance between guided and independent hikers.
We used Canmore GT-740 FL units due to their size, design, spatial and temporal accuracy, and battery life (Table 1). The units are small and have a simple design (Figure 1). The Canmore model has good spatial accuracy, detailed temporal data abilities (timestamp intervals), (see for model review White and others 2015), and extended battery life capabilities. These units are able to capture multi-day trips—a key component of this study. From our research, we found the Canmore units lasted approximately three days. If groups planned to be in the backcountry longer than this, they were given multiple GPS units in order to record the entire trip.
|Small unit||Unit is about 2.75 inches (7 cm) long, weighs less than half an ounce, and is similar in size to a USB flash drive.|
|Ease of use||Unit has a simple design with two buttons on the top, one to power on and the other to set manual waypoints. Buttons are stiff to press, making the unit difficult to accidentally power off.|
|Spatial accuracy||Unit is accurate to 8 feet (2.5 meter) CEP (circular error probably).|
|Timestamp interval||Timestamp intervals are set by the user and can be as short as 1 second between recorded waypoints. Note that setting shorter timestamp intervals will decrease battery life. We collected waypoints every 15 seconds.|
|Battery life||Battery life is about 48-72 hours per use (Stamberger et al. 2018).|
Data Cleaning and Management
During the 2016 field season, spatial data were systematically downloaded from the units and cleared for redistribution. On a weekly basis, the tracks were extracted and converted into .csv files. Following the field season, the .csv files were uploaded into ArcGIS 10.4. The spatial point data were then converted from the WGS 1984 to the NAD 1983 Alaska Albers coordinate system. As noted by others (Peterson et al. 2016), extracting the data from these units and into ArcGIS software is a multiple-step process and less streamlined than when using other units. Models that export the tracks as .gpx rather than .csv files may result in a more efficient data management process.
In all, 147 GPS units were distributed to 132 independent backcountry groups, but data cleaning processes trimmed the total number of recorded trips to 113. Three of the units were never returned and the remaining units not included in our sample had no or incomplete data and were therefore discarded. When units were returned with no data, backpackers often forgot to turn on the units at the start of their trip. Trips were deemed incomplete by researchers if the backpacker(s) did not enter the backcountry during any part of their trip. For example, a group may have decided to stay at a designated campsite or take a bus trip into the park.
The tracks in our sample (N=113) were cleaned for a more accurate representation of where and how far backpackers traveled. Three factors guided our decision to cut data points in the cleaning process. The first factor was frontcountry travel (i.e., visitor centers, along road/established trail networks or in tourist areas outside of the park). We assumed backpackers had started their trip in the backcountry when the GPS points diverged from the park road and into backcountry units. Thus, points not located in the backcountry were removed. Second, we removed points that were not physically feasible (e.g., a consecutive point in a 15 second span located several miles away). Third, we removed points that formed a dense cluster when backpacker movement appeared to be stagnant. This final cleaning step reduced the total trip distance for each group by an average of 4.7 miles (7.6 kilometers; t = 11.52, p < 0.01).
After the data were cleaned and prepared for analysis, the routes taken and campsite locations chosen by backpackers were examined using ArcGIS 10.4 software (Table 2). Routes were created using the point-to-line conversion tool, and campsites were delineated by placing a point in the location where backpackers were stationary from one day into the next. The spatial analyses described in Table 2 were driven by several motivations: (1) concentration of use (density, access point distribution, and distribution by backcountry unit), (2) impact on natural resources (land-cover overlay), and (3) adherence to park rules and regulations (viewshed analysis).
|Element||Variable||Research Question||Spatial Analyst Tool(s)||Description of Analysis|
|Route||Density||Where were low- and high-density areas located?||Kernel density||Spatial diffusion of GPS routes was analyzed using kernel density estimation (KDE; Korpilo et al. 2017).|
|Route||Access point distribution||At which point did users access the backcountry? How were the access points distributed?||Create routes,
Locate features along routes,
|The park road access corridor was converted into a route layer and point features were located along the road to capture where backcountry users departed from the park road and moved into the backcountry.|
|Route||Distribution by backcountry unit||How many miles were hiked in each backcountry unit?||Spatial join||The routes (line layer) were spatially joined to a polygon layer of the park’s backcountry units. The sum of miles hiked in each unit was calculated during the join.|
|Campsite||Density||Where were low- and high-density areas located?||Kernel density||The campsite data were analyzed to show concentration of use, particularly “hotspot” locations (Alessa et al. 2008).|
|Campsite||Viewshed analysis||Which campsites were located within the park road viewshed?||Viewshed,
Select by location
|The viewshed tool calculated the raster cells that were visible from the park road (Carver et al. 2012). Campsites within the viewshed were spatially selected and mapped.|
|Campsite||Land-cover overlay||On what types of surfaces did most users camp?||Raster to polygon conversion,
|A Denali land-cover layer, including 23 land-cover classifications, was spatially joined to backcountry campsite locations (Marion and Cole 1996).|
Route and Campsite Characteristics
On average, the backpacking trips lasted about three days, covering approximately 11 miles (17.84 km). The lengthiest trip recorded was ten days long. We found the GPS tracks were concentrated in specific locations of the park (Figure 2). Tracks were densest near the Eielson Visitor Center and the Toklat River Rest Area. There were less dense pockets of GPS tracks at the park entrance and near the end of the road by the Kantishna and Wonder Lake areas. Access points into the backcountry were also highly concentrated as over one third of backpackers (37.5%) accessed the backcountry from two stretches of the 92-mile park road (between miles 50-55 and miles 65-70). Backcountry Unit 13 (Mount Eielson) was most popular with 149.9 miles (241.2 km) hiked in this unit followed by Units 9 (East Toklat, 126.2 mi; 203.1 km) and 10 (West Toklat, 117.6 mi; 189.3 km).
The spatial distribution of campsites (n=203) exhibited a similar pattern to the density of GPS tracks with the majority of backpackers choosing to camp from the middle sections of the park road between Toklat Rest Area and the Eielson Visitor Center. The viewshed analysis (using ESRI’s Viewshed Analysis Tool) showed that about half of the campsites recorded were within view of the park road although NPS staff urge backpackers to camp outside of the road’s view (Figure 3). Lastly, we conducted an overlay between the campsites and land cover. This analysis illustrated that the majority of backpackers camped on vegetated surfaces. Backpackers most commonly camped on a low shrubby land-cover type (41.9%) followed by bare ground (17.7%).
Independent vs. Guided Trip Characteristics
In addition to our GPS tracking dataset, tracks of guided backcountry trips led by NPS staff and other educational groups were independently recorded during the 2016 peak season. We compared the characteristics of these guided trips to the independent backpacking trips we analyzed (Stamberger et al. 2018). Unguided independent travelers spent the most days in the backcountry (mean (M)=2.89, standard deviation (SD)=1.37). For this subgroup, two-day trips were most common. The NPS-led day hikes lasted only one day while guided educational tours averaged 2.30 days (SD=1.36). For the educational trips, two-day trips were also most common, followed by one-day trips with the longest trip lasting six days. Unguided independent travelers not only spent the most time in the backcountry but also traveled the farthest, averaging 11.08 miles. However, mileage was highly variable with the minimum being 0.39 miles and the maximum distance a group traveled being 37.34 miles. NPS-led day hikes averaged 2.67 miles (SD=1.09), and guided educational tours averaged 5.97 miles (SD=7.07).
Understanding the Findings
Despite the strong “Find Your Own Trail” messaging in Denali National Park and Preserve, backpackers’ GPS tracks were concentrated in a few locations. This concentrated pattern of tracks might occur for a few reasons. First, the densest track locations were at relatively high elevations. Hiking at higher elevations in Denali tends to be easier due to drier conditions compared to the boggy and brushy conditions at lower elevations. Second, backpackers were likely pursuing similar scenic vistas. Third, tracks were densest in renowned popular units (that is, people knew to go there; Keller and Toubman 2019). Finally, visitor use may have been more concentrated near established stops along the road, such as the Eielson Visitor Center, so backpackers could use amenities from these sites, such as bathrooms and water before starting their trip.
Tracking visitor use in the backcountry in Denali has been a priority since the adoption of the Denali Backcountry Management Plan in 2006 (NPS 2006). This long track record has shown park managers that use has changed in some areas (e.g., decreased visitation to Sable Pass), but others remain key attractions and have a potential for deterioration. Monitoring studies such as GPS tracking of visitor use in the backcountry are important for managers because they provide season-long (or multi-season) datasets. Our study provides guidance to backcountry staff when administering permits to reinforce the messaging of Find your Own Trail, and camp out-of-sight from the road. Campsite densities from this study furthermore provide direction of where to monitor landscape changes, especially in alpine tundra settings, for impacts related to human use and concentration: key indicators in the Denali Backcountry Management Plan (NPS 2006). This study also provided data comparing dispersion of NPS-led hikes, guided educational hikes, and independent hikers which suggests that these subgroups have very different experiences of Denali’s backcountry regarding time and distance covered.
Evidenced by this research, GPS trackers can be used to understand use patterns in remote and expansive areas where use is often less known. This studied expanded on previous GPS tracking research by capturing use patterns for multi-day trips. The future of GPS tracking has wide-reaching potential as the technology continues to improve and expand. Studies like these will continue to aid in the betterment of managing our public parks and protected areas.
Alessa, L., A. Kliskey, and G. Brown. 2008.
Social-ecological hotspots mapping: A spatial approach for identifying coupled social-ecological space. Landscape and Urban Planning 85: 27-39.
Beeco, J. A. and J. C. Hallo. 2014.
GPS tracking of visitor use: factors influencing visitor spatial behavior on a complex trail system. Journal of Park and Recreation Administration 32(2): 43-61.
Carver, S., A. Comber, R. McMorran, and S. Nutter. 2012.
A GIS model for mapping spatial patterns and distribution of wild land in Scotland. Landscape and Urban Planning 104(3): 395-409.
D’Antonio, A., C. Monz, S. Lawson, P. Newman, D. Pettebone, and A. Courtemanch. 2010.
GPS-based measurements of backcountry visitors in parks and protected areas: Examples of methods and applications from three case studies. Journal of Park and Recreation Administration 28(3): 42-60.
Edwards, D. and T. Griffin. 2013.
Understanding tourists’ spatial behaviour: GPS tracking as an aid to sustainable destination management. Journal of Sustainable Tourism 21(4): 580–595.
Gundersen, V. and O. Andersen. 2010.
Visitor counting and surveys in a dispersed use mountain area in Norway. Recreation, Tourism and Nature in a Changing World. pp. 67.
Gundersen, V., O. I. Vistad, M. Panzacchi. O. Strand, and B. van Moorter. 2019.
Large-scale segregation of tourists and wild reindeer in three Norwegian national parks: Management implications. Tourism Management 75: 22-33.
Hallo, J. C., R. E. Manning, W. Valliere, and M. Budruck. 2004.
A case study comparison of visitor self-reported and GPS recorded travel routes. Proceedings of the 2004 northeastern recreation research symposium (pp. 172–177).
Kidd, A. M., C. Monz, A. D’Antonio, R. E. Manning, N. Reigner, K. A. Goonan, and C. Jacobi. 2015.
The effect of minimum impact education on visitor spatial behavior in parks and protected areas: An experimental investigation using GPS-based tracking. Journal of Environmental Management 162: 53-62.
Keller, R. and J. Toubman. 2019.
Backcountry Management Plan Indicator Reports by Management Area and Backcountry Unit in Denali National Park and Preserve, 2017-2018. National Park Service: DENA.
Korpilo, S., T. Virtanen, and S. Lehvävirta. 2017.
Smartphone GPS tracking—inexpensive and efficient data collection on recreational movement. Landscape and Urban Planning 157: 608–617.
Leung, Y. F. and J. L. Marion. 1999.
Assessing trail conditions in protected areas: Application of a problem assessment method in Great Smoky Mountain National Park, USA. Environmental Conservation 26(4): 270-279.
Marion, J. and D. Cole. 1996.
Spatial and temporal variation in soil and vegetation impacts on campsites. Ecological Applications 6(2): 520–530.
National Park Service, Denali National Park and Preserve. 2006.
Denali National Park and Preserve final backcountry management plan: general management plan amendment, and environmental impact statement. U.S. Department of the Interior, National Park Service, Denali National Park and Preserve. Denali Park, Alaska
Peterson, B., M. Brownlee, and R. Sharp. 2016.
Understanding Visitor Use at Cumberland Island National Seashore. Report delivered to management staff at the Cumberland Island National Seashore. Department of Interior, National Park Service.
Stamberger, L. 2018.
Using GPS units to understand where backpackers travel in the wilderness. Available at: https://sciencetrends.com/using-gps-units-to-understand-where-backpackers-travel-in-the-wilderness/ (accessed April 10, 2020).
Stamberger, L., C. J. van Riper, R. Keller, M. Brownlee, and J. Rose. 2018.
A GPS tracking study of recreationists in an Alaskan protected area. Applied Geography 93: 92-102.
van Riper, C. J., L. Foelske, S. D. Kuwayama, R. Keller, and D. Johnson. 2020.
Understanding the role of local knowledge in the spatial dynamics of social values expressed by stakeholders. Applied Geography 123: 102279.
van Riper, C. J., L. Stamberger, C. Lum, and S. Kuwayama. 2017.
A study of values, environmental behavior, and GPS visitor tracking in Denali National Park and Preserve. Technical report prepared for the National Park Service. University of Illinois.
van Riper, C. J., S. Winkler-Schor, L. Foelske, R. Keller, M. Braito, M., C. Raymond, M. Eriksson, E. Golebie, and D. Johnson. 2019.
Integrating multi-level values and pro-environmental behavior in a US protected area. Sustainability Science 14(5):1395-1408.
White, K., M. Brownlee, N. Furman, and A. Beeco. 2015.
Climber access trail mapping and GPS visitor tracking in Indian Creek. Technical report submitted to the Bureau of Land Management.
Series: Alaska Park Science Volume 20 Issue 1 - Parks as Proving Grounds
Last updated: May 18, 2021