《Case_Studies_of_Simulation.pdf》由会员分享，可在线阅读，更多相关《Case_Studies_of_Simulation.pdf（54页珍藏版）》请在三一文库上搜索。
1、USDA Forest Service Gen. Tech. Rep. RMRS-GTR-143. 200517 Computer simulation models can be usefully ap- plied to many different outdoor recreation situations. Model outputs can also be used for a wide variety of planning and management purposes. The intent of this chapter is to use a collection of 1
2、2 case studies to illustrate how simulation models have been used in a wide range of recreation situations and for diverse planning and management applications. The types of recreation situations included in these case studies vary in the size and remoteness of the area being modeled, as well as in
3、the type of recreation. Case studies include: large backcountry areas: Bighorn Crags in the Frank Church River of No Return Wilderness, Misty Fjords National Monument, Colorado River through Grand Canyon National Park, Humphreys Basin in the John Muir Wilder- ness, and Isle Royale National Park; sma
4、ller frontcountry areas: Yosemite Valley in Yosemite National Park, Alcatraz Island in Golden Gate National Recreation Area, Arches National Park, Acadia National Park, Para- dise Meadows in Mount Rainier National Park, and the Twelve Apostles in Port Campbell National Park, Australia; overnight hik
5、ers: Bighorn Crags, Humphreys Basin, and Isle Royale; day hikers on trails: Yosemite, Arches, Acadia, and Mount Rainier; visitors to facilities: Alcatraz Island and the Twelve Apostles; bicyclers: Acadia National Park; whitewater rafters: Grand Canyon National Park; visitors on cruise ships and ocea
6、n kayakers: Misty Fjords; visitors in automobiles: Acadia. Perhaps the most basic use of computer simulation modeling is as a tool for describing current use pat- terns. The purpose of the first four case studies (Big- horn Crags, Misty Fjords, Grand Canyon, and John Muir) is to illustrate this use.
7、 Simulation can also be Chapter 4: Case Studies of Simulation Models of Recreation Use David N. Cole used to monitor crowding-related indicators, such as number of encounters, persons-at-one-time or per- sons-per-viewscape, either to describe the current situation or to determine whether standards f
8、or such indicators are being violated. The Bighorn Crags and John Muir case studies illustrate the estimation of encounter rates, while persons-at-one-time or persons- per-viewscape measures are estimated in the Yosemite, Alcatraz Island, Arches and Acadia case studies. Computer models can also be u
9、sed for predictive purposes. In the Isle Royale case study, simulation was used to help planners identify alternative man- agement actions that would be effective in reducing campsite sharing. This was accomplished by predict- ing the effects on campsite sharing of reductions in amount of use, chang
10、es in the spatial and temporal distribution of use and increases in behavioral restric- tions and number of facilities. At Yosemite, Alcatraz Island, and Arches, simulation was used to predict the maximum amount of use that can be sustained without violating crowding-related standards. At Alcatraz a
11、nd Arches, the effect of alternative public transportation systems on maximum allowable use was also predicted. Simulation can also be used to prepare for the future. In the Acadia and Twelve Apostles case stud- ies, use is steadily increasing, and future use levels are forecast to be much higher th
12、an they are presently. In these case studies, simulation models are used to predict the effect of increased use on crowding-related variables. In the Twelve Apostles case study, the effects of changes in park infrastructure are also predicted. Finally, the Mt. Rainier case study shows an innovative
13、way to collect data on visitors in a challenging situation. The case studies included in this chapter suggest that there is reason to be enthusiastic about the potential of computer simulation modeling as a visitor management tool. However, this enthusiasm must be tempered with appropriate realism a
14、nd caution. The application of simulation to outdoor recreation issues is still in its infancy, and there is much need for more learning and development. As with so many things, 18USDA Forest Service Gen. Tech. Rep. RMRS-GTR-143. 2005 the more we learn, the more we recognize the need to know even mo
15、re. The models presented in the case studies have only been partially validated. Simulation outputs have not been statistically compared to real- world observations. Numerical estimates and predic- tions are presented without numerical estimates of how much confidence one should have in these metric
16、s. More work is needed to ensure that the models devel- oped are as valid as possible, simulation runs are conducted correctly and outputs are appropriately interpreted. However, these case studies demonstrate the useful outputs that valid models can produce. Recreation Visitation and Impacts in the
17、 Bighorn Crags Portion of the Frank ChurchRiver of No Return Wilderness _ Randy Gimblett Suzanne Cable David N. Cole Robert M. Itami Purpose This case study demonstrates the use of agent-based modeling and simulation to describe recreation use patterns in a popular portion of the Frank Church River
18、of No Return Wilderness in central Idaho. The case study will demonstrate the data collection meth- ods, modeling, and simulation of backpackers and rec- reational stock users on multiple day trips. Particular attention is given to estimating encounter rates in the interior of this wilderness becaus
19、e encounter rates can affect the experience of wilderness visitors. Conse- quently, wilderness managers commonly want to moni- tor encounter rates and frequently develop standards for maximum acceptable number of encounters. Study Area The Frank ChurchRiver of No Return Wilderness Area is the larges
20、t contiguous wilderness area in the United States, outside of Alaska. The most popular portion of this wilderness for backpackers is an area known as the Bighorn Crags. The area also receives substantial use by groups traveling with pack and saddle stock. Rugged and remote, this country offers adven
21、ture, solitude, and breathtaking scenery. Like other popular wilderness areas, physical impacts from dispersed visi- tor use are evident throughout the area, and social impacts to visitor experiences are likely, but currently not well documented. To develop the information needed to better manage re
22、creation in the Bighorn Crags, we conducted inventories of all recreation impacts in the Crags (official trails, user-built trails, and campsites), and collected the data needed to build a computer simu- lation of the distribution of recreation use. Data Collection Procedures To build the computer s
23、imulation, data were col- lected on both visitor demographics and site charac- teristics. Site characteristics include a map of travel networks and popular destinations. Visitor CharacteristicsTrip diaries were used to collect the data on visitor itineraries needed to con- struct the simulation. Vis
24、itors were asked to take a diary with them and record on one side of the diary their group size, mode of travel, and date and time of entry and exit. They were also asked a series of attitudinal questions about trail, campsite, and management con- ditions. On the other side of the diary, a map was p
25、rovided so a group could record route information. Specific instructions for the map were as follows: Please indicate on the map where you camp, the number and type of encounter(s) you have and a notation at the edge of the map anytime you leave and re-enter the wilderness area. Please locate each o
26、f the campsites you visit as accurately as possible on the map. Place a C beside the campsite and a number that indicates the night of the trip that you camped at that location (Ex- ample C2 denotes the place where you camped on the 2nd night of trip). In addition to camp locations we would like you
27、 to indicate the number and type of encounters you have with other parties throughout your trip. Place an E to mark any encounters you have along the trail as they occur or while at camp at the end of the day. Associated with the E, provide one or more of the following notations to denote the type o
28、f encounter(s) you had, O (Other Party Camping), P (Packstock) or B (Other Backpacker) followed by a number which indicates the number of people in the group encountered. (Example EP10 means encounter with a packstock group of 10 people). To create trip itineraries in the format required by RBSim, a
29、ll spatial and relational data from the diary were entered into an Access database via a Web-based interface developed specifically to enter both types of data. Figure 1 provides an example of this interface and one of the trips that was entered into the database. This was a 4-day trip into the area
30、. C1, C2, and C3 were the three locations where the group camped, and DH repre- sents day hikes taken from each of the campsites. To represent this trip in RBSim, data were transformed into a sequence of travel routes and destinations (ge- nerically referred to as links and nodes). The sequence was
31、determined by the number on the left side of the destination notation. For example, the trip in figure 1 camped the first night at Harbor Lake. The next day began with a day hike to Bird Bill Lake, followed by a backpack to and camping at Sky High Lake. The next day began with a day hike to Terrace
32、Lakes, followed by a backpack to and camping the third night at Reflection Lake. The fourth day began with a day hike to Lost Lake USDA Forest Service Gen. Tech. Rep. RMRS-GTR-143. 200519 Figure 1Interface for capturing spatial data about the trip. and then a return to the trail head. No other visit
33、ors were encountered on the trip. This sequence of links, nodes, and durations at campsites are conformed into a data sequence that RBSim reads and uses to replicate this trip in the trace simulation. Site CharacteristicsGlobal Positioning System (GPS) tracks were used to inventory trails and camp-
34、sites to create a travel route and destination network. All trails in the study area were inventoried during the summer of 2003. Link data were then converted into ArcView format, and then converted into a format that conforms to standards established for RBSim. In some cases, it was clear that recr
35、eationists traveled offtrail between destinations. These informal routes were digitized and represented as part of the trail network. Campsite inventory data were also collected in the field. This consisted of creating a photographic record of the sites, using a GPS to locate recreation sites, and t
36、aking quantitative measures of variables such as soil erosion, vegetative cover, and amount and type of disturbance (Cole 1989). Similar to the trail data, campsite and trailhead nodes were inserted into the ArcView map. Since most of these campsites were situated beside lakes, a single node was est
37、ablished for each lake, and all campsites at the lake were associated with that node. Consequently, campsite statistics are provided for lakes rather than individual campsites. We were not confident that visitors could identify the exact campsite they used at a lake. Modeling Characteristics RBSim,
38、a recreation behavior simulator (Gimblett and others 2001) was used to describe patterns of visitor use across the landscape. The modeling ap- proach used in this study is a trace simulation using baseline fixed itineraries derived from the trip diaries. A typical trip for the Bighorn Crags simulati
39、on is described by an entry node, an exit node to the net- work, an arrival curve, a probability distribution of agent types, a list of destinations, and trip duration. On execution, the simulator reads all the trip itinerar- ies, schedules all trips, and then an agent representing the type of trip
40、moves from one destination to the next across the travel network using the shortest travel time between the two points. A large set of information for each agent is collected and stored in an Access database for later processing. Simulation Outputs Simulation output can be used to describe use level
41、s on trails and at lakes as well as encounter levels on 20USDA Forest Service Gen. Tech. Rep. RMRS-GTR-143. 2005 trails and at camping areas around lakes. There is considerable interest in the estimation of encounter levels because they are a common indicator and are difficult to measure in the fiel
42、d. To derive encounters along trails and at campsite destinations, a simulation was constructed. The simulation was run for a total of 89 days, from July 3 through September 29, 2003. A total of 75 trips through the Bighorn Crags were simulated. Average trip length was 3.6 days. Visitors typically a
43、rrived on Friday or Saturday and departed the wilderness on Sunday or Monday. Peak use oc- curred in late July and early August. Average group size was 2.8. Model outputs were visitor use levels and encounters by trail segment and camping area (lake). Statistics are the mean of 10 replications of th
44、e simu- lation. Trail encounters varied substantially between replications, while trail use, camp area use, and camp- site encounters did not. Trail UseTable 1 shows simulation output for three different trail segments. Total use is indicated by the total number of groups that traversed the trail se
45、gment during the simulated sampling period. The most heavily used trail segments are from the trail- head at Bighorn Crags campground to the trail inter- section that branches off to Welcome and Wilson Lakes. Encounters are recorded when one group over- takes another group in the simulation or where
46、 two groups pass each other in opposite directions. Table 1 shows the total number of encounters that occurred during the 89 day season, the mean number of encoun- ters per day, as well as the number of days on which encounters occurred. There is not a simple linear relationship between use levels a
47、nd encounter levels. Nor is there a linear relationship between the average number of encoun- ters and the days on which encounters occur. This illustrates the value of the computer simulations. Segments 64 and 66 have similar encounter levels, but both use levels and the number of days with encoun-
48、 ters are much higher for segment 64. Trail segment 64 is much shorter than segment 66. Apparently, as the length of trail segments decreases, so does the ratio between number of encounters and amount of use. Figure 2 displays the distribution of trail encounters across the Bighorn Crags. The maximu
49、m mean daily encounter level was just 0.51 encounter per day, and on many of the trail segments, no encounters occurred. Table 1Simulation output on use and encounter levels for selected trail segments. TrailTotalTotalEncountersDays with segmentgroupsencountersper dayencounters 612000 6410470.085.5 662660.071.4 This suggests that crowding on trails is not a serious issue in the Bighorn Crags. Campsite Use Table 2 provides information about use levels and encounters between groups at the more popular camp- ing areas in the Bighorn Crags. Campin