U.S. Army Corps of Engineers

The Corps defines a recreation visit as “The entry of one person into a recreation area or site to carryon [sic] one or more recreational activities” (Corps, 2025). In addition to recreation visits, Corps estimates overnight-use data, which are the number of nights that sites are occupied, the number of occupants per site, and visitor-hours. To better align with other Federal agencies and provide a consistent reporting approach, camping is reported as 1 visit per person per night (rather than counting multiple consecutive nights as 1 visit). Corps adopts the definition of a visitor-hour from the 1965 Federal Executive Policy Governing the Reporting of Recreation Use of Federal Recreation Areas: “the presence of one or more persons on lands or waters, generally recognized as providing outdoor recreation, for continuous, intermittent, or simultaneous periods of time aggregating 60 minutes” (U.S. Recreation Advisory Council, 1965, p. 2).

The Corps visitation estimates are measured by recreation area or project site area levels. Estimates are produced for each of the 4,750 project site areas. Visitation is calculated on a monthly basis at some sites and on an annual basis at other sites, following the fiscal year calendar. Visitation estimation is centrally controlled through procedures, policies, and guidance that are implemented by individual locations. Visitation is primarily estimated through five data sources and methods (Leggett and others, 2017):

Visitation is occasionally estimated using tally counts through visual observation, for example during special events or for school groups on buses. Visitation is estimated in dispersed-use areas by using additional methods, such as shoreline management permit data and household census data. It is extremely difficult to estimate visitation in dispersed-use areas. For example, adjacent homeowners may use lakes daily without ever crossing a site boundary or automated counter.

The Corps calibrates automated counters to ensure accuracy and develops conversion factors for automated counters. Calibration and conversion factors are developed using onsite surveys that take into consideration the types of traffic counters, vehicles, and lanes that are monitored. Persons-per-vehicle multipliers are updated based on visitor surveys that collect information about the number of people per vehicle and the percentage of vehicles associated with recreation. The agency also tries to avoid double counting visits by using onsite visitor surveys with questions about the percentage of vehicles departing for the last time and the length of stay.

Visitation data are stored in the internal Visitation Estimation and Reporting System (VERS). Annual visitation estimates from 2016 to the prior fiscal year (from the current fiscal year) are publicly available from a Corps website called “Value to the Nation” (https://www.iwr.usace.army.mil/Missions/Value-to-the-Nation/Recreation/), which can produce reports with visitation estimates at various levels (national, lake or river, State, Corps division, Corps district, and watershed levels; Corps, 2023b). This website also allows users to download a spreadsheet that contains more detailed visitation estimates and estimates by activity type. Annual visitation by Corps division is also available online from 2014 to the prior fiscal year (Corps, 2023a). The methods used to estimate visitation at each site are documented internally but are not publicly available.

For many years, Corps and other agencies have used estimates of recreational visitation for planning, management, and reporting purposes. Visitation is one of many data metrics used to ensure Federal appropriation investments are prioritized appropriately based on efficiency and performance. However, visitation is not the only consideration for establishing recreation budgets.

Bureau of Land Management

The BLM defines a recreational visit as:

The BLM also estimates the number of visitor-days and the number of visitor-days by activity, where 1 visitor-day standardizes recreation visits of different lengths into 12-hour visitor-days (Leggett and others, 2017), which is consistent with the definition of a visitor-day from the Federal Executive Policy Governing the Reporting of Recreation Use of Federal Recreation Areas (U.S. Recreation Advisory Council, 1965). For example, 1 visitor-day could be one person visiting for 12 hours or 12 people visiting for 1 hour each. The BLM also tracks the number of recreation permits issued (Leggett and others, 2017).

The BLM collects these visitation data at all open and active BLM sites as well as dispersed areas. There are more than 3,500 developed sites and more than 1,000 dispersed areas where visitation is estimated. The BLM has more than 150 field offices and offices at other designation types, such as national monuments and national conservation areas, and each one is responsible for developing its own site-specific methods to estimate visitation. Although this approach grants individual field offices and units flexibility to develop methods that work best for their unique context, the accuracy of the methods used to estimate visitation varies substantially across sites and field offices (Leggett and others, 2017).

Because each field office is responsible for developing their own methods for estimating visitation, there are a variety of methods used across the BLM. The three most common methods for estimating visitation are as follows (Leggett and others, 2017):

Because each office is responsible for developing site-specific methods to estimate visitation, the methods used to develop and implement conversion factors (for example, to convert the number of vehicles to the number of recreation visits) vary by office. Additionally, the methods used to reduce or eliminate double counting of visits vary by office and there is no single protocol used across the agency.

Visitation is reported annually at the end of each fiscal year, although some sites have data on smaller time scales, such as monthly. Processed data (including any adjustments and conversions) are stored in a database called the Recreation Management Information System (RMIS). These data are only available to certain BLM staff who have access to RMIS and are not publicly available. However, aggregated visitation data are reported at the national and State levels in the BLM’s Public Land Statistics reports (such as BLM, 2024). In spring 2024, BLM field staff who were involved in visitation estimation completed an internal spreadsheet to indicate methods used to collect data and estimate visitation at each site. This is the first step in the process of internally documenting the methods used to estimate visitation at each site.

The agency uses visitation data for the annual Public Land Statistics reports. In addition, the data are used to conduct impact analyses for the National Environmental Policy Act, support budget requests and grant applications, and help prioritize monitoring and allocation of resources.

Bureau of Reclamation

Reclamation considers a visit to occur “when a person enters Reclamation lands to engage in recreation on a given day” (Leggett and others, 2017, p. A-2). However, Reclamation often shares jurisdiction with other Federal agencies or nonfederal partners (for example, State, county, and city governments). In cases of shared jurisdiction, Reclamation usually cedes recreation management, including the definition and estimation of visitation, to the partner agency. Therefore, the definition of a visit could vary at partner-managed sites where different partners may use slightly different definitions (Ronnie Baca, Reclamation, oral commun., 2024).

Reclamation estimates visitation at each management area (for example, reservoir). Visitation estimates are reported annually by either Reclamation or partner agencies with shared jurisdiction. Given the site-level variation across partner-managed sites, staff at each management area are responsible for developing their own site-specific methods for estimating visitation. Although methods vary by management area, there are three main methods to estimate visitation (Leggett and others, 2017):

Because each management area develops site-specific methods to estimate visitation, the methods used to develop and implement conversion factors (for example, to convert the number of vehicles to the number of recreation visits) vary by management area. Additionally, Reclamation does not have a standard method to avoid double counting visitors who enter multiple times in 1 day; this is also up to each management area. Some management areas have methods to avoid double counting, but other management areas do not. There is no agency-wide documentation for how visitation is estimated at different areas; some management areas may have documented methods, but other management areas may not.

Visitation numbers for each area are reported annually in an ArcGIS program. The data are stored in a database on an ArcGIS server, and reports are accessed through an ArcGIS dashboard. This visitation dashboard is only available internally and is not publicly accessible. Combined annual visitation estimates for all Reclamation sites can be found in the Federal Lands Recreation Enhancement Act Report to Congress (U.S. Department of the Interior and U.S. Department of Agriculture, 2024). The agency considers visitation data for a broad range of decision-making and planning purposes and for reporting. For example, visitation data can help inform Reclamation on how to allocate resources at a regional level.

U.S. Fish and Wildlife Service

The FWS defines a recreation visit as “when a visitor engages in a particular recreational activity ([for example], hunting). A visitor engaging in multiple activities is counted as multiple visits” (Leggett and others, 2017, p. A-2). Although this definition is still in use, FWS continues to have discussions around the definition as it pertains to reducing the propensity to double count. For example, counting an individual only once on the same day for inclusion in annual visitation estimates while also maintaining the ability to estimate the number of different activity-type visits for the same individual. This metric also does not consider the length of stay; for example, a person on a 3-day camping trip or a 1-day camping trip would both be counted as 1 visit. In addition to total visits, FWS tracks the total number of visitor-days and the number of visits by activity. The definition of a visitor does include a time component; a person on a 3-day camping trip in a refuge would count as 3 visitor-days. The FWS primarily uses this visitor-days metric to estimate visitation.

The FWS estimates visitation at the unit level; this is most often at the national wildlife refuge level but also includes marine national monuments and wetland management districts. Visitation estimates are reported annually. Similar to the BLM, each unit is responsible for developing unit-specific protocols for visitation estimation. In the past few years, the staff at the headquarters for the refuge system have developed a collaborative program of visitation estimation research with university partners. The goal of this project is to understand how to improve visitation estimation accuracy through the adoption of more efficient methods and the development of greater staffing capacity and training across the system.

Because each national wildlife refuge, marine national monument, and wetland management district is responsible for developing site-specific protocols, the mix of methods used at each unit varies. Common techniques include (Leggett and others, 2017):

Because each unit develops site-specific methods to estimate visitation, the methods used to develop and implement conversion factors (for example, to convert the number of vehicles to the number of recreation visits) vary by unit. Additionally, the methods used to reduce or eliminate double counting of visits vary by unit, and there is no set protocol used across the agency. Recently, FWS has partnered with researchers from the University of Washington and Clemson University to improve, standardize, and simplify visitation estimation across the system. This effort includes identifying conversion factors used for counters across a subset of units. This will help inform visitation estimation training tools and resources, with an overall goal of working toward standardization across the agency. The project has five interrelated components: (1) interviews with FWS staff involved in visitor estimation; (2) a systematic literature review and synthesized decision tool; (3) engagement and capacity building with refuges; (4) developing a statistical modeling approach for reliably estimating visitation at refuges, including using geolocated social media and mobile data in the models; and (5) an interactive framework for a new visitation estimation toolkit (Dagan and others, 2024).

Visitation data reported by FWS staff are stored in an internal database called “Refuge Results” that is not available to the public. However, visitation numbers for individual units can be found in reports produced by the National Wildlife Refuge Visitor Survey. Each report contains visitation estimates for the same year that visitor surveys took place at each refuge (Dietsch and FWS Human Dimensions Branch, 2024). The agency uses visitation data for a broad range of decision-making and planning purposes. Some of these purposes include staffing and resource allocation, infrastructure development, and community outreach and engagement.

U.S. Department of Agriculture Forest Service

The FS defines a recreation visit as “one person participating in one or more recreation activities on a national forest or grassland for an unspecified amount of time” (FS, 2023, p. 9). People who do not recreate are not included in the visit count; for example, employees, contractors, and people who enter FS lands simply to use a restroom or obtain information are not counted. During a single recreation visit to an FS area, an individual may visit multiple sites within the national forest or national grassland. The FS produces separate estimates for national forest visits and individual site visits. Visitation is estimated across the FS as part of the larger NVUM program.

The NVUM program was piloted in the late 1990s and launched in 2000 to collect annual data on visitors to FS lands. “The NVUM program serves two concurrent goals: (1) to estimate the volume of recreation visits to units of the [National Forest System], and (2) to describe salient characteristics of those visits, including activity participation, visit duration, visitor demographics, and visitor satisfaction” (English and others, 2020, p. 65). The NVUM program is a large-scale effort, typically surveying 24 or 25 National Forest System units each year, with an average of around 23,000 surveys completed annually across the system (English and others, 2020). The forests are surveyed on a 5-year rotation, meaning that each forest will have data collected once every 5 years, and data will be representative regionally or nationally over any 5-year period (English and others, 2020). The method and effort have been fairly consistent for more than 20 years, which allows for a long-term consistent dataset of visitation estimates and other visitor use data.

The NVUM program chooses sampling sites and days within a forest using a stratified random sampling approach to get a representative sample across the forest. Sites are stratified based on site type and use levels. Site types consist of day-use sites, overnight sites, wilderness access sites, and general forest area access sites. Use levels are categorized by field staff for each site and each day of the year as not used or closed, low, medium, high, or very high, which considers both the time of year and time of the week, recognizing many locations have higher use on weekends or holidays. This sampling framework for selecting sites for data collection that are representative across the forest and throughout the year allows for accurate visitation estimates across the forest. Each forest has an average of 230–240 sample days during an NVUM year. For more information about specific methods used for NVUM, including site stratification and field sampling methods, refer to the article by English and others (2020).

Visitation is estimated as part of the NVUM program. Visitation estimates are produced for each forest once every 5 years. This process is centrally controlled (meaning national-level employees administer and oversee the program) and consistent across FS lands. Estimates are made using three types of information (English and others, 2020):

During interviews, visitors are asked if they are exiting for the last time, and only those who are finishing their visit are asked to complete a survey. The manual count from the sample day is adjusted based on the share of those interviewed who are exiting for the last time. Additionally, the FS eliminates double counting of visitors by asking how many FS sites visitors went to during their trip. Finally, visitors are asked how many people are in their vehicle and these data are used to convert traffic counts to visit estimates. These conversion factors are updated for each forest every 5 years with the NVUM cycle; the conversion factors use both the current year and prior sample year data, so a single sample year does not dominate the conversion factors (Sarah Cline, FS, written commun., 2024).

The NVUM data are stored in the “Results Application,” and data stored in this database are cleaned from the raw data to produce estimates. Estimates are then transmitted to the FS Natural Resource Manager Program (an internal FS information technologies program), which maintains the database. The “Results Application” can be used to produce reports and tables. Summarized data are publicly available through a web-based user interface, the NVUM website (https://apps.fs.usda.gov/nvum/results). The website allows users to select any forest and obtain summary statistics of data that were collected for any year NVUM data collection took place for that forest. Users can also retrieve results at a regional or national scale over any 5-year period, and the FS produces a national-level annual report that is available on the NVUM website. Because visitation estimates are produced using the same methods at each location, there is no need for site-specific documentation of methods; however, there is publicly available documentation describing methods used in the NVUM program, including how visitation is estimated (English and others, 2020). Visitation estimates have been used for various purposes within the agency, including congressional and departmental reporting, agency accomplishments and communication, executive dashboards, National Environmental Policy Act analyses, forest and strategic planning, natural disturbance effects analysis (for example, effects of major fires), forest and recreation management, and partnerships (English and others, 2020).

National Oceanic and Atmospheric Administration Office of National Marine Sanctuaries

This report includes information from the ONMS only and does not include information from other parts of the Department of Commerce, such as the National Marine Fisheries Service, International Trade Administration, or National Travel and Tourism Office. ONMS does not use the term “recreational visit” but instead defines a “person-trip” as equal to one person who makes a trip and a “person-day” as one person doing any recreation activity for any part of a day (Danielle Schwarzmann, NOAA, written commun., 2024). A visitor could engage in several person-days of different activities in 1 day. ONMS tracks different metrics at locations across the system including visitor-days (person-days) and visits by activity. The metrics that are tracked vary based on the specific management priorities of a given marine protected area.

Currently (as of 2024) and historically, ONMS has not collected consistent visitation data for the entire system. The ONMS network includes 17 national marine sanctuaries (NMS) and 2 marine national monuments. However, ONMS has collected consistent data at specific sites and locations for several years as part of socioeconomic research projects. The ONMS is in the process of developing a visitation monitoring program that will allow for consistent data collection. The ONMS is working with the Office of Management and Budget (OMB) on a NOAA “Compendium of Questions” that would allow for consistent collection of visitation data. The ONMS estimates visitation using two methods (Danielle Schwarzmann, NOAA, written commun., 2024):

For ONMS, the reporting scales are dependent on the management action or site-specific research need. Common scales are per facility (for example, visitor centers), sanctuary, specific habitat, or cultural resource or management area within a larger protected area. Data are most commonly reported at the monthly time scale, though the specific timescale is defined by the management scope and application of the data (Danielle Schwarzmann, NOAA, written commun., 2024). The ONMS uses survey questions to estimate the number of people on each vessel and, when other data on the number of boats, such as satellite imagery or photographs, are available, will continue this across sites to estimate the number of recreational boaters. In one previous project, NOAA partners used mobile device location data in the process of estimating visitation in the Florida Keys NMS in Florida (Schwarzmann and others, 2022).

The ONMS conducts recreational studies to support specific regulatory or policy needs. Examples include reports on Washington State resident use of the Olympic Coast NMS off the coast of Washington State (Leeworthy and others, 2016), whale watching in the Stellwagen Bank NMS off the coast of Massachusetts (Schwarzmann and Shea, 2020), and recreational fishing in the Florida Keys NMS in Florida (Schwarzmann and others, 2022). For these reports, ONMS produced technical appendixes detailing the methods applied. The ONMS stores the datasets for these studies internally and can provide them upon request. The National Centers for Environmental Information (part of NOAA) supports ONMS’s developing visitation monitoring program and may offer centralized data storage, including raw (unadjusted) visitor count data in the future. The National Centers for Environmental Information applies specific protocols to ensure that data are publicly available in a timely manner and to protect personally identifiable information and business identifiable information.

The ONMS visitation data are used to support rulemaking (including sanctuary expansions and updates to regulations). The data are also used for designating new sanctuaries, producing condition reports, and reporting official statistics, like the Outdoor Recreation Satellite Account (https://www.bea.gov/data/special-topics/outdoor-recreation), Marine Economy Satellite Account (https://coast.noaa.gov/digitalcoast/data/marine-economy.html), and Economics: National Ocean Watch (https://coast.noaa.gov/digitalcoast/data/enow.html).

National Park Service

The NPS defines a recreation visit as “the entry of a person onto lands or waters administered by the NPS except…for non-reportable and non-recreation visits” (NPS, 2024b). Funeral parties at national cemeteries, school groups, and other visitors, are reportable as “recreation” use because such uses are for the purpose for which the park was established. Visits originating on surface vehicles (trains, boats, other) and aircraft may be counted if they stop and disembark passengers on NPS-administrated territory. “The applicable rule is that one entrance per individual per day is countable” (NPS, 2024b). In addition to recreation visits, the NPS monitors the number of nonrecreation visits, hours of recreation and nonrecreation use, number of recreation and nonrecreation overnight stays, and other information, such as visitation associated with special events or a particular attraction (NPS, 2024c). Definitions for each term can be found online at https://www.nps.gov/subjects/socialscience/nps-visitor-use-statistics-definitions.htm.

Visitation is estimated only for areas administered by the NPS. An area is administered by the NPS when one or more of the following conditions are met: (1) the park superintendent has the authority to develop and enforce regulations on the property, (2) NPS funds directly support the management of the property, or (3) legislation or interagency agreements direct the NPS to administer the property. Official NPS reporting of visitation is not allowed for areas primarily operated by other Federal agencies or in areas specifically excepted by NPS management, legislation, cooperative agreements, and memoranda of understanding or other official documentation (for example, affiliated areas). Data collected in miscellaneous areas where the NPS has partial administrative responsibility or limited presence may be maintained as sources for comparisons but are not reported in the combined total statistics of those areas directly administered by the NPS. Visitation of areas for which the NPS provides services (for example, patrol or other emergency services) without any legislative, contractual, or other official and externally imposed requirement is not counted as NPS visitation.

The NPS uses a wide variety of technologies and methods for estimating visitation. Some parks find that manually counting visitors at a visitor center, on the grounds, or at other attractions provides a reasonable estimate for their park. Many parks, particularly those with entrances on travel corridors such as roads and trails, find it optimal to use automated counters. Where necessary, mathematical relationships may be used instead of manual or automated counts to estimate the number of visits to an area based on a count of visits to a more easily measured area of the park. The NPS estimates recreational visitation using four techniques (Leggett and others, 2017):

Data collection techniques are selected specifically for each park considering what is operationally feasible, data that are available monthly, and data that can be collected over many years. If these data do not directly represent visits or overnight stays, various adjustments must be applied to convert the “raw” measurements to visits or overnight stays. For example, when automated counters are used, parks need various conversion factors to change readings from counters to monthly visitor use estimates. These conversion factors are established by studies of visitor use. A study may consist of systematically observing the number of people in vehicles coming into the park, the number of visitors entering a visitor facility, or asking visitors questions about their visit. The NPS Social Science Program assists parks in the technicalities of such studies and reports conversion and correction factors needed for the valid and accurate reporting of visitor use statistics. Additionally, for most parks where multiple entrances per day are common, the NPS applies adjustment factors to correct for duplicate visits. There are updates underway using SEM survey data to build adjustments for same-day reentry counts and duplicate measurements. The NPS has also started using mobile device location data at some locations to develop multipliers and adjustment factors. These data (described in the “Mobile Device Location Data” section of this report) are only used to develop adjustment factors and are not used for continuous data collection or reporting. Mobile device locations are used to develop adjustments in certain locations when they are considered a better source than traditional sources, such as visitor surveys (refer to the first case study in the “Case Studies” section of this report).

The NPS visitation data are stored in a Microsoft Structured Query Language Server database managed by a contractor and owned by the NPS Visitor Use Statistics Team Lead. Data can be accessed internally and externally through the visitor-use statistics website at https://irma.nps.gov/Stats/. The NPS Visitor Use Statistics Program is centrally coordinated, and visitor-use statistics staff in the NPS Social Science Program are responsible for collaborating with parks to develop site-specific data collection methods. This ensures consistency and reliability of data collection across NPS units. Monthly and annual visitation estimates are available on the Integrated Resource Management Applications (IRMA) website (https://irma.nps.gov/Stats/) for both the national level and individual park units. The spatial scale for official reporting is the park unit, but most park units have measurements or estimates on a smaller spatial scale (for example, an entrance station, a visitor center, or a road) that are used as building-block elements of the broader park-level reporting.

Methods for estimating visitation at each park unit are also available online through the IRMA website (https://irma.nps.gov/Stats/). These count procedures are reviewed periodically, ideally every decade or when there are new considerations (for example, new construction or legislation) that alter visitor use (Pamela Ziesler, NPS, written commun., 2024). The counting and reporting instructions for each park describe the protocol for estimating visitation. However, the instructions do not provide further documentation on the estimation approach, such as how conversion factors were developed, for example, how to convert total vehicles to the number of recreation visits (Leggett and others, 2017).

Within the NPS, information about visitation allows park managers to handle the challenges of increasing or decreasing visitation and mitigate negative effects on cultural and natural resources related to visitor use (Pamela Ziesler, NPS, written commun., 2024). Visitation data are a measure of workload requirements for the cost-effective dispersal of park staff. The data support planning visitor services and related functions, scheduling maintenance and cleanup, tracking crime and accident rates, estimating resource deterioration and use concentration indices, monitoring public hazard zones, verifying fee revenue, informing reports to Congress and other stakeholders, and developing construction-contract specifications for visitor facilities. In addition, modern park management practices, such as decisions on the need for and monitoring of reservation or permitting systems, require information on visitor use. Outside of the NPS, accurate visitor-use information allows local communities and businesses to determine the types and quantities of amenities, goods, and services needed by visitors to nearby parks. Provision of community services such as public water supplies, water treatment, solid-waste handling, emergency services, food, lodging, and transportation services may all require accurate NPS visitor use statistics. Consequently, a detailed understanding of visitation levels in nearby parks is beneficial for the appropriate management of community resources. Visitation data also provide information to local communities about the potential economic effects of nearby parks (Pamela Ziesler, NPS, written commun., 2024).

Social Media Data Overview

During the past decade, social media use has continued to increase among the U.S. population. Both the number of social media platforms and the percentage of Americans using social media platforms have increased (Pew Research Center, 2024a). There have concurrently been large changes in the makeup of social media applications that people most widely use, as well as the ways that people interact with these platforms (Pew Research Center, 2024a). As some platforms have increased in use, gaining more users and postings over time, the use of other platforms has declined. Instagram, for example, has continued growing as an image-sharing platform since 2017 (Pew Research Center, 2024a), whereas Flickr use has decreased (Stuart, 2019; Goebel and others, 2023). Over the last 10 years (2014–24), the age of social media users has expanded to include older demographics, and users have shifted from interacting with social media primarily through their computer and web browser to using mobile phone applications (Cotten and others, 2022; Pew Research Center, 2024a, 2024b66).

During the same period that social media use has been increasing and evolving, researchers have been testing the potential for information provided by social media platforms to function as a source of recreational visitation data (Wilkins, Wood, and Smith, 2021). Generally, studies have concluded that data derived from various social media platforms are informative for estimating visitation, if care is taken to assess and model the data properly (Wilkins, Wood, and Smith, 2021; Ghermandi, 2022). This has remained true even as the specific sources, types, and availability of social media have shifted over time.

In 2017, Leggett and others (2017) described three types of social media platforms that could be used to infer recreational visitation data for public lands: (1) photograph sharing, where the geographical location (geotag) of a photograph shared on social media (for example, on Flickr) is used to infer the presence of a person; (2) location sharing, where the user can choose to geotag the location that corresponds to the place where the social media were generated (such as a post on Twitter, renamed “X” in July 2023); and (3) path sharing, where users can share the entire path they took on a trip, usually running/hiking or cycling (for example, on Strava). Additionally, beyond the photograph-sharing, location-sharing, and path-sharing platforms that were discussed by Leggett and others (2017), there is now a multitude of platforms that visitors use to share reports about their trips to specific recreational sites. These review-sharing platforms are often designed to serve a specific community or user group that wants to share information about a particular recreational activity, such as off-highway vehicle use, mountain biking, or hiking. The level of access that researchers have to social media has shifted over time for individual platforms and, as of 2024, most social media companies do not allow free data access to researchers.

Among the review-sharing platforms, Trailforks (https://www.trailforks.com) has emerged as a popular platform for people to share information about mountain biking trails and to report on their experiences. Similarly, the AllTrails platform (https://www.alltrails.com/) is commonly used to share trip reports and review trails for activities such as hiking and running. On a more local scale, a trip-reporting platform run by the Washington Trails Association (https://www.wta.org/) is widely used by hikers in Washington (Armstrong and others, 2022); similarly, the Colorado Trail Explorer is used to share trip reports for trails in Colorado (https://trails.colorado.gov/). Because trip reports posted on these platforms are geolocated and time-stamped, they provide a source of information on visitation. A study by Fisher and others (2018) found that the monthly counts of trip reports on the Washington Trails Association’s platform are highly correlated with on-the-ground counts of trail users at 16 trails across Washington. White and others (2023) determined that geolocated reviews on AllTrails were useful (in addition to data from other social media platforms) to model changes in visitation to public lands in the Columbia Gorge for a study on the effects of wildfires and associated closures in Oregon and Washington State.

The conclusion from research during the past decade is that geolocated social media are informative for visitation estimation, but no one data source is accurate enough to serve as a substitute for on-the-ground counts across a wide range of site types. Although the number of social media users and posts is generally positively correlated with the actual number of people in a place and time, there is high site-to-site variability in the relationship between the number of geotagged social media posts and actual visitation (Wood and others, 2013; Sessions and others, 2016; Heikinheimo and others, 2017; Tenkanen and others, 2017; Fisher and others, 2018). Furthermore, this relationship varies by data source and over time (Donahue and others, 2018; Winder and others, 2025). So, although a data source might be strongly related to actual visitation in one location, it may be weakly related to visitation in another location where an alternative social media data source may work better.

To make sense of variability in space and time, researchers have developed visitation modeling approaches that use equations to combine multiple social media sources into predictions that leverage their combined explanatory power while at the same time attempting to statistically correct for known issues. Importantly, these methods use onsite “ground-truth” data to calibrate the relationship between social media data and actual visitation. These calibrated relationships form the basis of statistical models that convert the data to estimates of actual visitation, potentially even at new sites or times where onsite data are not available, with varying success. Wood and others (2020) developed and tested this calibrated-relationship approach using social media posts from Flickr, Instagram, and Twitter (now X) with onsite visitation counts from 42 recreation sites on Federal lands in the Western United States. They found that including predictors derived from the three social media data sources substantially improved model performance, measured as the accuracy of visitation estimates at new locations, even when models were developed with data from one location (Washington) and applied in another location (New Mexico). The best performing models also included other variables known to affect visitation, such as seasonality, local weather conditions, and holiday timings.

Social Media Data Assessment

During the past decade, researchers have concluded that the location and timing of geolocated social media posts can be used to improve visitation estimation. For example, a study that evaluated the use of social media data for visitor estimation on public lands and waters in the United States showed that between 45–91 percent (approximately) of the variability in weekly visitation to a site can be determined by statistical approaches that rely on learned relationships between numbers of social media posts from multiple data sources and onsite counts (Wood and others, 2020). This variability likely reflects underlying biases in who is actively using different social media platforms and, therefore, how social media under- or over-represent visitation to different locations.

Although the overall conclusion is that geolocated social media are inherently useful for visitation estimation, there can be challenges to using these data in practice. One known limitation is that data quality and data access vary not just spatially but also over time (Wood and others, 2020; Ghermandi, 2022). During the last decade, the popularity of social media platforms has varied, and others are no longer available for use in visitation estimation because the companies that own them have made the data unavailable, often to monetize data access. Previously, Twitter (now X) allowed researchers to query and download an unlimited number of geolocated tweets and metadata for free (Tromble, 2021; Pfeffer and others, 2023). In February 2023, X started charging a fee for access to the same data (Developers, 2023). Similarly, in 2018, Meta (the parent company of Instagram) shut down its public Instagram Application Programming Interface (API), which had previously allowed access to geotagged posts, and replaced it with an API restricted to business use with less data availability (Gummadi, 2018). This general loss of social media data access was raised by Leggett and others (2017) as one of the risks of relying on data owned by private companies, and this continues to be a concern. The monetization of previously free data access will likely continue to make it more difficult for researchers to access social media data. This issue, and many other issues with digital mobility data, are not unique to social media. Additional issues and limitations on implementation with other types of digital mobility data are discussed in their respective subsections below and in the “Implementing a Visitation Estimation Method” section of this report.

Mobile Device Location Data Overview

Since the 2017 report by Leggett and others was published, there has been a shift in the type of mobile device location data that is available. The main source of data available in 2017 was derived from call detail records (CDR). These data were generated by cellular service providers based on the location of the hardware (the “tower”) that connected with the cellular phone at the time of a phone call or text message. In 2024, the most commonly available location information from mobile devices (including mobile phones, tablets, and smartwatches) is provided by applications (apps) that use location-based services (LBS). The recorded location is often determined by the device using a Global Positioning System (GPS).

Very few studies have researched the use of CDR to estimate recreational visitation, and none of the known studies have been conducted in the United States. Fisher and others (2019) were the first to test the relationship between CDR data and onsite visitation estimates at parks and other nature-based tourism destinations. Using sites on Jeju Island, South Korea, the authors found that CDR data provided by Sun Kyung Telecom (SK Telecom Co., Ltd.)—one of three major mobile communications providers in South Korea—were positively correlated with field-based counts (Fisher and others, 2019). Other studies using CDR data to study tourism patterns outside the United States are discussed in Zaragozí and others (2021). In 2019, major telecommunications companies in the United States voluntarily stopped selling CDR data (Krebs, 2018) and shortly after, in 2020, the Federal Communications Commission issued a notice of liability for forfeiture and admonishment and then fined four U.S. companies for selling customers’ location data without their consent (Federal Communications Commission, 2024). Accordingly, CDR data are unlikely to be a source of visitation data in the United States going forward.

During the last 5 years (2020–25), as the number of smartphone users and smartphone applications has risen, there has been an increase in the availability of mobile device location data from LBS running on those applications. Simultaneously, an industry has arisen that is devoted to the buying and selling of those data. The applications generating these data provide a variety of services to the device user such as weather information, social networking, games, and a multitude of others. These applications are location-aware because the developers integrate software, often by using a Software Development Kit (SDK), which then interfaces with GPS hardware or uses other information, such as surrounding wireless access points, to determine the location of the device. This data collection requires users to grant permission for the app to access the location of devices running the Android and iOS operating systems.

Each recorded location contains, at a minimum, a latitude and longitude, timestamp, and user identifier, but can also contain information on the speed of the device, information about the device itself (such as make and model), and a measure of the GPS precision, as well as other information. These data are then transmitted back to the application developer or, if there is no cellular connectivity available, optionally cached on the mobile device and uploaded once connectivity is reestablished. To reach the market, the location data received by the application developer is then de-identified (personal information, such as names, is removed) and sold, typically through third-party data vendors.

There has been a rapid proliferation of third-party companies buying and selling mobility data derived from SDKs since 2017. These vendors package data into a wide range of products. Some companies sell unaggregated and minimally processed location data, which contains latitude and longitude coordinates of each individual device (according to the anonymized device identifier described in the previous paragraph). Other data companies sell data that have been aggregated to certain geographical areas or points-of-interest. Still others apply considerable processing to the data. Many vendors will, for example, use the number of devices detected in a given area to extrapolate the total number of people in that area.

A 2025 study comparing five sources of geolocated social media data and three sources of mobile device location data created by visitors to United States Federal public lands indicated that, relative to social media, LBS-derived data products often capture a larger proportion of total visitors (Winder and others, 2025). Nonetheless, mobile device location data are similar to social media in that the raw data represents a fraction of visitors to a location. The same study observed between 9–288 percent of visitors counted onsite in a sample of United States national wildlife refuges are recorded in one vendor’s location dataset, showing that mobile device location data can show much smaller or much larger visitation numbers than those observed on the ground (Winder and others, 2025). The percentage of total visitors captured by a given dataset—and potentially also the representativeness of the dataset—at a site depends on several factors. First, individuals must carry a mobile device for their location to be recorded. The device also needs to be running an application that determines and stores its location using GPS data (through a mobile application), and the user needs to opt in to location sharing. When applications are not able to cache data, a connection to cellular service is needed to reliably capture location data, and such connections can vary across sites. Some data vendors also adjust the dataset to attempt to correct for changing device sample volume across space, but whether and how adjustments are made is not usually shared with the data purchaser. Finally, because each location-data vendor buys data from a subset of applications and SDKs, and because it is not feasible for a vendor to buy all data on the market, there will be a bias in the app data purchased by a given vendor, which can differentially affect data reliability at different sites.

Despite the increasing availability of LBS-derived data, few studies have evaluated the use of the data to estimate recreational visitation. Most research has involved comparisons with estimates based on observational data (for example, ticket sales and traffic counters). Merrill and others (2020) compared total visitation estimates from a third-party location-data vendor for 18 water recreation areas in the Eastern United States (such as beaches and boat ramps) to visitation measured using various observational methods. The extrapolated estimates from the vendor were about four times greater than the observational estimates. Tsai and others (2023) and Winder and others (2025) estimated visitation in 38 U.S. national parks and 13 U.S. national wildlife refuges, respectively, and both observed that visitation according to location data derived from LBS that was sold by two different vendors showed a wide range of correlation with actual visitation across the entire United States, from less than zero to 0.996 (Pearson’s correlation values). Filazzola and others (2022) measured the relationship between empirical visitation estimates and an “activity index” based on data derived from apps providing LBS and observed generally strong correlations between the two in urban green spaces in the Greater Toronto Area.

Research into methods for using LBS-derived data to estimate recreational visitation has reached similar conclusions as studies evaluating social media—that LBS-derived data is most useful for producing estimates of visitation when included as inputs into statistical models, along with covariates such as weather and seasonality, and parameterized using on-the-ground counts (Merrill and others, 2020). Multiple studies have found that, similar to social media, LBS-derived location data on their own may not be a suitable proxy for recreational visitation (Tsai and others, 2023; Winder and others, 2025), but when used as a model input along with other variables and calibrated using onsite data, LBS-derived data can provide valuable information for land managers.

In general, studies evaluating the use of third-party location data from LBS for measuring recreational visitation conclude that, although mobile device location data are a promising data source to complement more traditional onsite counting methods, they should be calibrated with onsite data and ideally used to develop predictive models for similar locations where onsite data are not available. Furthermore, out-of-sample model testing, where model performance is evaluated on a portion of the initial dataset that was held out from model training, should be used to avoid overfitting and ensure models can be used successfully for new sites or time periods without onsite data (Merrill and others, 2020; Wood and others, 2020).

Mobile Device Location Data Assessment

Mobile device location data derived from LBS have been shown to be most valuable when used as one of multiple sources of information in models predicting visitation. For example, variables that are known to affect visitation, such as weather, as well as other proxies for visitation, such as geolocated social media posts, can improve models that use LBS data to predict visitation (Merrill and others, 2020; Winder and others, 2025). These data may be particularly useful for situations where visitors are dispersed across large and remote areas because LBS data have the potential to fill gaps in visitation estimates where onsite counting is unfeasible, as long as best practices, such as calibration with onsite data from other sites and out-of-sample testing, are followed.

There are several additional considerations and challenges to using mobile device location data for visitation estimation. First, there are logistical and operational considerations, including the monetary cost of acquiring data from third-party vendors, which can be from tens to hundreds of thousands of U.S. dollars depending on the company, the geographic extent, the purchaser, and the amount and type of data. Companies generally sell data either as a single, one-time purchase of data for a defined geography (for example, a single park) or as part of an ongoing data subscription for any geography as long as the subscription is active. Terms of service may put limits on data sharing, which can also factor into the cost (for example, if data cannot be shared with multiple field offices).

After data are acquired, data storage and processing pose potentially significant costs. When data are aggregated for a limited number of geographies or points of interest, files are generally small enough to store on personal computer hard drives, but raw mobility data, especially for a large area such as the entire United States, can require hundreds of terabytes of disk storage and cloud storage systems. These services can incur costs for storage as well as access to the data. Processing such large datasets may also require cloud computing resources that have associated costs. Additionally, purchasing, transferring, storing, processing, and analysis of mobile device location data need to be done by personnel with the technical expertise required to perform these tasks. Finally, privacy issues associated with mobile device location data must be considered, including removing sensitive information, securing storage, and creating publicly available policies for working with private data. This issue applies to other forms of mobility data and is discussed in further detail in the “Best Practices for Novel Methods” section of this report.

Beyond the operational considerations, it is important to consider issues associated with representation and sampling bias that are inherent in mobile device location data. These issues will vary from one data vendor to another, and it cannot be assumed that data for a site purchased from one provider will perform similarly to data purchased from a different provider. Moreover, as discussed in the “Mobile Device Location Data Overview” section, data derived from LBS are limited to the subset of visitors who have a mobile device, use an application that collects location data, and consent to share their data. If a particular user group is less likely to use a device (for example, older visitors or foreign visitors), their visits will be underrepresented in the dataset. This bias could affect the accuracy of absolute measures of visitation and comparisons among locations that differ in the makeup of visitors. Additionally, technical aspects of the mobile applications and SDKs that provide LBS can affect data representativeness, as can proprietary decisions made by the third-party vendors that resell location data.

Application developers and third-party data vendors do not provide much information about how location data are generated and their proprietary data processing steps. For example, third-party location data vendors generally do not share specific information on which specific mobile applications—or even the number and type of applications—are used in their commercial data products. If, for example, a third-party vendor resells data collected through an application that is popular with recreational fishers, then the dataset may have high correspondence to visitation at sites where fishing occurs but will be less representative at sites where there is no fishing. Data companies typically buy data from many mobile applications and SDK developers as one way to increase representativeness. However, with the lack of transparency from developers and vendors, it is difficult to gauge whether these decisions are responsible for the observed bias and changing volumes of mobile device location data over time (Winder and others, 2025). Further, because these decisions are made by each data vendor individually, datasets from one vendor cannot be assumed to be interchangeable with those from another.

Without knowledge of which mobile applications are used to create a dataset, it is unclear how data representativeness is affected by differences in the technical design of each mobile application from which third-party data vendors source data. Each developer of a mobile application or SDK that provides LBS makes different decisions, such as whether applications must be running and open on the device to determine a location or whether locations are collected in the background on a regular schedule, which are factors that depend on permissions the user has given to the app. Either way, the frequency with which an application determines a location affects the utility of the data. Applications that determine a location once per day, for example, are underestimating actual visitation to an area compared with an application that collects a location once per hour or more frequently.

One design decision made by developers of mobile applications and SDKs is whether to cache location data when a device is out of range of cellular service. This decision is likely affecting the amount and potentially the representativeness of location data from Federal public lands where there is often limited cellular connectivity (Lawson and others, 2023). The extent to which applications have the capability to cache data is unknown, because this information is not shared by application developers nor the third-party companies selling LBS-derived data.

An additional unknown factor with third-party location data is the amount and nature of data processing that vendors apply before reselling the data. Data can generally be bought either in a raw form or in an aggregated form. During aggregation, data may undergo significant processing, and sometimes a population-level estimate is extrapolated from the sample. However, those processing methods are usually considered proprietary and not shared with customers. Even data sold as unaggregated locations of individual, anonymized devices can still be altered and processed by the company, for example, by adding small amounts of noise to the data to prevent unauthorized sharing or reselling of data (also known as salting or watermarking) or by processing to remove data artifacts and to preserve privacy. The technologies, application developers, and third-party data resellers that bring location data to the market will all vary over time. As applications are launched and discontinued, there is turnover in the data that underlie commercial data products, and the representativeness or biases associated with those data will change. Similarly, the ways that companies process data will change. For example, the algorithms used to extrapolate from a sample to a population may change, and this information may or may not be divulged to customers. Costs to buy, store, and process the data are also not fixed, which is difficult to factor into long-term study plans and budgets. Furthermore, there may be changes in the legalities and social acceptance of using LBS data that may create risks that decisionmakers will need to weigh.

Research on the use of mobile device location data for estimating visitation has identified many potential issues with data sold by third-party vendors, including biased representation of visitors, instability over time, and lack of transparency on data sources and methods applied by vendors. There are many unanswered questions about when and where mobile device data can be used to accurately estimate visitation. As such, recent studies have cautioned against using mobile device location data that are available from third-party vendors without calibration using onsite count data (Winder and others, 2025).

Remote Sensing Data Overview

Leggett and others (2017) presented several data sources under the category of remote sensing, which is the gathering of data at some distance away from the site of interest. These sources include satellite or aircraft imaging, time-lapse photography, and Bluetooth or Wi-Fi device detectors. None of these sources have been used extensively in the intervening years as a source of visitation data.

Aerial surveys were used to estimate changes in recreational visitation to the coastlines of Louisiana, Mississippi, Alabama, and Florida after a major oil spill (Tourangeau and others, 2017), but high costs have so far prevented this method from being widely adopted (Andrew and others, 2021). As an alternative to aerial surveys, drones provide a new option for monitoring visitation from the air in locations where on-the-ground methods are difficult to implement, such as marine settings. Commercial drone technology requires an operator within line-of-sight on the ground but can fly in an array of weather conditions and, compared to aerial survey equipment, can fly for longer periods of time and closer to the surface for the purpose of collecting high-resolution images (Tang and Shao, 2015; Andrew and others, 2021). Data from drones are used to monitor animal populations in remote settings with good results (for example, Johnston and others, 2017; Hodgson and others, 2018).

Satellite data have been used in a limited number of studies of recreational visitation since 2017. The temporal resolution of satellite data is constrained by the rate at which the satellite orbits over the site and collects an image (Andrew and others, 2021). Kendall and others (2021) used counts of boats identified in satellite images collected once every 4–5 days to develop a model predicting visitation at a marine protected area, although they did not calibrate their satellite counts with onsite data.

In other settings that are not recreation sites, studies have investigated the use of Bluetooth detectors to measure crowd flows (Kitazato and others, 2018; Al Anbouri and others, 2019) and building occupancy (Park and others, 2018). In principle, the approach is similar to established methods that use technologies such as passive infrared beams or magnetometers to detect people and traffic. However, this approach has only been researched as a method for measuring visitation to Federal public lands and waters at one site (Otak Team and others, 2022).

“Connected vehicles” are emerging as one new source of potentially useful information on traffic in and around Federal public lands. The data originate from modern vehicles that are increasingly equipped with GPS-enabled technologies and software used to detect the vehicle’s geolocation. The data collected by these vehicles are transmitted back to the car manufacturer who then sells the data to third-party data aggregators and vendors, who in turn resell the de-identified data as a commercial product, provided the vehicle owner has given consent. The only known study of recreational visitation using connected vehicle data was conducted in Grand Teton National Park in Wyoming by the NPS (Otak Team and others, 2022), but more studies are expected (Lawson and others, 2023). It is possible that these data (and other types of remotely sensed data) are included in mobile location datasets sold by many third-party vendors, but this is difficult to verify because the vendors generally do not disclose the sources of their composite dataset.

Remote Sensing Data Assessment

Because there are so few examples of remote sensing data used for visitation monitoring at recreational sites, especially ones that research the validity of the method (for example, by comparing remotely sensed data to direct observations), it is difficult to assess the utility of such data. As shown in table 6, satellites and aircraft have high labor and training requirements, so data from these remote sensors can be expensive to purchase, but the high accuracy and ability to obtain data at inaccessible sites may offset those costs. Additionally, as more governmental satellites make images available for free, the costs associated with satellite-based methods will likely decrease (Levin and others, 2015). The greater flexibility of operating drones compared to piloted aircraft also makes drones a promising future data source, recognizing tradeoffs in the spatial extent and resolution that can be covered using drones versus aerial platforms. Other remote sensing data sources, such as connected vehicles and Bluetooth detectors, have yet to be extensively tested as reliable sources of visitation data. These data may face many of the same issues as discussed for mobile device location data, including temporal instability, sampling bias, cost, and privacy concerns.

Community Science Data Overview

Although not discussed by Leggett and others (2017), community science approaches may be valuable for generating useful data on the recreational use of Federal lands and waters (Cheung and others, 2022; Lia and others, 2023). Broadly, community science (also known as citizen science and participatory science) involves people—who are not professional or academic researchers in scientific research—collecting data and performing other research activities. Commonly used in fields such as ecology and public health, community science programs are often geographically focused and rely on local volunteers as participants (Cheung and others, 2022). Community science is often used to enhance a research program by augmenting the amount of data collected and increasing public interest (Conrad and Hilchey, 2011; Bonney and others, 2016). Modern digital technologies like smartphone apps and other online platforms have provided more accessible opportunities for community science, which have the added benefit of easy data collection and increased reliability (Lia and others, 2023).

Community science is not yet used extensively in outdoor recreation research, though it holds promise as a tool for measuring recreational visitation and filling data gaps (Cheung and others, 2022). One motivator for implementing a community science program is to free professional staff from data collection obligations. This is potentially most valuable for land managers with large and remote management units where recreationists may visit sites even more regularly than field staff. Additionally, because recreation data is often easily observable—such as parking lot or campground occupancy—volunteers may not need special training and expertise (Cheung and others, 2022). Because people who are recreating outdoors are often enthusiastic about the outdoors and outdoor activities, they might be motivated to participate in community science and provide regular data contributions, especially when there is a clear connection between their participation and some benefit for the places where they are recreating (Lia and others, 2023).

There have been several efforts to develop technologies that facilitate participation by reducing barriers to entry and encouraging continued participation and, therefore, provide for sustained data contributions. Most recently, Lia and others (2023) developed and tested a novel method for engaging visitors in recreation monitoring using a chatbot to converse with visitors through text messages sent to the visitors’ mobile phones. A chatbot is a computer program that interacts with people by exchanging messages (Adamopoulou and Moussiades, 2020). The chatbot that Lia and others (2023) developed facilitated a community science effort specifically aimed at visitation estimation. The program encouraged visitors to recreation sites in national forests in the United States to submit counts of vehicles at trailheads. The counts reported by participants were highly correlated with ground-truth data gathered using cameras, and as much as 12 percent of visiting parties participated, regardless of whether there was cellular service at the site or not. Overall, the study concluded that chatbot technology, and community science more generally, are potentially valuable approaches for gaining high-quality data for visitation estimation.

Community science efforts commonly use custom smartphone applications. One such application is eBird, which allows community scientists to document the location and timing of bird sightings, both for their own record and to share with other bird watchers, as well as scientists. Although visitor estimation is not an explicit goal of eBird, the data are also potentially useful for studying recreation because the reporter is recording their presence in a particular place at a specific time (Echeverri and others, 2022). iNaturalist is another application for sharing sightings and identifications of organisms with the goal of documenting species’ distributions and biodiversity but can also be used to infer the presence of people.

One study has evaluated whether bird observations (as a proxy for the presence of a person) recorded using eBird are related to actual visitation. Winder and others (2025) compared estimated visitation at 13 FWS refuges with the number of eBird checklists from the same refuges and found that, although the two metrics were generally correlated, the relationship varied across refuges. Chen and others (2022) used eBird data to estimate visitation in the absence of ground-truth data and observed a decrease in birdwatching trips to municipalities in Mexico as a response to deforestation. Although not estimating recreational visits directly, several other studies have used eBird (Roberts and others, 2017; Jayalath and others, 2023) and iNaturalist (Costadone and Balzan, 2023; Cao and Hochmair, 2024) to answer research questions focused on outdoor recreation patterns and preferences. These studies demonstrate that community science may be a valuable but underused source of recreational visitation data.

Community Science Data Assessment

Many of the drawbacks of community science data for visitation estimation are like those drawbacks outlined previously for social media and mobile device location data; that is, there is a clear potential for sampling bias. Community science platforms are best at capturing patterns of recreation when performed by visitors who are interested in the purpose and objectives of the program, such as birding in the case of eBird and natural history in the case of iNaturalist. Furthermore, there will be a selection bias toward individuals who have the ability and time to volunteer, the knowledge and training necessary to perform the requested tasks, and the skill to use the technologies, such as smartphone applications, that support the science.

Beyond the selection biases, there can be concerns about the quality of data collected through community science programs (Burgess and others, 2017; Cheung and others, 2022). Programs that ask volunteers to follow specific protocols—like counting people in a specific area or at a specified time—may be difficult to follow, resulting in miscounts or other inaccuracies in the data or metadata. Yet, studies repeatedly observe that, with proper training, community science programs generate useful data, including one aimed specifically at visitation estimation (Lia and others, 2023).

Another consideration with using community science is the expertise and resources that are needed to coordinate a program, recruit participants, and develop technologies that facilitate data collection. To ease the burden, there are existing communities and technologies, such as the eBird mobile application, that can be leveraged. Still, considerable technical skills and resources are necessary to develop technologies for collecting and storing data. However, compared to the cost of onsite data collection by field staff or purchasing mobile phone location data or devices such as vehicle and trail counters, community science data may be a relatively lower cost option.