There are numerous types of heritage, including natural and cultural heritage(U.N.E.S.C.O., 2021). Natural heritage refers to heritage assets that were created by biological and/or physical processes(U.N.E.S.C.O., 2021). An example of a natural heritage site is the Giant’s Causeway and Causeway Coast in Northern Ireland(U.N.E.S.C.O., 2021). Cultural heritage encompasses human-made intangible and tangible heritage assets(U.N.E.S.C.O., 2003), such as traditions, languages, oral histories, rituals, manufactured objects, technologies, buildings, and structures. Tangible heritage assets are those that can be physically interacted with, for example, archaeological artefacts, buildings, and clothing(U.N.E.S.C.O., 2021). Heritage tourism is often driven by the presence of tangible heritage assets as a physical place or item which people can visit and see as part of a trip. Intangible heritage assets are not physical entities but exist in cultural practices and memory. Folklore, languages, dances, beliefs, and cooking techniques are all examples of intangible heritage(U.N.E.S.C.O., 2003). Intangible heritage can be directly associated with tangible heritage assets or can exist independently.

Heritage is important because it provides history and context for communities to form an identity around, as well as being valuable for telling stories of the past and how societies have evolved over time(Council, 2021). Protective measures can help to ensure that heritage assets and cultural memory are not lost. For example, in the UK, National Landscapes are areas of environmental and natural significance which are legally protected to preserve them against loss of heritage and natural assets(Landscapes, 2024). Heritage organisations worldwide can access the United Nations Educational, Scientific and Cultural Organisation (‘UNESCO’) Convention for the Safeguarding of the Intangible Cultural Heritage to source information to assist them in preserving intangible heritage assets(U.N.E.S.C.O., 2003).

In 2019, the heritage sector in England contributed £31 billion to the UK economy(England, 2019) . Heritage Visitor Attractions (‘HVAs’) are a key part of the visitor economy, including museums, heritage sites with admission fees, and free-to-enter heritage sites(Millar, 2009)(Digital et al., 2017). Free-to-enter outdoor heritage sites are a type of HVA that are often visited by people for recreation and education(Millar, 2009).

Ancient and scheduled monuments have been preserved for the appreciation and enjoyment of people today and for future generations(England, 2021). They make up and add value to the Historic Environment. The Historic Environment includes all areas of an environment changed or influenced by humans(England, 2021).

Monuments are physically built structures or areas located on a site contributing to its meaning and history(Marschall, 2010). Heritage sites are areas of land that hold significance because of their past use. Monuments can be located inside a heritage site, and heritage sites are set within a cultural and/or historic landscape(England, 2021). Some outdoor heritage sites have visitor centres with staff and volunteers to assist visitors. There are also outdoor heritage sites which do not have these facilities and are free-to-enter by visitors. Without a permanent staff and volunteer presence, these sites are less likely to have frequent condition, livestock, visitor, weather, and wildlife monitoring programmes. The owners and managers of these sites may have limited time and financial resources to monitor them, as their focus is generally on primary income streams. For example, some heritage sites can be located within farmland where farming activities are the main economic focus of the site.

Outdoor heritage sites are exposed to a range of threats, including damage by livestock, visitors, weather conditions (including extreme weather events), and wildlife. It is not always practicable to shield a site from these threats.

The raising of livestock is important to provide food supplies, produce materials, and generate income, but grazing animals can cause damage to the cultural and historic landscape(England, 2023). Livestock can cause soil erosion, ground poaching, and erosion scars in the ground. A document published by Historic England in 2023 provides guidance for livestock management when animals are grazing on archaeological sites with an aim to reduce the development of erosion scars and ground poaching from hooves(England, 2023). Grazing livestock contributes to cultural heritage and historic land use practices, and in some areas around the world, farming practices such as these are protected. In the UK, many free-to-enter outdoor heritage sites provide grazing sites for livestock; for example, cows graze within the site of Long Meg and Her Daughters Stone Circle for part of the year.

Visitor presence at heritage sites also brings threats. Erosion of footpaths around heritage sites can be caused by visitor footfall(Rodway-Dyer & Ellis, 2018). Monuments and features can be focal points for visitors and can result in areas of high footfall around the base of these structures, which can lead to concentrated soil erosion. Visitors can also disrupt agricultural industries in the area with the potential of damaging crops, spooking livestock, and blocking access routes with their vehicles whilst visiting a heritage site(Young, 1999). Another threat to heritage sites is vandalism by visitors. Writing or carving into surfaces, theft of historic material, and damaging works are all forms of vandalism(Qwatekana, 2021). English Heritage has published a guidance document looking into the consequences of increasing visitor numbers at heritage sites, which includes studying the loss of heritage assets and erosion(Heritage, 2024).

Extreme weather events can lead to increased deterioration at outdoor heritage sites. Wildfires, flooding, and hurricanes have all caused permanent loss of heritage assets at sites around the world. Flooding can cause irreversible structural damage and losses of heritage assets. English Heritage has published a report listing several of its heritage sites that are at high risk of flooding, to enable them to address this issue (Pearson, 2013).

Wild animals and plant life at outdoor heritage sites can cause deterioration of heritage assets. Microorganisms such as lichen are often found growing on stone heritage structures, and can cause chemical and mechanical erosion to stone surfaces (Inkpen & May, 2006). Other examples of common microorganisms found at heritage sites are algae, actinomycetes, cyanobacteria, and fungi. Moss can be found growing on stone surfaces, which may cause discolouration.

In summary, there is a wide range of threats to heritage sites, and it is not always possible to eliminate exposure to them. Deterioration of outdoor heritage assets cannot be eliminated; however, it can be slowed down. A cost-effective toolkit could be implemented to inform the development of heritage management strategies, which can be put in place to prolong the existence of heritage assets, including sites.

Heritage management strategies involve a range of approaches, including regulation and manipulative controls, to reduce the impact of threats to heritage sites.

A key part of heritage site management is understanding the site’s carrying capacity. Carrying capacity is an evaluation of how much pressure, for example, by visitors, can be put on a site without its assets being damaged or destroyed(Mansfield, 2018). It is a constant balance that needs to be found between the conservation needs of a site and the needs and demands of visitors (Figure 1). To avoid reaching a point where a site has exceeded its carrying capacity, management strategies need to be implemented by site managers.

UK Heritage sites are managed by a number of bodies at different levels, including the UK Government Department for Digital, Culture, Media and Sport (‘DCMS’) (which oversees the Listing of buildings and Scheduling of monuments) (Digital et al., 2013), Historic England, societies and specialist advisory groups, heritage asset owners and site managers such as English Heritage and the National Trust, as well as private land owners. International stakeholders may also be involved in the management of heritage sites, including the International Council on Monuments and Sites (‘ICOMOS’)(I.C.O.M.O.S., 2025)and UNESCO(U.N.E.S.C.O., 2025).

UNESCO designates certain sites as World Heritage Sites (‘WHS’) on an international scale. To meet the criteria to become a WHS, a site must prove its ‘Outstanding Universal Value’(U.N.E.S.C.O., 2021). Once WHS status is achieved, the site must be protected and maintained under the World Heritage Convention 1972, which is a legal framework to help safeguard heritage sites(U.N.E.S.C.O., 2021). On a national level, the DCMS creates policies and guidelines for the protection of heritage assets such as monuments (Digital et al., 2013) and the UK Parliament makes laws to protect heritage assets from damage and destruction(England, 2021).

Managers of heritage sites can implement regulatory controls, such as admission fees, opening hours, restricted routes, and site closure on certain dates and times. However, it is often difficult to put these controls in place at free-to-enter sites, many of which are not staffed. This is why legal guidelines for the protection of heritage sites and monuments are vital for their preservation. The DCMS published a paper outlining the need to monitor the loss of heritage assets(Sagger, 2021).

Free-to-enter sites often implement manipulative controls to encourage visitors to act a certain way without the presence of staff or volunteers(Mitchell, 1979). Visitor flows are the routes frequently taken by people through a site and can be recorded by observing how people behave (Beritelli et al., 2020). Once a heritage site’s visitor flows have been identified, manipulative controls can be put in place to encourage shifts in the visitor flow to reduce the risk of further damage (Cvelbar et al., 2018).

An example of manipulative controls is the positioning of interpretation panels at certain parts of a site to encourage people to visit them, rather than other, more vulnerable areas of the site(Mitchell, 1979). Interpretation panels are commonly used for presenting contextual information about a heritage site to the public, and can also influence their chosen route around a site. These controls are particularly useful when there is no permanent staff or volunteer presence on site. Free-to-enter outdoor heritage sites have less frequent monitoring and less staff and volunteer presence. These were therefore identified as sites that would benefit from more frequent monitoring to help inform future site management decisions. A cost-effective method to monitor sites that is accessible for beginners and provides clear results would be beneficial to small heritage organisations, particularly those that are volunteer-led. The results of such monitoring should help to inform future decisions in the choice of heritage management strategies to implement at sites.

The PhD research aimed to design a method for recording the physical condition and visitor routes at five selected free-to-enter Cumbrian outdoor heritage sites. Then, to develop a toolkit based on the resulting method.

The research objectives were firstly to write a literature review and evaluation of existing research to inform the design of the methodology. Next, a number of sites were selected for the research based on chosen criteria, and then a method for monitoring the selected sites was devised. Once the method was designed, surveys were carried out at least four times at each site, then the results were processed, interpreted, and compared. Finally, vulnerable areas at the sites were identified, suggestions for conservation strategies were made for each site, and the toolkit was designed.

A suite of criteria was used to assess the suitability of sites for study. These included access to the site, presence of a monument on site, proximity to roads and parking areas, type of regulatory and manipulative controls, type of site, site ownership, and management. It was also important to gather data about different types of heritage sites, so a range of site types were considered for selection. To create a comparative study, some sites were located inside the English Lake District UNESCO WHS and National Park(Park, 2022), whilst others were outside the boundary. All the selected sites were located in Cumbria, in northwest England.

A vital stage in confirming the selected sites was gaining permissions from site owners and managers. After the selection process was complete, Ambleside Roman Fort, Castlerigg Stone Circle, Kendal Castle, Long Meg and Her Daughters Stone Circle (‘Long Meg’ hereafter), and Shap Abbey were selected for the study (Figure 2). These five sites contain different types of monuments from a range of time periods, and the size, ownership type, and topography of each site vary. Within the selection are two stone circles, which have different ownership and management types and vary vastly in size.

Numerous site surveying techniques were reviewed, including terrestrial photography, camera traps, video surveillance, aerial LiDAR scanning, aerial photogrammetry (see Glossary), and using satellite imagery. The most cost-effective method from those reviewed was aerial photogrammetry using an Unmanned Aerial Vehicle (‘UAV’), which could capture details of the ground surface and changes to the site, whilst being relatively low cost for the quality of the results achieved.

To be able to survey the sites, an aerial photogrammetry method was chosen(England, 2017), so that the area could be assessed by producing 3D models of the site through photography. Aerial photographs were captured using a UAV with an integrated ‘red, green, and blue’ (‘RGB’) camera. A UAV weighing less than 250 grams was used, as this is within the Open Category (Class A1) of UAV, as designated on the Civil Aviation Authority website (Authority, 2025). This type of UAV was chosen to make the aerial surveys accessible to first-time UAV pilots, and for them to have the confidence to carry out the survey using low-cost and lightweight UAVs. Permission to use the five sites and to fly a UAV over them was granted by the landowners and site managers. A Research Ethics application was approved by the University of Cumbria’s Ethics Panel. Risk Assessments were carried out and approved in advance of the commencement of the fieldwork.

To capture the correct set of aerial photographs to produce a 3D model, Dronelink (Dronelink, 2024) was used to design overlapping grid patterns (Figure 3A-B) and flight paths for the UAV. The flight path ensured photographs would have overlap (forward 80% and side 70%) (Bhambri, 2016) so that the photogrammetry software, Metashape Standard (Agisoft, 2025), could align the photographs to assist it in stitching them together to produce a resulting 3D model. Figure 3B shows how overlapping photographs are taken to be used to generate a 3D model. The flights also included taking photographs with different camera angles, -90° (vertical) and -70° (oblique), in order to capture details of taller parts of the monument on site(England, 2017).

The altitude of the flight impacted the detail which could be collected through the aerial photography (known as the ‘Ground Sampling Distance’) and affected the flight time(Pix4D, 2025);(England, 2017). It was important to consider flight time, as the small UAV only had approximately 30 minutes of battery life and would sometimes need to be used for two flights in one day(England, 2017). A lower altitude of flight would gather more detailed photographs; however, this would take longer than a higher altitude flight, as it would require more photographs(Pix4D, 2025). Testing was carried out to determine the best altitude for each site.

Some monuments were large and in more topographically complex settings, so only certain sections of these sites were surveyed with the UAV. These sites were Kendal Castle, Long Meg, and Shap Abbey. If the project were to be carried out again, with the experience in UAV piloting gained through this fieldwork, it would be possible and preferable to survey the entire sites.

Aerial surveys were carried out multiple times at each site, with the aim of them being once a month for six months of the year. However, weather conditions sometimes prevented UAV flights from taking place, so flights had to be rescheduled for a date with dry conditions. The repeated process of collecting photographs over time allowed multiple 3D models of each site to be made, which could then be compared to one another to identify areas of physical change and interesting features on the site that may not be visible from ground level.

The aerial photographs were transferred from the SD card in the UAV to a computer, where they were each saved into a folder named after the date of the photography. The images were then uploaded into Metashape Standard (Agisoft, 2025), and then multiple stages (alignment of the photographs, building the dense cloud, building the mesh, and building the texture of the 3D model) of processing were completed to create a digital 3D model. The resulting model was saved in the dated folder, and then the point cloud (the individual points that form the basis of a digital 3D model) was exported for use later in CloudCompare (CloudCompare, 2025). This process was repeated for every aerial survey that took place, generating multiple 3D models of each site. Figure 4 shows an example of a 3D model generated of part of the Long Meg heritage site, focusing on the tall standing stone pictured in Figure 5a.

A range of methods for recording visitor numbers and visitor routes at sites was considered. Using a Global Positioning System (‘GPS’) in conjunction with 3D models of a site could have been a useful method, but the GPS devices were costly (Zubiaga, 2019). Alternatively, photographs that had been geo-tagged for social media to locate spots that had been visited by people could have been used (Miyasaka, 2018). Another possible method involved visiting sites with a group and observing their behaviour in-situ (Roura, 2011). Infrared monitors and/or camera traps could have been used to count visitors, but these only cover certain points in a site, rather than gathering route data, and these pieces of equipment need to be regularly checked and maintained, including checking the batteries (Pettebone, 2010)(Lupp, 2021).

After reviewing possible different ways of recording visitor routes and numbers, the methods chosen for this research were those that used low-cost equipment and techniques that were easy to learn. This approach was used so that staff and volunteers could carry out the surveys with minimal training time and negligible equipment costs. These factors are valuable to heritage organisations with limited budgets. The chosen method was based on the paper-and-pencil method (Yalowitz & Bronnenkant, 2009)and tally counting (Harnik & Kimball, 2004), taking place in situ on site (Roura, 2011).

Visitors were counted using a handheld click counter (Harnik & Kimball, 2004), and their routes were recorded on site during set time periods. These time periods varied from morning to afternoon and took place on significant dates of the year, such as the Summer Solstice (21st June) and Beltane (1st May), and during school holidays as well as term time. As far as practicable, recording of visitor numbers and routes at the sites took place over 6 months. Visitor routes were recorded using paper maps and drawing on the route taken by a visitor in pencil. The visitor route lines were then digitised by scanning them and tracing the image using Illustrator (Adobe, 2025), so that they could be compiled and analysed.

The data gained from recording visitor routes helped to determine areas of high foot-traffic on sites and popular times and dates of visits to the sites. By digitising the data, the routes could be overlaid onto the 3D models produced by the site surveys (Zubiaga, 2019). This allowed for robust conclusions to be made about the impact of visitor routes on the physical sites.

Examples of the data collected are shown in Figures 5b and 5c. Figure 5b shows a map of the site of Long Meg, with the tall standing stone (Figure 5a) circled in red. This tall standing stone stands outside the main stone circle structure and has circular carvings on its surface. It is seen as a focal point at the site. Figure 5c shows the visitor routes overlaid onto a section of the map of the site, covering the area surveyed with the aerial photogrammetry.

In addition to the recorded visitor routes, visitor numbers were recorded at each site, along with notes on weather conditions and significant dates, such as the Summer Solstice (21^st June) or Beltane (1st May). Table 1 presents the other data collected at Long Meg related to visitor numbers.

As multiple datasets and data types had been produced, it was necessary to combine them to analyse them. This was achieved in various ways. First, each 3D model was visually analysed and annotated, highlighting areas of exposed soil (ground erosion), standing water, and other relevant topographical features. These were then compared to one another in order to see if the areas of ground erosion changed in size, or if the amount of these patches of exposed soil changed.

Next, the digitised visitor route lines were overlaid onto the aerial view of each site’s 3D model.Figure 6a shows an example of visitor routes overlaid onto a 3D model of Long Meg. Overlaying the digital layers was completed using PowerPoint (Microsoft, 2025), setting the 3D model as the back layer and the visitor route lines as the front layer of the image.

The 3D models were also analysed using CloudCompare (CloudCompare, 2025) by taking two point clouds (a group of data points which form the basis of a 3D model) of the same site and inputting them into the software. The two separate point clouds were aligned together, and then the software processed them to produce a coloured model, where the colours indicate the degree of difference between the two models. Figure 6B shows an example of a CloudCompare (CloudCompare, 2025) model produced through this research, comparing two point clouds from Long Meg, one from 3rd April 2023 and the other from 12th November 2023. Digitised visitor routes could also be overlaid onto CloudCompare (CloudCompare, 2025) models, providing the ability to further compare the visitor routes to physical changes recorded on the sites.

By carrying out the visitor surveys at different times of day across different seasons, it was possible to identify quiet times on site and more popular times, including special days of the year. Using the digitised visitor route maps, it was possible to identify the total number of routes recorded at a site and the number of routes that intersect areas of ground erosion.

A key strength of this method for monitoring the condition of outdoor heritage sites and visitor behaviours is its replicability. The researcher can reuse the equipment as many times as needed with no extra cost. This would appeal to voluntary groups and heritage organisations with limited resources wishing to gain data about their site over a long period of time. The visitor recording methods are low-cost and easy to teach to staff and volunteers, and the UAV flight uses entry-level equipment and easy-to-use flight planning software. The successful testing of the method by Nenthead Mines Conservation Society showed that the instructions in the toolkit were easy to follow and produced useful results. In addition, the method can be used at a range of site types. It is not confined to only being used at sites in Cumbria or the UK, but has the possibility of being used at heritage sites worldwide.

The potential benefits of this method include the use of the resulting images outside of the main purpose of the research itself, including replicating 3D models for outreach and educational projects, widening access to the heritage site and its associated stories to remote audiences. The use of this method by small heritage organisations will help align their records with those of larger organisations such as Historic England, who use photogrammetry to record a range of different heritage assets under its guardianship (England, 2017).

The method records useful data, providing a record of the site and popular areas of it, which can be referred back to when comparing future results of site surveys. It provides results that can be interpreted and analysed to better understand the potentially vulnerable areas of heritage sites. Having this knowledge can then aid in decision-making around implementing strategies to better manage the protection of vulnerable areas on site, such as the installation of interpretation panels in specific areas to draw visitors away from vulnerable areas. The visitor route data is also useful in assessing the efficiency of existing management strategies, such as the location of site entrances and their impact on visitor flows (Cvelbar et al., 2018), or how effective the placement of interpretation panels is at channelling foot-traffic.

It also provides a set of documented evidence of visitor numbers (how popular the site is) and the potential visitor impact on the site. These results can be used to support the organisation’s future funding applications and interpretation strategies. In addition, the method successfully provides a method to monitor losses of heritage assets over a period of time, which is seen as a requirement for better understanding the future of our cultural heritage(Sagger, 2021).

There are a range of weaknesses of this method, which future developments will aim to reduce. To fly a UAV, dry and non-windy weather is required, so it may not be possible to gather aerial images during wetter seasons. This was a limitation experienced when the researcher carried out the fieldwork. In addition, a budget needs to be set aside for the equipment and software costs needed for this research, as even though it was designed to be cost-effective, it still requires the purchase of equipment and a time commitment.

The chosen method has limitations in the level of detail it can provide at ground level. As shown in Figure 7b, some details are lost in the generation of the 3D models; for example, where the grass level meets a stone, the 3D model blurs these together. This loss of detail in the 3D models limits the ability to interpret results on a small scale (i.e., at the centimetre level), but the 3D model still provides enough detail to identify changes on a site-wide level. The RGB images captured by the UAV camera do not measure the distance from the camera to the ground surface, but the use of multiple photographs to create a 3D model (photogrammetry) provides a topographic rendering for visual analysis. There are sensors available that would collect topographic measurement data, for example, LiDAR(Uribe, 2021). However, these were more expensive and less accessible for entry-level users, so they were not used in this research. The repeated use of LiDAR would build a portfolio of models where physical changes to the site could be detected on a finer scale; this level of detail is not required for this research project.

The quality of 3D models generated via aerial surveys (photogrammetry or LiDAR) depends on the Ground Sampling Distance (‘GSD’)(England, 2017). The GSD is calculated using the sensor’s specifications and the altitude of the UAV capturing the data (Pix4D, 2025), with higher altitudes producing lower resolution data and lower altitudes producing higher resolution data (England, 2017). Flying at lower altitudes over a set takes longer than flying over the same area at higher altitudes, and therefore uses more battery power (Pix4D, 2025). This is also a consideration when using aerial surveying techniques.

The paper-and-pencil method for recording visitor routes is based on the observer’s perception. This was carried out by one researcher for each of the sites, so the results were recorded in a consistent manner; however, there may be a risk that if routes were to be recorded by numerous different people, they would be inconsistent. A drawback of the paper-and-pencil method was also the time required to digitise the recorded routes (Yalowitz & Bronnenkant, 2009).

This method has a range of opportunities generated alongside the main body of results. The 3D models produced can have a range of uses, not just for evidencing physical changes at sites. Other uses include marketing for heritage sites, by showing the 3D models on screens at the site or on the site’s website. 3D models can be used for virtual tours of the site online or in Augmented and/or Virtual Reality. The 3D models could also be 3D printed and used for merchandise for heritage sites, the sales of which could raise money to help maintain the site.

In addition, the toolkit can be adapted for other issues that are not yet covered by it. For example, it can be used specifically to monitor water damage on sites. In addition, the aerial images gathered can be used for wider research in fields such as archaeology, archaeoastronomy, digital heritage, ecology, and more. The research can provide a basis for cross-disciplinary studies into heritage sites.

Improving the method and the toolkit in the future, whilst ensuring it remains cost-effective, may be possible as more sophisticated technologies become cheaper, for example LiDAR on UAVs, which will enable greater detail of sites to be captured for the generation of 3D models and can be used in conjunction with the existing photogrammetry method to enhance the quality of the data collected(England, 2017).

There are also a number of threats to the method. For example, in the future, the CAA may increase its requirements regarding the flight of UAVs. In addition, some of the software providers may change the nature of payment, i.e., switching from a one-off cost to a subscription plan for use of the software, which may prove to be more expensive. These threats may increase the time and cost of monitoring outdoor heritage sites.

The site monitoring process requires a person to pilot the UAV, so it is useful to have more than one person trained to do so, in case a pilot is unavailable to carry out the flights on a regular basis. The visitor monitoring process requires data to be gathered in-situ over periods of time in the outdoors, which may not be appealing to some staff and volunteers, which could impact the amount of visitor data collected.

Overall, the strengths of the method outweigh the weaknesses, as it provides a cost-effective way of monitoring the loss of heritage assets at outdoor sites, as confirmed by the external trial user, Nenthead Mines Conservation Society. All of the sites presented different challenges and characteristics; however, the method applied was able to successfully capture the physical site condition and the visitor data to a consistent quality. The period of monitoring could be made less frequent, or remain the same, but continue on an ongoing basis, so eventually a dataset allowing comparisons in physical changes to a site and visitor behaviours over a period of years, rather than months, can be made.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sector/ individuals.

Ethics application ref 22/07. Issued by the University of Cumbria Ethics Committee.

The author(s) declare(s) that there is no competing interest.