Integrating Energy Transition into Protected Landscapes: Geoinformatic Solution for Low Visual Impact of Energy Infrastructure Development—A Case Study from Roztoczański National Park (Poland)

Abstract

Energy transition, encompassing the development of renewable energy sources and associated power transmission grids, may significantly impact landscape visual resources, particularly those legally protected. Large-scale energy transitions require a mandatory visual impact assessment procedure, which utilises proximity and visibility analyses to comply with legal regulations and achieve minimal visual impact. While design stage proximity provides full compliance with the given country’s legal acts, the following visual impact analysis is more about demonstrating the low visual impact of design variants. Notably, at the energy infrastructure planning stage, the information on visual landscape resources remains insufficient; hence, avoiding conflicts is particularly challenging. To address this issue, a geoinformatic framework for Visual Landscape Absorption Capacity (VLAC) is proposed to support the sustainable planning of energy infrastructure right before the visual impact assessment. The framework involves identifying sensitive and valuable vantage points across the analysed landscape and determining the dimensions of energy infrastructure to be developed in a sustainable way regarding visual landscape resources. This paper presents a case study from Roztocze National Park in Poland, a protected area under significant pressure from solar farms and accompanying power transmission lines development. The results provide a critical assessment of the existing transmission lines (110 kV) and solar farms in relation to landscape visual resources, while also identifying three key areas where further infrastructure development can occur without landscape resource degradation. The framework geocomputation is based on digital elevation models, enabling easy replication in other locations to support the decision-making process and facilitate sustainable energy facility planning, thereby minimising potential conflicts with landscape resources.

1. Introduction

Energy transformation is increasingly recognised as one of the most pressing challenges of our time [1,2,3]. This shift aims to overhaul the global economy, steering it away from high-emission practices towards a model that is not only diversified but also sustainable, particularly in its use of natural resources. In Europe, this shift is being guided by Directive [4], which sets the framework for the continent’s energy future.

Focusing on Europe, this transformation is largely driven by ambitious climate targets set by the Council of the European Union. This is particularly true for Poland, wherein this research paper case study is being located. The Polish energy sector stands as one of the most carbon-intensive in Europe, with fossil fuels still accounting for a staggering 85% of its energy production. This leaves Poland’s energy infrastructure among the most emissions-heavy in the region [5].

Yet, the decarbonisation transition is on the horizon. Poland, like many other European countries, is following two main energy pathways. These include the development of nuclear power [6] and the expansion of renewable energy sources (RESs). The latter is the more prominent of the two [7,8], although it still lags behind other EU-27 countries [9]. Both transition strategies require significant development of energy infrastructure. This includes, in particular, the expansion of power transmission lines (or grid).

Importantly, the shift towards RESs also brings a less-discussed, yet critical, consequence: changes in landscape physiognomy, which are the focus of this paper. Basically, the impact of the energy transition on environmental aspects is extensively addressed in the scientific literature [10,11,12,13]. For instance, the repurposing of agricultural land for industrial energy production [14,15] is associated with numerous alterations in species composition and changes to animal migration patterns, which are not always predictable [16]. On the other hand, the change is not necessarily a negative—research has proved that photovoltaic farms (PVs) can actually support increased biodiversity, especially among small avifauna, defying common expectations [17].

The question underlying this study is whether a similar synergy between the development of energy infrastructure and natural—specifically landscape—resources can be achieved. This paper explores the impact of the energy transition and, as a result, the expansion of technical infrastructure on scenic landscape features, such as viewpoints [18]. Particular attention is given to the influence of high-voltage transmission lines (HV-lines) and photovoltaic farms (PV-farms) on landscape physiognomy.

The analysis is conducted using the case study of Roztoczański National Park (RNP), an area subject to high investor interest in PV installations due to favourable solar conditions [19] and implemented local strategies towards climate change adaptation [20]. Notably, the implementation of energy investments should be agreed with the primary conservation objectives of the RNP, namely the protection of the landscape scenic resources [21]. Energy security in a country (in this case, Poland) often takes precedence over landscape values, and the energy transition is being enforced through a special act (e.g., [22]), simplifying environmental procedures. For this reason, the framework proposed in this research, despite such special acts, contributes in practice to the protection of the scenic resources of the landscape.

The social assessment of the energy transition process necessitates further commentary to justify the need for protecting scenic values. While the increasing need for energy transformation has led to a growing public acceptance of PV-farm development within the landscape [23,24,25,26,27,28,29,30,31], this trend has become so pronounced that terms such as “photovoltaic landscape” [32,33] and typologies of solar energy landscapes [34] have emerged over the last decade. However, from the perspective of the Polish landscape typology [35], such typology or lexical approach appears entirely justified, as it represents an attempt to mitigate landscape degradation—degradation that, from a landscape typology point of view, may be classified as an “industrial landscape”, provided that its entire structure and function has been shaped by human activity [36]. According to the author, it is more accurate to refer to a landscape with PV-farms rather than a “photovoltaic landscape”, unless the entire landscape is occupied by photovoltaic infrastructure. Therefore, this study refers to the physiognomic values of the cultural landscape [37], which are perceived mainly through the sense of sight—recognised as the dominant modality in the perception of the multisensory landscape [38,39]. Above mentioned lexical and terminological issues, not exempt from the obligatory visual impact assessment procedure (VIA) [40] and, if necessary, implementation of visual mitigation techniques of energy infrastructure facilities in order to protect the landscape’s visual resources as best as possible. This also constitutes the primary focus of this research, ensuring that landscapes with photovoltaics, or more broadly with energy infrastructure, do not become degraded, especially since, in some cases, this must also occur in nature conservation areas.

Consequently, the issue addressed in this article is illustrated through the RNP case study, while being situated within a broader geographical context [41] concerning all protected areas engaged in the energy transition. Accordingly, the proposed framework for integrating the energy transition into protected landscapes has the potential for global application. It constitutes a novel approach supporting the spatial planning of energy infrastructure development and serving as a prerequisite to the VIA procedure.

In energy transition spatial planning, precision is achieved through the use of off-the-shelf GIS technology and GIS science [42]. The VIA (visual impact assessment) relies on geospatial calculations, including visibility analysis, commonly executed using the viewshed algorithm [43]. It is surprising that, despite the widespread availability of GIS software, including open-source options [44] and a number of scientific publications on landscape visual absorption capacity [45,46,47], guidelines for HV-line development [48,49,50] still fail to incorporate a solution based on Jacobs’ (1975) [46] theory. This is particularly noteworthy, as the absorption capacity can technically be quantified using the viewshed GIS algorithm. Importantly, to maintain consistency with the multisensory understanding of landscape [51], i.e., space perceived not only through the sense of sight, this work clarifies the concept of landscape capacity, emphasising that it pertains to the visual capacity of the landscape and that is why the Visual Landscape Absorption Capacity (VLAC) term is used in this work.

The VLAC quantifies the landscape’s ability to absorb new features without compromising its physiognomy [45]. The concept of VLAC responds to the demand for more sensitive measures of the total range of costs and benefits associated with landscape development proposals, as well as the need for new evaluation techniques. It is also linked with the non-monetary aspects of landscape aesthetics, as referenced in the science of cultural ecosystem services [52,53]. The VLAC is also expected to be incorporated into policy and decision-making priorities [46].

In practice, the calculations focus on determining the maximum permissible height of a vertical structure so that it remains invisible from designated viewpoints. The method for determining this height, i.e., the height above ground level (AGL), is described in greater detail in the methodological section. One of the few analyses of landscape capacity type is the Visual Threshold Carrying Capacity proposed by Kyushik [54]. First, it identifies viewpoints and viewsheds, next determines the permissible level of landscape development. Also, Amir and Gidalizon [55] did not employ GIS tools in the calculation of absorption capacity; instead, they utilised an expert-based method. The approach was based on criteria derived from the group of physical changes and the group of biophysical characteristics and resulted in the printed map depicting six levels of visual absorption capacity. Regarding GIS-related VLAC calculations, Ozimek [56] employed the viewshed algorithm to determine the AGL height, interpreting it within the context of the potential absorption of new building development. The approach introduced by Ozimek [56] is further developed in this paper. Notably, according to the author’s knowledge, no more scientific evidence can be cited regarding the use of AGL analysis in the protection of landscape visual resources, especially those in protected areas. Other original studies focus on animal ecology [57], though these fall outside the scope of this paper. While viewshed analysis and its probable [58] and cumulative [59] version implications are well established in the scientific literature, including works discussing the visual impact of RESs [60,61], less attention has been paid to the possibility of calculating the AGL and interpreting the results within the VLAC context.

Moreover, the aforementioned comment regarding the limited recognition of VLAC also applies to existing guidelines for planning the routing of power lines [35,48,49,50]. These documents consider a wide range of landscape conditions and offer instructions for conducting a successful visual impact assessment (VIA). However, they do not offer spatial information indicating where HV-lines should be routed to minimise visual impact, which is addressed in this research study.

Recent research in the field of transmission line routing has primarily focused on the optimisation of the networks, particularly grid flexibilities that enhance the integration of variable renewable energy sources [62,63,64], or improved battery storage within RES networks. Moreover, most of these studies rely on synthetic network models, and only a few, such as Kenz [65], examine real-world electricity transmission systems. The contribution of GIS technology to these studies is significant, but it focuses primarily on quantifiable economic aspects [66], without accounting for the costs associated with landscape services [67], specifically those cultural [68].

This paper addresses a research gap by applying the VLAC theory [45,46] within a geoinformatics framework aimed at achieving low visual impact in energy transition, specifically HV-line and PV-farm location planning. The advantage of this framework lies in its precedence over the standard VIA procedure, offering spatial information through a suitability map. Although this spatial planning solution does not eliminate the issue of visual impact caused by energy infrastructure, it enhances the likelihood of energy infrastructure spatial expansion without conflict with visual landscape resources.

2. Methods

The framework aims to propose methods for advancing the energy transition without compromising other resources—particularly the visual quality of the landscape. Its conceptual foundation lies in identifying areas valuable for their visual landscape characteristics and, from this perspective, determining the Visual Landscape Absorption Capacity (VLAC). The general conditions outlined above define the structure of the framework, while allowing flexibility in the application of geocomputational tools and spatial analysis settings. These can be adapted based on the format of input data, the desired level of detail, and the selected mapping scale.

The methodological framework comprises three sequential stages, which collectively support a comprehensive analysis of VLAC in accordance with which the energy transition is being designed at the landscape scale (Figure 1). First, the visual landscape resources have to be identified, with a focus on the determination of vantage point (VP) locations and observation azimuths. The VP establishes the perspective for subsequent visibility analysis and consequently, identifies areas where the visual impact of energy infrastructure is undesirable. The existing energy infrastructure mapping is a parallel task utilising open-access spatial databases (not distinguished in the flowchart GIS routine task).

The second stage involves calculating the landscape’s visual absorption capacity using GIS software, specifically the viewshed algorithm. The third stage results in a suitability map, distinguishing the following: areas unsuitable for energy transition due to the high visual sensitivity of landscape scenic values; conditional areas where infrastructure may be permitted within specific height constraints; and areas deemed suitable for vertical energy investments that would not adversely impact scenic resources. The spatial knowledge of these three suitability zones, although not yet required by any legal regulation, supports investor preparedness for the mandatory visual impact assessment procedure and helps minimise landscape intrusion from the outset.

The proposed framework provides a structured yet adaptable methodological path that foregrounds visual landscape resources in the spatial planning of energy transition. A detailed description of the tools, input parameters, and datasets used in its application to Roztoczański National Park is provided below.

2.1. Method for Mapping Energy Infrastructure

The preparatory step of the framework involves creating an inventory of existing energy facilities, typically using open-access spatial databases and off-the-shelf 3D GIS software. The framework was developed using ArcGIS Pro v3.4, chosen for its efficient handling of 3D point clouds. Importantly, the entire framework can be replicated using open-source software such as GRASS or SAGA GIS, particularly for DSM generation, and QGIS for viewshed analysis below the horizon line. Additionally, any geoinformatics programming tools may be employed for this purpose. For the RNP case study, three types of energy infrastructure elements were mapped: existing high-voltage power lines (100 kV), existing photovoltaic farms, and cadastral parcels with permits already issued for future PV installations.

The precise location of the high-voltage power lines (HV-lines) was obtained from Poland’s open topographical object database (BDOT, 1:10,000), specifically the SULN02 vector layer (Esri shapefile format). Additionally, as a base map for HV-line mapping, an orthophotomap was used (RGB, 0.25 m GSD, acquired in 2024). The datasets were downloaded from geoportal.gov.pl, using the local coordinate system EPSG:2180. To measure the height of HV-lines and power poles, the 3D point cloud was used. The point cloud data were acquired via airborne laser scanning (ALS) in 2015. The acquisition was commissioned by RNP and made freely available for research purposes through RNP services. The case study point cloud dataset comprises 414 LAS files (totalling 300 GB of disk space). The data were classified in accordance with LAS version 1.2 standards [69].

The classified 3D dataset was filtered to generate a virtual representation of the energy infrastructure, recorded within the first class of the point cloud (unsigned points). Due to the relatively short length of the analysed HV-lines (45.3 km), their height measurement was not automated [70] but made manually using a 3D view of GIS software, specifically a local scene using the EPSG 2180 coordinate system. The measurement process was initiated by placing six 2D points at equal intervals between consecutive pylons, with an additional point at each pylon’s peak. Next, using the Z information from the 3D point cloud, the points were interpolated to the ground surface and manually adjusted to match the virtual height of the HV-lines; any errors were corrected manually. The illustrative visualisation (Figure 2) demonstrates how the height of the energy infrastructure can be represented using a 3D point cloud at 12 points/m² density and conjunction with vector layers. The inventory of HV-lines was executed in accordance with the technical requirements of the viewshed algorithm, which is designed to operate preferably on vector point data. Additionally, a 7.7 m buffer zone was created along HV-lines to verify the allowable 2D distance to nearby buildings, as specified by the Polish Norm [71]. The results of this analysis are presented in the quantitative summary in the Results section (Section 4.1).

Regarding energy infrastructure, ground-based PV-farms were also inventoried. In contrast to HV-lines, the BDOT does not include this type of renewable energy facility, and as such, their inventory was conducted manually using high-resolution orthophotomap (2024). PV-plants are typically situated within specific cadastral parcel boundaries; therefore, the orthophotomap was first overviewed, and subsequently, the boundaries were imported as PV plots once identified. The cadastral parcels’ boundary import was executed using the Q-GIS plugin released by EnviroSolution, entitled “GUGiK Data Downloader” (version 1.2.7). In this case study, the entire area of the cadastral parcels is fully covered by PV panels. In addition, the locations of planned PV installations were incorporated into the database; this data was downloaded from the geoportal https://polska.e-mapa.net (accessed on 21 May 2025), which provides information based on data from municipalities across the country. All PV-plants (the cadastral parcels) were mapped as 2D polygons (EPSG 2180), distinguishing between existing and planned installations. Regarding the vertical dimensions of PV panels, a height of 3 m was adopted, in alignment with the typical PV height above ground level in the analysed case study.

2.2. Method for Mapping Visual Landscape Resources—From Viewpoint to Vantage Points

The starting point of the framework is the mapping of visual landscape resources (Figure 1). In this case study, the task was carried out using a 3D point cloud and RNP map documentation, in accordance with RNP’s Landscape Protection Plan [72].

Admittedly, visual landscape resources can be inventoried using traditional field methods; however, these methods require significant time and human resources. Therefore, the framework involves the use of geospatial data, with field studies being reserved solely for the assessment of geoprocessing results. Thus, geoprocessing constitutes the starting point of this section, specifically the conversion of the 3D point cloud into a raster format, as required by the subsequent viewshed algorithm. By filtering LAS classes, a Digital Terrain Model (DTM) and a Digital Surface Model (DSM) were generated, with a pixel size of 0.5 m (ref. for LAS to raster conversion). The pixel size was chosen to match the spacing of the 3D point cloud, which ranged from 0.16 to 0.36 m. A pixel size smaller than 0.36 m could have resulted in the absence of data pixels in the DTM or DSM rasters; consequently, a 0.5 m Ground Sampling Distance (GSD) was adopted, with an additional margin. The DSM layer was used as the input layer for the subsequent visibility analysis, specifically the Visual Exposure Index (VEI) calculations, viewpoint location, and vantage point (VP) delimitation, as well as above ground level (AGL) analysis.

Basically, viewpoints, scenic routes, and viewing axes are considered significant visual landscape resources [18], among which viewpoints are regarded as the most essential, as scenic routes are composed of multiple viewpoints. Consequently, viewpoints were designated for further detailed analysis. The RNP Landscape Protection Plan delimited a set of 20 viewpoints, which was adopted without major changes; however, any set may be extended to include other valuable views within the framework, if deemed necessary. Unfortunately, the locations of RNP viewpoints were provided only on a 1:25,000 scale map; therefore, their precise and unambiguous identification within the GIS software was not possible due to insufficient spatial accuracy. Thus, viewpoint location [72] was treated as a proxy requiring verification—the procedure for delimiting the viewpoints’ precise location was based on the VEI and tourist routes alignment [73] in reference to the averaged location pointed in the RNP Landscape Protection Plan.

The VEI identifies areas within the landscape that offer high visual exposure, specifically locations from which extensive panoramic views can be observed. VEI is being calculated on a DSM raster using a grid of potential observer points (POPs) and cumulative viewshed [59] analysis. Importantly, the POPs are excluded from forested areas, building rooftops, and other surfaces that do not constitute natural or typical positions from which a person would observe the landscape. Regarding the RNP case study, the POP grid spacing was set to 100 m, points were elevated 1.6 m above ground level, and incorporated into a cumulative reversed viewshed calculation. The resulted raster’s pixel values corresponded to the total number of POPs from which the analysed pixel can be observed. By assuming intervisibility between the observer and the observed area [74], the VEI, through its higher values, indicates preferred locations for siting viewpoints [73]. Importantly, the VEI values were not the sole criterion within the VP delimitation process. The second factor is the distance to the tourist routes network; it was assumed that a viewpoint must be accessible to tourists—either via a signed trail or other path. Data on tourist routes were manually vectorised based on two sources: the Tourist Map of RNP, issued at a scale of 1:50,000, and the geoportal of the Lublin province (gis.lubelskie.pl).

At last, following the VEI spatial distribution, tourist routes, and viewpoint locations, the manual delimitation of vantage points (VPs) was executed using a high-resolution orthophotomap. Preference was given to landscape-specific sites (e.g., tourist shelters or trail intersections), which were simultaneously evaluated in terms of their above-average visual exposure. Prior to the final VP selection, repeated 2D viewshed analyses were conducted for each candidate location to ensure the widest possible panoramic view was provided. The agreed VPs were saved in SHP format (PUWG 2180) and exported as coordinates to a GNSS receiver. Using these coordinates, the designated locations were navigated to in the field. The surrounding landscape was then assessed, and the positions of the proposed VP were adjusted accordingly.

Each VP location was verified with sub-metre accuracy using the Hi-Target Q-star 8 RTK-GNSS receiver; for the RNP case study, a panoramic photo from the VP position was also taken. The finalised VPs were subsequently used as vantage perspectives for VLAC calculations (see Section 2.3). Since the VP set may be freely modified by the framework users (e.g., to include other visually significant locations such as those offering views of cultural landmarks), the term vantage points is adopted herein, in place of the more landscape-specific viewpoints.

2.3. Method for VLAC Calculation

The VLAC calculations are executed for a previously defined set of VPs, first individually for each VP, and then cumulatively. Consequently, an ID was assigned to each VP, as they were utilised as input data for subsequent analysis.

VLAC refers to the landscape’s ability to incorporate new elements without compromising its physiognomic identity [45,46]. In practice, the maximum allowable height of the new element (herein HV-line or PV-plant) is calculated to ensure it remains invisible from the VP perspective. In 2D mode, these calculations are made within a viewshed routine, resulting in both visible and non-visible pixels, for which the above ground level (AGL) raster is optionally generated. The principle of AGL calculations is graphically presented in Figure 3.

The AGL pixel values correspond to the height at which the object must be elevated to become visible from the VP location. When considering a new vertical investment (e.g., an HV-line pylon or a PV-plant), exceeding the AGL raster value will result in more or less visual impact. The impact level depends on three main factors: how significantly the AGL value is exceeded, the number of VPs with exceeded AGL, and the distance from the VP. Furthermore, factors like colour, transparency, shape of the infrastructure object, weather conditions, and others contribute to the visual impact level. However, such detailed considerations belong to the visual impact assessment procedure, which is beyond the scope of the framework described in this research, except for some insights regarding the number of VPs affected by the visual impact. In simple terms, the VLAC calculation method is based on AGL analysis from the VP perspective. For the RNP case study, the AGL raster was calculated for each VP from a set of 20; the calculations were performed on the GPU (GTX 4060, 16 GB RAM), significantly reducing computation time. A point file with the VP locations and a DSM were used as input data. The AGL was calculated using a 360° azimuth to capture the entire view from each individual VP. For all VPs, uniform input parameters for the AGL analysis were used, namely, an observer offset of 1.6 m, a target offset of 0 m, Earth curvature correction, and a default value of 0.13 for the coefficient of refraction of visible light in air. The resulting 20 AGL rasters were individually visualised on the map (this componential result map is presented in Appendix A). VPs that AGL overlapped with HV-lines or PV-plants were marked as primary, while the remaining were ranked as secondary. Since the AGL of individual VPs overlaps, it is essential to determine the minimum value of the overlapping AGL pixels. This calculation was performed using the cell statistics tool, with the local minima option. The result, referred to as cumulative AGL, was visualised using a colour scale.

2.4. Method for Energy Transition Suitability Map

Since the AGLs of individual VPs overlap, it is crucial to determine the minimum value of overlapping pixels to ensure that the planned energy investment will not impact any VP’s view. These calculations were performed using the cell statistics tool with the local minima option. The result, referred to as cumulative AGL, was visualised using a colour scale.

The preparation of the energy transition suitability map requires prior information on the spatial extent of planned energy facilities, specifically their height. In the RNP case study, a height of 3 m above ground was assumed for PV-farms. Importantly, the assumed value of 3 m is a generalisation derived from a review of publicly available technical drawings of PV installations and two field measurements obtained using a laser rangefinder. In practice, the heights of PV installations vary. Therefore, when applying this framework to other case studies, the PV installation height parameter should be adjusted to reflect local conditions. Furthermore, for HV-lines, the height was set at 16 m, calculated as the average of all regularly distributed points along the analysed HV-lines (as described in Section 2.1). Similarly to PV installations, the framework applies height averaging in this context as well. This parameter should be carefully adjusted for each new case study region, taking into account its specific geographic characteristics. Thus, the energy transition suitability map method is a tool tailored to specific energy infrastructures, here for PV-farms and, in the second case, for HV-lines. At this point, the framework proposes the reclassification of cumulative AGL into three classes: suitable, not-recommended, and undesirable zones. For PV-farms, these correspond to zones up to 3 m, between 3 and 4 m, and above 4 m, respectively. However, the extent of the not-recommended zone is determined individually for each case study. It is designed to account for potential measurement errors due to the vertical accuracy of the DSM, rather than to serve as a specific recommendation or restriction. The zones were visualised on a 2D map using an intuitive colour scale As the study aims to propose a framework rather than to plan new energy infrastructure in practice, no new energy facilities were designated (such a task typically falls within the scope of energy network designers). Instead, existing HV-lines were assessed, with a focus on identifying locations resulting in minimal impact on scenic resources, presenting the results in quantitative terms (HV-line length) as well as viewshed conformation. Similarly, in the case of PV-plants, a quantitative assessment was made regarding existing PV-plants that satisfy the energy transition suitability map criteria, as well as those located in the third zone were also identified; the findings were subsequently summarised. Additional comments were also formulated regarding solar radiation, which constitutes a key information layer for the planning of profitable PV-farm locations. This commentary was included to highlight that the optimal location of PV-plants should take into account not only technical factors such as solar radiation and proximity to the power grid, but also aesthetic considerations, thereby respecting the scenic resources of the visual landscape.

3. Case Study Characteristic

RNP was established by the Regulation of the Council of Ministers dated 10 May 1974, concerning the creation of the RNP. The park initially covered an area of 4800.65 hectares, of which 518 hectares were designated as strict nature reserves: Bukowa Góra (128.45 ha), Czerkies (165.89 ha), Nart (212.06 ha), and Obrocz (11.60 ha). The predominant feature of the park’s flora is its forests, which cover approximately 95% of its surface, including communities of fertile Carpathian beech forests and upland fir forests. The primary purpose of establishing the RNP was to protect the ancient, natural forests that are remnants of the once vast primeval forest that once covered the Roztocze region. From the moment of its establishment, the park has been surrounded by a protective zone. The role of this zone is to safeguard the park from external threats, including the movement of pollutants in water and air, the spread of invasive species, as well as threats to the landscape’s physiognomy. This includes preventing investments that introduce anthropogenic elements that negatively impact the visual resources of the landscape. A distinguishing feature of the landscape of both the Western and Eastern Roztocze mesoregions (Figure 4B) is its gently undulating terrain, interspersed with numerous loess ravines that form a unique, densely connected network, rare on a European scale. The elevation difference within the park and its buffer zone is 157 m, with the highest peak being the “Glimowiżna” hill (357.6 m), located within the park’s buffer zone. The high visual value of the landscape under study is also influenced by the agricultural character of land use in the park’s buffer zone, where long, narrow strips of fields create distinctive and picturesque linear patterns. The park’s attractive landscape, combined with the scenic valley of the Wieprz River and a rich offering of tourist services, makes it highly popular. It is estimated that up to 50,000 people visit the park annually [75]. Simultaneously, extensive sunlit uplands, which were once used for agriculture, are now increasingly becoming sites for renewable energy investments, driven by the ongoing energy transformation. In accordance with the aforementioned legal acts, the park administration reviews local planning decisions concerning public and private investment projects, including those related to energy transformation. To date, 10 negative decisions and 21 positive decisions have been issued regarding PF-farms, with the largest PV-farm, operational since 2023 (27 MW capacity). Furthermore, one project concerning the development of HV-lines within the park’s buffer zone has been negatively reviewed.

RNP staff, working in collaboration with energy investors, must find a balance between preserving the visual value of the harmonious landscape and the need for energy transformation. This framework provides a practical solution to the challenges faced by both sides and helps achieve a sustainable energy transition while maintaining the park’s ecological and visual landscape resources with no impact.

4. Results

4.1. Existing Energy Infrastructure

In accordance with the methodological description (Section 2.1), power lines and photovoltaic (PV) panels were inventoried. As a result, the spatial distribution of the energy infrastructure is presented in Figure 3. The high-voltage (HV) line does not traverse the core area of the RNP; however, a section measuring 45.37 km in length is located within the buffer zone. Based on measurements averaged from 746 measurement points, manually positioned along the HV-line upper line, the average height is 16 m above ground level. The network is further supplemented by a medium-voltage (MV) line with a total length of 187.53 km, of which 5.56 km runs directly through the RNP, while the remaining 181.97 km are situated within the buffer zone. From the perspective of energy grid development, the availability of the HV-line and the transformer station is of key importance. The highest density of HV- and MV-lines occurs in the northeastern part of the study area, where an electrical station is also located. The proximity of this substation strongly influences the preferred siting of PV installations. A total of 143.54 hectares of PV installations have been inventoried, with an additional 49.48 hectares planned, comprising 15 polygons, with areas ranging from 1.11 to 8.13 hectares. All installations are located within the buffer zone of the RNP; however, this does not preclude their potential visual impact on the core area of the park.

The MV-line network is evenly distributed throughout the study area, whereas the HV-line is situated in the northern part of the RNP buffer zone. An additional analysis of compliance with the PN-EN 50341-1:2013-03 standard [71] revealed that there is one farm building and part of one residential building located closer than 7.7 m to the HV-line.

4.2. The Results of Visual Landscape Resources Mapping

The forested character of the case study (forests cover 30,342.5 hectares, which is 66% of the analysed area, but 96.7% is located directly within the park boundary) means that visual landscape resources are predominantly located outside the park boundary, within the park’s buffer zone. Consequently, high VEI values are also associated with the park’s buffer zone (Figure 4A). The highest VEI values were recorded in the northeastern part of the buffer zone (maximum of 6.2%), in Kąty town. According to the Landscape Protection Plan [72] viewpoint no. 19 is located there; however, high visual exposure pertains to an extensive area across the whole Kąty plateau (Figure 5B). The second prominent location exhibiting high VEI values (2.7%) is Żurawnica (northern part of the buffer zone), which corresponds to the site of viewpoint no. 18. The VEI high values zone, extending along the northern border of the buffer zone, needs additional comment—in this case, no viewpoints were proposed due to the area’s limited tourist accessibility (pointing viewpoints in the middle of a farm field was considered pointless). Notably, in the context of energy transition sustainable spatial planning, the between HV-lines/PV-farms and high VEI value zones (the northern part of the buffer zone) suggests potential visual conflicts (although their identification is not the primary focus of this study). All remaining 20 viewpoints correspond to locations designated in the Landscape Protection Plan, as outlined in the methodology of this research. Thanks to VEI maps, field verification (Figure 6), their location, and exposure are confirmed with high accuracy. A detailed list of all resulting viewpoints is provided in

Table 1. The resulted VP list (based on viewpoints proposed in the RNP Landscape Protection Plan, coordinates given in the local EPSG 2180 coordinate system).

VP Number	Native Viewpoint Name (In Polish)	X Coordinate	Y Coordinate	Z Coordinate	VEI (%)
1	Bukowa Góra	779293.478176	309997.356125	301.55	0.61
2	Biała Góra	782840.2345	310004.056808	272.58	0.61
3	Pola Obroczy	784863.636854	311124.616341	275.85	0.69
4	Poręby	784939.447481	310218.698949	334.43	1.67
5	Góra Niedźwiedź	788235.174472	313890.876288	297.49	0.17
6	Góra Księża Choina	789003.46428	314153.201962	299.33	0.18
7	Góra Wieprzec	790177.477029	316257.992831	272.13	0.18
8	Guciów	787509.394271	309803.045279	296.49	0.10
9	Wysoka Góra	781639.392638	304961.217414	320.75	0.65
10	Florianka	782361.061501	306230.674007	272.99	0.00
11	Piaseczna Góra	779620.171333	311377.684349	281.15	0.50
12	Wzgórze Polak	775357.35804	311030.712304	310.21	0.42
13	Góra Młynarka	787898.281443	302016.298036	331.74	1.86
14	Tartaczna Góra	782030.118185	313174.605841	285.4	0.30
15	Punkt nad Tereszpolem	776723.15593	308210.601726	322.12	1.07
16	Senderki	787635.559025	305356.964034	342.73	1.61
17	Felkowa Góra	777761.567214	316147.371747	298.25	0.47
18	Żurawnica	780691.740048	316239.574832	291.13	3.13
19	Kąty	791021.935447	320026.271369	285.86	2.06
20	Wzgórze Adamów	792572.863489	312113.875269	340.55	0.32

Chmielewski and Grabowski [73] argue that high VEI zones are more important than local maxima. They also note that viewpoints must be accessible and easy to find, which explains why not all listed viewpoints have high VEI values. For instance, the maximum VEI in Kąty is 6.2, but the viewpoint was optimally placed where the VEI is 2.06.

. It is important to note that the 20 viewpoint set has not been expanded in order to maintain consistency with the Landscape Protection Plan official documentation. This set constitutes the VP from the perspective of which the further analysis of VLAC will be considered (Results, Section 4.3).

4.3. The Visual Landscape Absorption Capacity Results

4.3.1. VLAC for Single VP

Not all VPs analysed under this case study provide a view on existing HV-lines or PV-farms. Therefore, in this result section, the VLAC will only be discussed in relation to VPs 19, 18, and 17—those for which energy infrastructure facilities are visible. This explains the relationship between the exceedance of VLAC and the visual effect proved with the used of panoramic photo documentation. As previously demonstrated in Section 4.2 (Figure 6B), some PV-farms are visible from VP position no. 19. Specifically, the resulting viewshed indicates the visibility of four PV-farms at a distance of approximately 3 km (Figure 7, purple viewshed area). The VLAC values, for these four PV-farms, range from 0 to 7.45 m, which, with the height of the PV construction 3 m above ground level, results in the VLAC exceeding, thus impacting the panoramic view. Put simply, it results in visual conflict. The remaining six PV-farms are located in an area where the VLAC exceeds 3 m, meaning that the PV constructions are not visible from the observer’s position; this is further corroborated by the purple field of visibility indicated in Figure 7. One more existing PV installation exceeds the VLAC threshold (3 m)—a single PV-plant located in Żurawnica and visible from VP no. 18. The visual conflict is not as significant as those mentioned above, as the PV installation itself is much smaller (1.1 ha) and located 1.6 km away from the VP. However, the VLAC for this site ranges from 0 to 5.9 m (Figure 8C), thereby impacting the spectacular visual resources of this landscape.

To summarise, five of the 15 inventoried PV-farms are located in areas with insufficient VLEI, seven are situated in accordance with the VLAC, and therefore do not cause visual conflicts, while the remaining three (one existing and two planned) are located outside the field of view of any of the 20 VPs considered in this case study.

Moving on to the VLAS results concerning the HV-line, for which the threshold value is 16 m, its exceedance is observed from the VP 19, 18, and 17 perspectives. Of these three, the view from VP 18 is the most impacted and will therefore be reported in detail. A photo documentation taken in VP18 (Figure 8A) shows that HV-line infrastructure, despite using all mitigation techniques (green painting colour), still creates a visual conflict. The section of 1.28 km of 16 m high HV-line is visible since VLAC values for this section are not greater than 3.1 m. In other words, the VLAC is considered to be 13 m too low to visually absorb this linear energy infrastructure. Additionally, the resulting VLAC values are presented as a point layer to better illustrate the phenomenon of visual capacity (Figure 9B), while Figure 9C presents the VLAC raster calculated for the entire analysis area. The resulted maximum values reach 783 m, which aligns with the perspective of a single VP; however, in practice, as multiple VPs are considered, a cumulative VLAC does not result in such extreme values anymore (as outlined in Section 4.3.2). Notably, the landscape absorption zone for objects exceeding 16 m is located at a considerable distance from VP 18 (dark area in Figure 9C), and a substantial part of the HV-line runs through this area. Concluding, photo documentation confirms the high reliability of VLAC calculations; in areas with insufficient VLAC, the energy infrastructure indeed creates a visual impact.

In the case of VP 19, due to the significant distance (approximately 2.5 km), the vertical shape of the HV-line pylons remains discernible (Figure 6C) in the panoramic view, thereby indicating another instance of VLAC exceedance. Additional results for VP 19 are provided using the cumulative VLAC approach in the subsequent section. Furthermore, the VLAC results indicate an exceedance of the threshold value, also in the case of VP 17. However, the photographic documentation taken in 2024 does not fully confirm the calculations made using source data acquired in 2014. Thus, the influence of the timeliness of spatial data on the VLAS is presented in the Discussion section, with VP 17 used as a representative example.

Nonetheless, the above-presented results pertain to the perspective of a single VP, whereas the framework for a sustainable energy transition consistently advocates the use of multiple VPs. Therefore, the subsequent results focus on cumulative VLAC, which is considered more reliable. The component results concerning the AGL of individual VPs are presented in graphical form in Appendix A (Figure A1).

4.3.2. Cumulative VLAC Results

Cumulative VLAC encompasses all VPs designated within the framework, and the result of such calculations for RNP is presented in Figure 10A, where the AGL values range from 0 to 112.1 m. Briefly, three distinct zones are clearly identifiable, which can accommodate new energy infrastructure without conflicting with the scenic resources. To the east of VP 19, a triangle-shaped area of agriculture results in cVLAC values reaching up to 93 m. Additionally, the northeastern part of the buffer zone is also identified as a suitable area for the implementation of energy transformation investments up to a height of 40 m. Similar sites, although much smaller in terms of area, are also located in the western part (west of VP 17) and, with a considerably reduced absorption capacity, in the southwestern part as well. The highest cVLAC, however, occurs in the Bondyrz–Glamówka locality in the southeastern part of the study area, where the maximum value of 112.1 m was recorded. The results are further illustrated in Figure 10B, where only areas with cVLAC values exceeding 16 m are shown. In conclusion, with regard to the RNP, areas with high scenic potential are not present within the RNP itself. However, the eastern part of the buffer zone is more suitable for the implementation of energy transformation tasks than the western part. Another favourable factor is the distance from the park ridge.

4.4. Resulted Suitability Map

The final outcome of the three-step framework is a set of maps, prepared here as an example for sustainable energy transformation, specifically for PV-farms with 3 m height (Figure 11B), and for the development of a network of HV-lines at a height of 16 m. With regard to the latter map, it is evident that moving (or planning) the HV-line by approximately 1 km to the north, towards the outer boundary of the buffer zone, would certainly eliminate visual conflicts (ref. to Figure 8). A quantitative summary of the HV-line sections within particular suitability zones (

Table 2. A quantitative summary of energy infrastructure placement within the suitability zones of the RNP case study.

PV-Farm Suitability Map				HV-Line Suitability Map
Zone name	AGL	Zone area (ha)	PVs	Zone name	AGL	Zone area (ha)	Length (km)
Visual impact risk	<3 m	1628.08	6	Visual impact risk	<16 m	7870.96	22.38
Conditional	3–4 m	627.39	1	Conditional	16–17 m	452.55	0.95
Allowed	>4 m	12,433.31	8	Allowed	>17 m	6365.27	22.04

) facilitates a moderate assessment. Excluding sections in forested areas and focusing on the open landscape only, half of the HV-line length (22.38 km) is in conflict with the visual resources of the RNP landscape. Notably, the proposed framework’s objective is to identify suitable zones—the “allowed” zones for the HV-line covering 6365.27 ha, which constitutes 13.86% of the entire analysed area. Regarding the development of PV-farms, 6 out of 15 farms diminish the visual resources of this landscape. Conversely, 27.08% of the area is considered suitable for energy transition, taking into account visual landscape resources. The key result, however, lies in the spatial distribution of the blue “allowed” cVLAC zones, which indicate areas for sustainable energy transition. The objective of this work is not to discuss these zones in detail within the context of the RNP, but to present the methodology for their computation. In conclusion, the northernmost part of the buffer zone permits energy transition; however, the interpretation of the results must consider certain technical limitations of the method, which should be taken into account when replicating the methodology in other case studies (see next section).

5. Conclusions and Discussion

The proposed framework provides a geoinformation layer—an energy infrastructure development suitability map that streamlines spatial planning by incorporating visual landscape protection aspects. Compared to methods focused primarily on network efficiency and stability [77,78,79], the suitability map enhances preparedness for the obligatory visual impact assessment procedure. The framework advances beyond previous viewshed-based approaches [80] by incorporating AGL measurements, enabling the prediction of visibility outcomes based on planned infrastructure height, thereby supporting energy planning that minimises visual landscape impact. However, due to the previously mentioned limited recognition of the practical aspects of the VLAC theory [45,46] and the pioneering nature of the proposed framework, comparing the results with those of other authors appears to be an unattainable task.

In a broader context, environmental concerns drive the transformation of energy systems; however, their implementation, while aiming at the sustainable management of all resources, remains a complex challenge. Existing energy transition models have low concerns about land use, landscapes, and biodiversity [33]. A contribution is made by this research to the body of knowledge on scenic resource protection and integrating energy transition into protected landscapes, with RNP as a case study. Although the proposed solution does not eliminate the visual impact of energy infrastructure, it increases the likelihood of grid expansion proceeding without conflict with visual landscape resources.

Since 2023, Poland has been undergoing a reform of its spatial planning system, which includes the implementation of landscape audits for voivodeships, alongside local and general spatial development plans. As part of landscape audits, viewpoints and priority landscape areas have been identified for each voivodeship. The results of the audits are available as web map services, offering a valuable resource for integrating all viewpoints (thus VPs). Technically, this would enable the calculation of VLAC across the country, particularly for priority landscapes. The creation of such an information layer would extend beyond the scope of sustainable energy transformation, serving as a key information layer for supporting the national spatial planning system with reference to VLAC. The VLAC information layer can be shared as a generalised web map service to improve accessibility—particularly for use in public consultations or by investors during the early stages of energy infrastructure planning. Legislative requirements call for the preparation of technical guidelines, with this publication representing an initial step towards achieving this goal. Taking this possibility further, Poland, as an EU member, could propose a framework at the level of international recommendations. However, expanding the analysis to a broader scale introduces additional technical challenges, which are beyond the scope of this discussion.

The presented suitability map, similarly to other spatial information layers (e.g., [81,82]), classifies areas as more or less suitable for power grid development herein in the perspective of landscape resources. However, the resulting layer integration into automated transmission network planning processes [83,84] requires further research and parameterisation. The resulting suitability map was prepared in the variant for HV-lines and PV-farms. Theoretically, implementing a PV version is somewhat more feasible, as their planning is less spatially extensive and, consequently, less automated in comparison to HV-lines. In practise, the PV-farm planning can be addressed using multi-criteria analysis methods, using GIS software [85] or handy WebGIS [86] with the suitability map included as one of the input layers.

What requires the most careful consideration are the limitations of the method, particularly those related to the source data used, the inherent characteristics of the viewshed algorithm, and the degree of subjectivity involved in viewpoint mapping. As spatial data becomes outdated from the moment they are acquired, this inherent limitation intensifies over time. The case study is similarly affected, having been based on a 3D point cloud collected in 2014, now 11 years old. An intensive process of secondary plant succession [87] was observed during viewshed results field verification, which occurs in the VLAC, being underestimated (Figure 12). It is intuitively understood that the older the data, the greater the potential biases that may affect the results. However, the scientific literature does not provide specific information on how selected land cover types influence the accuracy of viewshed analysis over time. Sobala et al. [88] demonstrated that land use changes over time affect viewshed results, particularly the dynamics of changes in non-forest areas, where viewsheds exhibit considerable diversity. However, assessing the loss of accuracy in viewshed results due to varying degrees of data obsolescence remains an unresolved issue. This is especially true across different land use and landform configurations. The issue is further complicated by the fact that viewshed analysis, even when performed with current data, has inherent limitations. Factors such as vertical structure and the number of view curtains contribute to these inaccuracies. Better results tend to be obtained when examining landscapes devoid of vegetation, as the raster used for viewshed calculations simplifies the shape of vegetation [89], introducing errors into the results. This remains a significant challenge for viewshed algorithms.

The impact of source data resolution on viewshed accuracy is widely discussed in the scientific literature, both for 2D data sources (such as DSM or DTM) [90,91] and 3D data, including voxel-based models [92]. These limitations also apply to the VLAC method. While the framework proposed in this manuscript was tested in Central Europe, its application in other geographical regions requires consideration of local factors—such as weather conditions and temporal variations that may affect visibility analysis [93].

Incorporating the aspect of data temporal accuracy further complicates the problem. It presents an intriguing research challenge. In VIA analysis practice, it is recommended to use the most current data available. This is often achieved through a fusion of aerial data and terrestrial scanning data, or by combining these with data acquired through photogrammetric methods [94].

Moreover, the framework defines infrastructure parameters only by planned height, omitting factors such as colour, background, or, in the case of HV-lines, their openwork structure. These aspects, though visually important, are considered secondary and should be assessed individually once detailed technical specifications are available. While the cVLAC framework does not account for such custom features, it does not dismiss their impact on the final visual impact.

Another limitation relates to VLAC mapping in forested areas. In this framework, such areas are masked; however, this is not the only possible approach. In the case study, the high-voltage (HV) line also passes through forested zones but remains below the tree canopy. As an alternative to masking, a canopy height model [95] can be incorporated, allowing the assumption that only infrastructure exceeding tree height contributes to visual impact.

The final limitation also relates to vegetation and arises when the observer (the VP) is located near a dense clump of vegetation. Even a single tree or bush can form an opaque visual curtain, behind which an area of artificially elevated VLAC is generated. This effect is evident in Figure 9A, where VLAC values increase with distance behind VP 18. The 2D viewshed has been criticised in the literature for its limited accuracy and tendency to overestimate visibility [96]. Consequently, a robust VLAC assessment should incorporate multiple VPs rather than relying on a single one, as minimum statistics resolve the discussed issue.

The selected case study (RNP) highlights the consequences of a flawed spatial policy. HV-line investments are classified as strategic national projects, triggering a special act that prioritises national energy security. As a result, landscape protection is marginalised. Despite this, the case study shows that the proposed framework can effectively address real challenges linked to energy development in protected areas. Therefore, implementing the framework, even in an advisory role, can help balance energy needs with landscape conservation, also in other geographical regions.

Concluding. The transition to renewable energy and new power grids can affect the visual resources of protected landscapes, so implementation of VLAC into spatial planning is needed to balance energy development with nature conservation.

A new mapping framework helps identify the best places to build energy infrastructure with the least visual impact, supporting smarter and more sustainable decisions.

While the method has some limits, it provides a useful tool that can be adapted and improved to help reduce conflicts between energy projects and scenic landscapes.

Making this visual landscape information widely accessible, for example, through web maps, could help involve the public and investors early on, leading to better energy planning that respects the environment.