Exploring the Possibilities of Implementing the ALS-Based 3-30-300 Concept for Urban Green Space Management in Small Municipalities

Abstract

This study examines the applicability of the 3-30-300 rule in five medium-sized Polish municipalities. The rule states that residents should be able to see at least three trees from their homes, neighborhoods should have at least 30% tree canopy coverage, and public green spaces should be within 300 m. The method proposed in this study shows that the tree visibility component of the 3-30-300 concept is the most fluctuating index, and it strongly depends on the settings of the algorithm parameter, as well as on the placement of artificially generated observers. This demonstrates the complexity of the issue and the need to further specify the nuances of the 3-30-300 rule. The work shows that all variables of the 3-30-300 rule can be calculated based on publicly available data, such as point clouds, which are increasingly being made available for free for research and implementation purposes. The study concludes that the proposed solution is effective in assessing the availability of green spaces and emphasizes the need for qualitative improvements in the management of urban green spaces. While the 3-30-300 rule can serve as the foundation for future urban planning, complementary strategies are needed to ensure long-term sustainability and better access to green spaces.

1. Introduction

Numerous actions are being undertaken to adapt cities to climate change with the aim of improving the living conditions of their inhabitants. Many of them have also been implemented as part of European Union law [1]. The World Health Organization (WHO) has developed several indicators regarding this topic, e.g., health indicators related to air quality and extreme weather events, indicators of access to green spaces, and indicators of health adaptation [1]. One of the tools for adapting cities to the effects of climate change and improving the living conditions of a growing number of inhabitants is the appropriate design of urban green spaces—from local to city-wide spatial levels [2].

Greenery in urban areas is an inseparable element of the urban fabric and composition [3]. The concept of urban greenery, in the broadest sense, encompasses all biologically active areas in a city. It includes both managed green spaces (parks, squares, green areas, etc.) and unmanaged greenery (forests, isolated green spaces, woodlots, etc.) [4]. In a narrower sense, urban greenery can only be understood as the population of trees in the above-mentioned areas. Regardless of the definition, the role of urban greenery in shaping urban space and improving the quality of life of residents is significant [5].

According to the concept of ecosystem services, urban greenery can provide services in four categories: provisioning, regulating, cultural, and supporting [6,7]. In urban areas, timber-harvesting activities related to local timber resources are of marginal interest [8]. As surveys show, city residents value the regulating and cultural services provided by green areas and trees in cities much more [9]. Urban greenery reduces the risk of the urban heat island effect by providing shading, evapotranspiration, and a reduction in thermal amplitude [10,11]. Trees also reduce surface runoff and, thus, the occurrence of local flooding due to heavy rainfall [12]. In addition, urban greenery, especially trees, effectively reduces air pollution, including particulate matter [13]. Reducing pollution decreases mortality due to respiratory diseases [14]. Another regulatory service provided by greenery is noise reduction [15]. It has been known for many years that urban greenery has a positive effect on the health and well-being of city residents [16]. Numerous studies at the intersection of various disciplines and scientific fields have shown that contact with nature and with trees improves one’s psychophysical state [17], reduces stress [18], and reduces the risk of depression [19]. This positive effect has been demonstrated in recent years during the COVID-19 pandemic [20]. In addition to direct contact with nature, the mere possibility of seeing green spaces, e.g., through a window, also has a positive effect on the lives of residents [21]. Urban greenery also provides important cultural services related to aesthetic values, complements the urban composition, and inspires and forms the identity of the place [22]. The cultural service of recreation in green spaces is extremely valuable for city dwellers [23,24], especially as access to green spaces in highly urbanized areas varies greatly, and there are inequalities in access for different social groups [25]. This is due to investment pressure, the need for new residential areas, low-quality spatial planning, and a lack of urban standards and land values. These factors often cause green areas to change their purpose to other types of land use. Green areas are becoming a deficit asset, the demand for which significantly exceeds supply.

All of the ecosystem services mentioned above directly or indirectly affect the well-being of citizens in place of their residence and the standard of living of society [26]. It should be mentioned that urban greenery is an important element of a nature-based solution (NbS), as it can provide a wide range of ecosystem services if properly placed and planned. Therefore, the appropriate identification of green resources and making decisions that support the modification of resources to better adapt to climate change should be one of the challenges for city managers. NbS-based activities provide benefits for residents and nature, and one of the elements that NbSs could improve is fair and adequate access to urban green spaces for city residents [27,28]. This statement was reflected in the 2030 Agenda—the world development strategy adopted by the United Nations in 2015 [29]. One of the goals was defined as ensuring easy and universal access to safe green areas, taking into account various groups that use the space, including women, children, the elderly, and the disabled. The Polish Ministry of Climate and Environment also emphasized the availability of green spaces as a criterion that can be used by local authorities for various strategic decisions in the handbook published in January 2022, which contains natural and climatic indicators for sustainable development [30].

Existing studies on the thresholds defining the sufficient amount of greenery per capita, the availability of managed green areas, and their size differ both in the literature and in strategic and planning documents [31]. The guidelines of the World Health Organization (WHO) indicate that areas located at a distance of more than 300 m from forests and managed green areas should be treated as “excluded” [32,33]. Furthermore, this distance is counted only from forests and green areas whose area is larger than 1 ha. This results from numerous studies on the minimum size of a green area that has appropriate characteristics to provide certain ecosystem services [34,35]. Kabisch et al. [36] determined the demand and supply of urban green spaces by calculating the number of people within a radius of 300 and 500 m around green areas of over 2 ha. Cardinali et al. [37] indicated the distance of up to 100 m as one with a positive impact on health and physical activity. Greenery in the vicinity of 500–1100 m from the place of residence also has a positive effect on health. Different guidelines were included by Polish legislation on urban standards when amending the Act on Spatial Planning and Development [38]. According to the provisions, the distance from public green areas should be less than 1500 m in the case of public greenery with a total area of no less than 3 ha, and it should be within 3000 m for areas over 20 ha. In addition to quantitative issues, researchers point to the need to create high-quality green areas [39].

Sustainable urban management that involves actions related to climate change adaptations can make use of synthetic indices. In particular, one such indicator, the “3-30-300 rule” presented by Konijnendijk et al. [40], has gained prominence in recent years. This guideline aims to ensure equitable access to green spaces by establishing the following thresholds: (1) at least three well-established trees within sight of every home, school, and workplace, (2) at least 30% tree cover in every neighborhood, and (3) no more than a 300 m distance from any residence to the nearest public green space. What is notable is that the author of the concept did not specify a calculation method for the above components, which means that different research teams are currently proposing their own solutions. To assess access to green spaces using the “3-30-300” rule, some authors [41] have conducted questionnaire surveys. Others have conducted visual and spatial analyses of images provided by respondents [42], online databases and services [43], and remote sensing data [44].

The individual components of the 3-30-300 index have also been realized differently. Nieuwenhuijsen et al. [45] assessed the greenness of the window view with a question about whether participants could see trees from their homes. They combined this information with data on the number of trees within 15 m of a residence using a central geodatabase. The locations of buildings from OpenStreetMap and the locations of trees from urban databases were used to determine the visibility of trees in 20 m [46] or 30 m buffers [47]. Ling [43] determined the total number of trees of a “decent” size that could be seen from the front window of each residential building using Google Street View, Google satellite images, or 360-degree views provided by an estate agent. Zhang et al. [42] calculated window views using the Green View Index of photos taken by respondents from their most viewed windows with the best viewing angle. Daland [44] proposed viewshed tool analysis, an algorithm that calculates all areas visible from a given point based on a raster elevation model.

The second component of the 3-30-300 rule, i.e., canopy cover, was also determined in various manners. Nieuwenhuijsen et al. [45] used the Normalized Difference Vegetation Index (NDVI) in 500 m buffers to estimate canopy cover in the direct vicinity. Zhang et al. [42] applied Sentinel-2 data with a spatial resolution of 10 m. Ling [43] used a similar method that was implemented in i-Tree Canopy. This tool randomly places points on Google Earth images, allowing the user to decide on their class. Browning et al. [48] recommended Digital Surface Models (DSMs) from LiDAR (Light Detection and Ranging) in conjunction with high-resolution imagery and high-accuracy land cover maps for the measurement of green space locations and the determination of canopy cover. Daland [44] applied zonal statistics to calculate the area occupied by trees based on the canopy height model.

The proximity of green areas has been a subject of consideration for many years. Nieuwenhuijsen et al. [45] calculated the linear distance to the nearest green area. Ling [43] determined the walking distance from the place of residence to the nearest boundary or path of a park or green space using Google Maps. The park or green space had to be public, free of charge, and at least 1 hectare in size. Daland [44] performed a cost distance analysis by calculating the shortest weighted distance from each cell of a cost surface raster to the nearest source location. Browning et al. [48] proposed several available types of Geographic Information System (GIS) data as a source of information to identify the locations of urban green spaces in order to determine their spatial relationships. Distances can be operationalized based on “as the crow flies” (i.e., Euclidean) or with more sophisticated analysis, which takes into account transportation networks (e.g., streets and junctions) and access points (e.g., centroids or entry points for the urban green space). Another approach was used by Zhang et al. [42], who calculated the average NDVI value within a 300 m buffer around check-in locations using the Google Earth Engine platform.

Taking into account the range of various approaches to realizing the 3-30-300 concept, it seems that remote sensing (RS) can be an appropriate tool for the determination of urban-greenery-related indices at the level of both individual trees and their features, as well as groups of trees [49]. Segmentation of individual trees was performed by Fekete and Cserep [50], who successfully determined single trees’ species, location, health status, crown parameters, and other features that were relevant from the point of view of managing greenery resources. Tree detection could also be performed for smaller areas using terrestrial laser scanning data [51], while mobile laser scanning could be used for linear objects [52]. Entire cities are mapped using airborne laser scanning (ALS) [53] or aerial image content analysis [54]. Depending on the complexity of the data and methods, the tree detection accuracy ranged from 62 to 100% [55,56].

Until now, the above-mentioned projects have mainly been undertaken for larger cities [46,57]. Nevertheless, the share of small and medium-sized municipalities in the structure of urban areas in the European Union (EU) is considerable. In 2023, there were almost 8000 towns in the EU, home to 95 million people. Two-thirds of these were densely populated, i.e., with 1500 inhabitants/km² and with populations between 5000 and 50,000 inhabitants. The population of towns was distributed almost evenly across the three size categories; just over a third lived in small towns (5000–10,000 inhabitants) and medium-sized towns (10,000–25,000), while just under a third lived in bigger towns (>25,000) [58]. We can assume that large urban agglomerations have both policies and resources for urban greenery management. However, this cannot always be the case in every town, the main reason for this being the limited budget and human resources available. On that account, there the reason for investigating the possibilities of transferring the 3-30-300 concept to the needs of smaller residential areas emerges.

In this paper, we present an original method for identifying urban green space resources in cities of different sizes based on the 3-30-300 concept. Our approach utilized free and open data such as OpenStreetMap and the increasingly collected public and available non-commercial point clouds [59]. The major focus of our work was to test a new approach to determining the ‘3’ component of tree visibility from buildings. We proposed a method based on so-called observers, i.e., artificially generated positions of tree observations. Since we also have information on the number of observers, we were tempted to identify so-called ‘high-priority trees’ that could be of greater value to a larger number of residents.

The proposed approach offers the opportunity to replicate the methodology in analyses for cities that do not have the financial and human resources to collect LiDAR data. In contrast to most work on this topic, we mainly use only one type of remote sensing data as a source. The presented solution makes it possible to identify places where there is a deficit of green spaces, and, therefore, appropriate management measures are needed. Moreover, the objective nature of the proposed methods allows comparisons between cities/regions thanks to the application of ALS data.

2. Materials and Methods

2.1. General Concept

The main idea was to develop a solution that would make it possible to efficiently determine the availability of urban green spaces according to the guidelines of the 3-30-300 concept. The solution was intended to be universally applicable but dedicated foremost to smaller municipalities, where detailed inventory data on urban vegetation are scarce. Our proposal is based on the use of ALS point clouds and the OpenStreetMap database (Figure 1). The application of GIS and RS tools and data should provide standardized and objective results so that municipalities can be compared according to the guidelines of the 3-30-300 principle. In the following subsections, the implementation of the individual components of the 3-30-300 concept is presented in detail.

2.2. Study Area

The analysis focused on five towns: Świdnik, Wyszków, Czempiń, Jasień, and Mrocza (Figure 2,

Table 1. Selected characteristics of the towns presented in this study.

Characteristics	Czempiń	Jasień	Mrocza	Świdnik	Wyszków
Area [km2]	5.0	3.6	4.8	20.4	20.9
Population	4137	5127	4044	36,806	26,042
Population density [people/km2]	824	1420	846	1803	1252
Green areas [%]	0.8	0.6	1.07	8.1	2.1
Forest cover [%]	1.6	2.2	6.1	9.6	3.4

) which are representative of the demographic and urban structure in Poland. Almost 38 million people lived in Poland at the end of 2023. Around 60% of them were in the country’s 1013 towns. The urban structure is dominated by towns with less than 5000 inhabitants (44%), while this accounts for only 3% of people living in towns. The total number of small towns (5000–10,000 inhabitants) and medium-sized towns (10,000–25,000) is similar—around 400—but almost half of all urban inhabitants of Polish towns are agglomerated in these groups. Every third inhabitant lives in a city with a population of more than 50,000 [60].

The selection of these cities resulted from the fact that approximately 80% of Polish cities have fewer than 50,000 inhabitants. As mentioned before, similar towns usually do not have sufficient tools for greenery monitoring, which makes them proper study objects. The general characteristics of the study areas are presented in Table 1 and Figure 2.

2.3. Data Source

Lidar data from the Head Office of Geodesy and Cartography [61] were primarily used for the analyses in this project. Airborne scanning missions were carried out in different time periods: in Czempiń and Wyszków in 2011, in Świdnik in 2011–2013, in Mrocza in 2015, and in Jasień in 2021. The point cloud density of about 4 points/m² was consistent across all of the towns. The points’ average elevation error was 0.15 m. The point clouds had already been classified by the data provider according to the LAS 1.2 standard [62]. In this study, ALS data were used for treetop detection, single-crown segmentation, and DSM generation, which served as a basis for analyzing the components of the 3-30-300 principle. Other remote sensing data were derived from the OpenStreetMap database [63], namely, the spatial layers of residential buildings and green areas. The green areas were additionally verified by city officials.

2.4. Point Cloud Processing

The point clouds were processed in the R environment version 4.2.3 [64] using the lidR library [65,66]. In order to detect the treetops, the height of the point clouds was normalized in the first step. This process removed the influence of the terrain on the measurements above the ground. This made it possible to extract and compare the heights of the aboveground vegetation. For this purpose, a k-nearest neighbor (KNN) approach with an inverse-distance weighting (IDW) algorithm for spatial interpolation was used.

The point clouds were further processed with a local maximum filter. This was a lidR algorithm for individual tree detection inspired by the work of Popescu and Wynne [67]. The threshold for trees was set to a height of at least 5 m. The resulting layer of points representing treetops was then manually evaluated. To improve the accuracy, the positions of treetops along the roads were manually adjusted.

In addition, a rasterized Digital Surface Model (DSM) with a resolution of 0.5 m was created from the point cloud. This layer was created based on the Pit-free algorithm developed by Khosravipour et al. [68], which is based on the calculation of a series of classical triangulations at different heights.

2.4.1. Tree Visibility

In this study, the term “visibility” should be understood as a direct linear and unobstructed line of sight between the observer and the treetop. The potential visibility of the trees was simulated from each side of each building. The visibility of a tree was determined in two phases.

In phase “A”, superfluous vertices, i.e., those whose removal did not change the overall orthogonal shape of the building, were deleted from the polygons. The polygons were then converted into lines representing the contours of the building. These lines were then cut into individual sections, and their lengths were calculated. Eventually, only lines whose length was greater than half the average length of all sections of a particular building were included in further analysis. In this way, only the main walls forming the generic body of the building were identified. This operation was carried out to reduce the noise caused by tiny sections of a building facade where the presence of a potential window seemed unlikely.

In the next step, an extruded point was created at the center of each section at a distance of 0.5 m perpendicular to the elevation line and at a height of 1.7 m, which corresponds to the height of an average person (white dots in Figure 3A). Henceforth, throughout the article, such points are referred to as “observers”. An additional advantage of this approach is that it is easy to identify so-called “blind” observers (red dots in Figure 3A), i.e., places where a building has no windows because it shares a wall with another building.

Phase “B” (Figure 3B) consisted of drawing a line of sight between the treetops and each observer. If the line crossed the building (gray lines in Figure 3B), this meant that the observer could not see the tree. For the line where the view was not obstructed by the building, a terrain profile was extracted from the DSM raster. At this stage, the visibility condition was met if the treetop was higher than other objects in the line of sight. An observer passed the visibility constraint if the condition was met for three or more trees. Per-building results were distinguished depending on the assumption of how many observers in a building needed to see three trees in order for the entire building to be considered as fulfilling the “3” component. Due to the lack of a precise definition of this component, we proposed three variants: (I) at least 1 observer sees at least three trees, (II) at least half of the observers (each) see at least three trees, and (III) more than 90% of the observers fulfill the criterion of the first component.

Additionally, a backward analysis was conducted to identify high-priority trees. By reversing the visibility condition, we recorded how many observers could see each tree from how many buildings. With this slight alteration, we were able to identify the priority trees to which the community should pay special attention.

2.4.2. Canopy Coverage

The sf library for the R programming language was used to determine the “30” component [69,70]. The tree canopy cover was generated from the point cloud by transforming points of class 5 according to the LAS specification [62] into a raster with a 0.5 m resolution. Next, the proportion of pixels representing tree crowns in the total area of every building’s neighborhood was calculated. As it is subjective to clearly define the appropriate neighborhood radius, variants with five distances were chosen: 30, 60, 90, 120, and 150 m from the residential buildings. The final result was the percentage of the area occupied by tree canopies in these buffer zones.

2.4.3. Distance from a Public Green Space

As in the previous analysis, R version 4.2.3 [64] was used to determine the “300” component. First, the building and green space layers were downloaded from the OpenStreetMap database using the osmdata library [71]. Then, the polygons were processed using the sf library [69,70]. Only green areas larger than one hectare were considered. The choice of green spaces with a larger area was related to the guidelines developed by the WHO [32]. The datasets of potential green space polygons were then cross-checked with city councils to ensure that they were publicly accessible to all residents. The green spaces were classified into three categories: small (1–5 ha), medium (5–10 ha), and large (over 10 ha). Then, the Euclidean distance was measured from the outline of every building to the boundaries of the green space with a straight planar line. This gave us a slightly better insight into the distance that residents have to travel to the different types of green spaces, although the interpretation remained relatively simple.

3. Results

The results themselves were not the primary aim of this study, but rather the presentation of the possibilities that arise when RS data are included in the analytical process. The visualization and final analysis of the results were conducted using the QGIS 3.34 software [72]. The possibilities of data interpretation are presented together with the results in the following subsections on the individual components of the 3-30-30 concept. In addition, a synthetic measure was proposed in different variants, as the lack of the rigid constraints of the 3-30-300 method makes the results susceptible to the threshold criteria.

3.1. The “3” Component—Tree Visibility

Table 2. Percentage of buildings meeting the criterion of visibility of at least three trees depending on the proportion of observers.

Town	Tree Visibility Component for the Buildings * [%]
	I	II	III
Czempiń	88.0	55.4	26.8
Jasień	88.6	53.9	26.2
Mrocza	89.7	55.4	28.2
Świdnik	80.7	49.9	23.6
Wyszków	80.5	51.1	25.8

* Variants: (I) at least one observer can see three or more trees; (II) at least half of the observers (each) can see three or more trees; (III) at least 90% of the observers (each) see at least three trees.

shows the share of buildings from which at least three trees are visible. In all towns, more than 80% of the buildings met this criterion in variant I. The higher the percentage of observers inside every building who were able to see at least three trees, the lower the percentage of buildings that fulfilled this criterion. In the strictest variant, where it was assumed that >90% of observers see the corresponding number of trees, the percentage of buildings that met this criterion fell to around 25%.

Interesting results were provided by our modification of the visibility component that determined the high-priority trees. Recording information about how many observers can see a particular tree provides an interesting tool for urban decision makers. In this way, we can point out trees and areas (Figure 4) that need extra attention due to the large number of people who see these trees from their windows every day. For example, the proportion of trees seen by 10 or more observers from their windows was less than 1% in four out of five towns. The highest value was in Czempiń, where it reached 4.5%. The maximum values (i.e., the total number of observers) of high-priority trees ranged from 15 in Jasień to 26 in Świdnik.

3.2. The “30” Component—Tree Canopy Cover

The tree canopy cover in the vicinity of the building was analyzed in five different distance zones (

Table 3. Percentage of buildings that meet the tree canopy cover criterion with the different neighborhood size options.

Town	Share of Buildings Meeting the Criterion of Tree Canopy Cover [%]
	30 m	60 m	90 m	120 m	150 m
Czempiń	2.5	2.4	2.7	2.7	2.9
Jasień	4.0	10.7	12.6	13.8	16.0
Mrocza	0.1	0.1	0.4	0.1	0.0
Świdnik	1.1	1.1	1.2	1.3	1.4
Wyszków	7.9	8.6	8.4	8.4	8.5

). Only in Jasień did the enlargement of the buffer around the building lead to a significant increase in the percentage of buildings meeting this criterion. The greatest gain (6.7 percentage points) was observed when transitioning from 30 to 60 m. In other towns, the change in the size of the neighborhood area had hardly any effect on the criterion. In Wyszków, the proportion of buildings that met the criterion was highest (7.9%) when the neighborhood had a 30 m buffer zone. In the other neighborhood variants, the percentage of buildings that met the 30% canopy cover criterion was highest in Jasień. In Mrocza, Świdnik, and Czempiń, the percentage of buildings in the different variants did not exceed 0.4, 1.4, and 2.9%, respectively.

3.3. The “300” Component—Access to Green Areas

The analysis showed that 21.8% of the residential buildings in Wyszków and even 77.3% of the buildings in Jasień were located within a radius of 300 m from green areas of more than one hectare (

Table 4. Percentage of buildings meeting the distance criterion in different variants.

Town	Share of Buildings Meeting the Criterion of Distance [%]
	Size of Green Area			Total Number of Green Areas
	≥1 ha	≥5 ha	≥10 ha	≥1 ha	≥5 ha	≥10 ha
Czempiń	38.3	29.6	5.0	4	1	3
Jasień	77.3	53.6	14.9	7	3	2
Mrocza	46.7	32.6	0.0	5	2	2
Świdnik	61.9	32.3	31.2	18	2	5
Wyszków	21.8	12.9	9.5	10	9	1

). If we took into account the distance from larger green areas, the number of buildings that fulfilled this criterion decreased significantly. When considering green spaces of more than 10 hectares, only Świdnik fulfilled this criteria. A peculiar situation concerned Mrocza, where, despite having two large green areas, there were no residential buildings within 300 m.

3.4. Results of the 3-30-300 Rule

The two variants of the results for the analyzed towns shown in

Table 5. Percentage of residential buildings in relation to the number of criteria met.

Town	Variant I *				Variant II **
	0	1	2	3	0	1	2	3
Czempiń	7.6	57.7	32.6	2.1	42.24	52.57	4.84	0.35
Jasień	2.7	26.9	56.2	14.2	36.00	47.84	11.52	4.64
Mrocza	4.8	54.0	41.2	0.0	44.60	55.40	0.00	0.00
Świdnik	8.5	40.2	50.0	1.3	35.51	47.53	16.03	0.93
Wyszków	15.8	67.4	16.7	4.9	45.44	44.32	5.98	4.27

* At least one observer in the building sees three trees, the canopy cover is indicated at a distance of 150 m from the building, and the distance to green areas is at least 1 ha; ** at least 50% of the observers in the building see three trees, the canopy cover is indicated at a distance of 150 m from the building, and the distance to green areas is at least 10 ha.

differ in the criterion for the number of observers and the area of green spaces. The neighborhood criterion for the calculation of the tree canopy cover was left unchanged, as this variable did not depend on the distance assumptions for the selected towns. Regardless of the variant, the percentage of buildings meeting all 3-30-300 criteria was not high in the towns analyzed. In variant I, 14.2% of the buildings in Jasień met all three components, while in Mrocza, no residential buildings met all criteria. The vast majority of buildings (around 80–90%) met one or two criteria. The highest percentage of buildings that did not fulfill any of the criteria was recorded in Wyszków (15.8%).

The tightening of the criteria in variant II and the consideration of a higher number of observers in the building (at least 50%) who had to see at least three trees, as well as the distance from green areas of at least 10 ha, lead to a decrease in the percentage of buildings that fulfilled the criteria. In Jasień, the percentage of buildings meeting the three criteria fell by almost 10 percentage points, while in Wyszków, it only decreased by around 0.6 percentage points. To better illustrate the results, the distribution of the individual components of the 3-30-300 rule in the two variants discussed is shown in Figure 5.

The criterion evaluation is also shown in aggregated form on the grid level (Figure 6 and Figure 7). A square in the grid provides information on the number of criteria for the majority of buildings that fall within that square. In all towns, 6.1% of the squares in variant I were counted as meeting all criteria, 36.7% were counted as fulfilling two criteria, 51.9% were counted as fulfilling one criterion, and 5.7% were counted as not fulfilling any criteria. In variant II, the corresponding percentages were 3.4%, 10.4%, 52.8%, and 33.5%.

4. Discussion

The 3-30-300 concept is a relatively young idea introduced by Konijnendijk in 2022 [40]. However, it has been scientifically justified, especially in relation to the “30-300” components [48]. In the case of the “3” component, many papers point to the benefits of visual access to green spaces [73]. One should bear in mind that the 3-30-300 idea is a general concept that does not include specific definitions regarding the qualitative features of green patches, such as their biodiversity, health conditions, or landscape/sight values. From one side, this flexibility enables the adaptation of the concept to the very specific characteristics of a town/city. As this is a relatively new concept, work is still underway to develop a methodology for calculating the indicators that it contains. Most research in this area, therefore, still consists of specific case studies with relatively small samples that differ in the methodology used. This can be somewhat of an obstacle both for scientific analysis and for common policies in terms of comparisons between cities or regions.

The aim of this article was, therefore, to propose an objective path for the future implementation of the 3-30-300 concept, which can be applied relatively easily with the available non-commercial ALS data [59] or with drone solutions that have recently become more common and competitive in many areas [74]. The issue has gained importance as some documents that recommend the incorporation of the 3-30-300 concept into urban spatial planning have recently been published [75]. Furthermore, there is a significant share of small and medium-sized towns among the urban areas of the EU. Approximately 95 million people live in 8,000 towns, constituting around 21% of the entire EU population. Two-thirds of these towns are densely populated, i.e., with 1500 inhabitants/km², with the total number of inhabitants ranging between 5000 and 50,000 [58]. We perceive the aforementioned facts as a foundation for justifying one of the subjects of this study, i.e., an exploration of the current state of adaptation to the 3-30-300 concept by smaller municipalities.

As stated by Hummel et al. [76] and Browning et. al. [48], the financial cost of ALS-based urban greenery inventories may be out of reach for some municipalities. However, ALS data might be useful for municipalities where tree surveys are not a priority, for instance, in urban domains connected with flood risk assessment, logistic network management, spatial planning, risk assessment, emergency management, advertising, research, promotion, and visualization. Therefore, a complex ALS solution can be a source of multidisciplinary data, which is a crucial element of a city management system.

The application of remote sensing tools and data to the realization of the 3-30-300 concept can theoretically provide spatially continuous information about every tree within an area of interest (AOI). This seems to be a great advantage over field-based inventories, as used by Croeser et al. [57], where only publicly accessible trees were surveyed (usually those along main streets). Furthermore, a larger number of green spaces requires the deployment of numerous surveyors on site, which incurs additional costs. In addition, field crew members vary in experience and may be prone to subjectivism. In contrast, the cost of objective RS-based methods does not increase linearly with the extent of an AOI [74]. Such an approach could help to further optimize tasks related to urban greenery inventories. Furthermore, the 3D character of RS data makes them an attractive material for publicity and visualization for future planning.

As a rule, most methods require successive optimization steps either before or during their implementation. The 3-30-300 rule is a general concept that offers many possibilities but also poses some challenges in practical implementation. From a technical point of view, an accurate assessment of the visibility component seems to be the most complex part of this method. Gathering and automatically classifying in situ images from each window seems most reliable [48] but is not practically feasible. Alternatively, one could conduct a survey, but it should always be analyzed with some confidence intervals regarding the response rates, their credibility, and their bias. In such situations, RS techniques could be very useful. As Croeser et al. [57] stated, three-dimensional spatial data (e.g., point clouds) could provide a robust material for visibility assessment in a single window. On the other hand, this would require enormous computational resources.

A more detailed analysis would require an examination of the viewing angle, i.e., the relation of a fixed/floating viewpoint to the window size, as well as the inclusion of any other facilities, such as balconies or terraces. Moreover, human eyes also have a main focus spot, prioritizing objects around the central part of the field of view over those situated at marginal regions. Therefore, to ensure the feasibility of the analysis, some aspects pertaining to visibility assessment had to be generalized in our study. Namely, there was only one observer for each side of the building, standing on the ground floor (1.7 m). In our approach, buildings as entities could meet the visibility requirement in the three variants described in the methodology. Again, this was just an arbitrary distinction originating from the lack of a precise definition of the visibility component. Therefore, this aspect should also be analyzed in more detail if the method is to be introduced at a larger scale. Nonetheless, the authors believe that the assumed simplifications could influence the results of the visibility assessment in larger cities, where there are more high-elevation buildings than in towns with low-rise buildings. All in all, the visibility component of the 3-30-30 method still requires further analysis regarding the definition of this component, as well as the development of specific algorithms that would facilitate its accurate determination.

Knowledge about the spatial distribution of trees can help to view the aspect of visibility from a reciprocal perspective. It is also possible to identify high-priority trees, i.e., trees that are seen by many observers (Figure 4). This type of information could be useful for planning revitalization measures or other urban infrastructure investments. For example, our approach makes it possible to highlight weak spots, i.e., places where new planting is needed according to regulatory requirements and the actual expectations of residents.

Two main approaches are usually related to the determination of the “300” component: simple Euclidean distances and network analysis. The latter, which is relatively more accurate, is, at the same time, more complex. Proximity of a green area, e.g., a park, does not always ensure easy access. For instance, in the case of large fenced areas, the entrances constitute converging points. In order to determine the entry distance, detailed information about the spatial distribution of routes, pavement, junctions, etc., as well as their types and qualities, would be needed. On the other hand, one could reap the benefits of a park without accessing it just by being in its vicinity and/or by enjoying a view of it. This was the main reason why the simpler (Euclidean) approach was applied in this study. In our opinion, this might be a reliable proxy for a generic analysis and could be implemented in a greater number of municipalities, as few probably possess and would be willing to share detailed data on their communication networks.

Lastly, it seems paramount to emphasize that, currently, the 3-30-300 rule is a purely quantitative concept. It says little or nothing regarding the qualitative characteristics of green areas, which very often take precedence over pure numerical juxtapositions. Moreover, relying only on numerical summaries can lead to erroneous decisions. On this occasion, it is important to mention some qualitative features that could relatively easily expand the potential of the 3-30-300 method by using RS data. A complex spatial survey of the condition of trees could be carried out by calculating vegetation health indices such as the NDVI. Modern, relatively inexpensive unmanned aerial vehicles can be equipped with multispectral optical sensors that enable the creation of point clouds and increase the range of spectral information. This facilitates the ability to classify species, which could help determine biodiversity and differentiate subjective perceptions of the attractiveness of certain green spaces. This could also help identify and promote smaller or unknown places that are worth visiting.

Urban spatial management has become an important aspect for many cities. One of the biggest challenges is the provision and maintenance of high-quality green spaces. Other challenges include adapting to climate change and mitigating hazards resulting from extreme weather conditions, e.g., heavy rainfall, drought, and high temperatures. The United Nations’ strategies and local conditions require municipal governments to develop strategies, policies, and programs to address the above issues [29]. For example, after 2024, all municipalities in Poland with at least 20,000 inhabitants will have to develop a local climate change adaptation strategy [77]. Therefore, in this article, the authors explored the possibilities of the methodological implementation of the 3-30-300 rule, which offers certain solutions for urban spatial management focused on the well-being of citizens. The use of a single remote sensing dataset leads to a relatively simple and standardized implementation of the proposed methodology in municipalities. It is all the more universal as the ALS data allow for analysis at flexible resolutions, the results of which could support decision-making chains at different scales (individual trees and buildings, streets, parks, neighborhoods, cities, agglomerations, and regions). Finally, with the proposed approach, it is relatively easy to mark the areas that are far from the minimum thresholds of the 3-30-300 rule. Such information can be helpful in objectively prioritizing tasks related to new initiatives and investments.

Local conditions play a crucial role in meeting the 3-30-300 provisions, e.g., when dense urban structures or unfavorable hydroclimatic conditions affect planting success and vegetation mortality rates [78]. The potential opportunities for shaping the 3-30-300 rule are also related to the size and density of urban infrastructures [79]. The growth of cities puts excessive pressure on existing infrastructure, making the creation and maintenance of existing green spaces an expensive and difficult task [80]. In order to overcome the aforementioned difficulties in fulfilling the criteria of the 3-30-300 rule in larger cities, several concepts have been presented by different authors. For example, Verheij et al. [81] point out the possibilities of ensuring public access to privately owned green spaces. Another approach described by Karteris et al. [80] utilizes the vertical dimension by proposing the implementation and development of green roofs.

It should also be considered that social acceptance of changes and investments can be an important factor. Therefore, measuring the actual use of green spaces and involving residents in the decision-making processes seems to be of utmost importance in a just society [82]. Such an approach proves invaluable when it comes to negotiations between different parties, such as decision makers, investors, and residents, where heated debates, often infused with subjective but important emotions, can be supported by impartial information.

5. Conclusions

The method presented in this study is a proposal of the adaptation of the 3-30-300 concept that demonstrates its practical applicability to the situation of smaller municipalities. In contrast to previous studies, the overriding premise of our method is based on the use of point clouds throughout the analytical process as the main source of data. The relatively simple implementation provides objective results that can be compared between towns, which, in turn, could be a valuable and objective decision-making tool. The method presented in this study was limited to a quantitative analysis based on the spatial distribution of green spaces. However, many aspects related to the assessment of qualitative features, as well as people’s subjective perceptions, were discussed, and they certainly need to be taken into account in future adaptations.