Geospatial indicator analysis as a tool for managing urban blue-green infrastructure in cities

The general model developed for a method of selecting blue-green infrastructure solutions with applied indicators, based on the concept of ESS, can be described as structured steps:

• 1)

  selection of indicators to locate areas of the city with reduced resilience and conduct of spatial analyses oriented to threats and derivatives of climate phenomena within the city's structure;

• 2)

  analysis of the results of calculated indicators within the adopted value range system for spatial units of the city;

• 3)

  ranking of the threat of adverse effects of climate change and increasing urbanization;

• 4)

  determination of ESS and selection of actions to mitigate or reduce the resulting risks, depending on the values and interrelations of the indicators.

The method is based on the use of selected indicators to identify a problematic area or area suitable for potential implementation. Indicators inform about the location of particularly strong intensity and density of problematic areas of a given type. Identification of such areas is carried out on two levels. The first level involved a spatial visualization of the elements of land use or functions of the area in the place of their actual occurrence regardless of the administrative division.

The second level involved the use of a municipal auxiliary unit corresponding to the LAU2 scale—statistical district, neighbourhood or settlement—for ranking and prioritizing intervention areas. In the literature, there are many cases of spatial analyses concerning broadly defined geographic indicators based on other spatial units [Birch et al. 2007; Wanga et al. 2018; Burdziej 2019]; however, selected spatial units can be effectively managed and funded, for example, through a green citizen budget or integrated neighbourhood development plans. Therefore, the indicators were calculated for spatial units selected by the city, most often districts.

For each city, a consistent set of indicators was applied to describe and rank districts and to indicate areas where hazards exist. High temperature, the occurrence of an urban surface heat island, strong sealing density causing rapid rainwater runoff, areas with poor quality and low state of irrigated vegetation, and areas with poor accessibility to green space, were adopted as the main threats to cities resulting from intensive urbanization related to a changing climate. These are factors that can significantly affect the health and quality of life in the city; however, they can affect also the condition of built-up areas and the quality of public spaces [WHA61.19 2008]. In response to the defined threats, six groups of indicators were identified: sealing, Surface Urban Heat Island (SUHI), the Normalized Difference Vegetation Index (NDVI), biologically active area (BAA), availability of green space, and development and selected forms of land use (e.g., parking lots, historic buildings.) Individual indicators were calculated as follows:

• a)

  S1: average sealing level;

• b)

  S2: average sealing intensity level;

• c)

  S3: ratio of the area of sealed surfaces with values above 60% relative to the spatial unit area;

• d)

  S4: ratio of built-up area to sealed surfaces with values above 60% relative to the spatial unit area;

• e)

  SUHI: ratio of the SUHI area to the spatial unit area;

• f)

  BAA: ratio of the BAA to the spatial unit area;

• g)

  NDVI: ratio of the area where NDVI exceeds 0.2 to the spatial unit area;

• h)

  AV: ratio of residential area not covered by a 300-m AV buffer to residential area;

• i)

  B: dominant type of building development for spatial units;

• j)

  P: parking area in the spatial unit [m²];

• k)

  HB: share of historic buildings in the building area [%].

Indicators a–c and e–h represent threat indicators. The remaining indicators (i–k) are identified as auxiliary indicators, i.e., identifying conditions.

The method of calculating individual indicators affects their interpretation; therefore, the precise calculation method is presented in the results section.

Based on the calculated indicator values for the districts, a matrix was created to position the value of each indicator in relation to the respective city districts. To determine the value ranges, a quantile classification method was applied, dividing the data into six classes ranging from 0 to 5 (Table 1). The first set denotes the values for which a higher group is associated with higher risk, and the second set represents values for which a lower group is associated with higher risk.

The two factors linking spatial analysis stages with real GBI planning are:

• 1)

  levels of classification calculated based on district indicator values (indicator value interval groups), and

• 2)

  ESS or their derivatives, identified as missing, mitigating, or minimizing a given threat, depending on the type of problem and its intensity occurring in the district or specific location.

How to select GBI solutions in combination with a set of indicators is shown in Table 2. The application of ESS in the method of selecting solutions allows for targeting actions that need to be implemented for specific conditions. Green and blue solutions provide a variety of ESS beneficial for urban environments. Some of them have a particular impact on the examined threats related to climate change. These include water flow control and prevention of excessive stormwater runoff, temperature regulation (mitigating the urban surface heat island effect), and mitigation of extreme phenomena such as floods or droughts. Elements of GBI, in the context of ESS, serve significant functions such as water retention in soil, rainwater storage or capture, groundwater infiltration, regulation of evapotranspiration processes, and absorption of kinetic energy. The above functions have been adopted as the main differentiating directions for actions. As part of the project, a group of experts developed a catalog of 40 green and blue solutions directly linked to the analyzed factors.

Table 2.

Indicators supporting the selection of GBI solutions should comprehensively identify the physical space of occurrence of hazards and sensitivities that weaken the city, understood as a multiple complex structure or an organism composed of interrelated networks [Chmielewski 2010; Filip 2015]. The decision-making process aided by indicators helps identify potential hazards as a consequence of climate change or conditions that intensify its effects in urbanized space. Seven main indicators, supplemented by four additional ones, were analyzed across four cities: Olsztyn, Rzeszów, Sztum, and Bielsko-Biała. The results of the study in this paper are presented based on the city of Olsztyn only.

In the first group of indicators, the density of sealing was addressed, which has become an important issue within the contemporary context of climate change. Strong surface sealing generates numerous environmental challenges in cities, including the SUHI effect associated with the intense heating of artificial surfaces. Additionally, the overload of the sewage system caused by the rapid runoff of rainwater over sealed surfaces leads to more frequent floods. Sealing analysis allows for the identification of built-up areas or surfaces that are impermeable to rainwater infiltration or allow infiltration to a very limited extent. As part of the sealing analysis, four indicators concerning spatial units were determined and interpreted:

a) S1: The average sealing level describes the percentage value of average sealing for all surface types and land cover types, including green areas, relative to the spatial unit area. This allows for a preliminary estimate of the condition of the city unit's surface coverage causing rapid runoff of rainwater into the sewage system, and therefore the amount of rainwater that cannot be infiltrated. The indicator ranks districts in terms of the level of occurrence of the sealing problem, but taken alone it cannot be the basis for selecting adaptation actions: areas with significant disparities in land use result in a high generalization of the indicator. A low level of the indicator may indicate a large area of green spaces within the unit, but it does not exclude the occurrence of sealed areas requiring intervention.

The average sealing level (S1) was calculated based on the raster database of the Copernicus Program, determining the percentage of sealed surfaces within raster pixels, and it is calculated according to the formula: x¯=1n∑i=1xi {\rm{\bar x}} = {1 \over {\rm{n}}}\sum\limits_{{\rm{i}} = 1} {{{\rm{x}}_{\rm{i}}}} where n – number of observations: the number of pixels in a district, and x_i represents single observations x_i...x_n: the average sealing value for a single pixel.

The sealing density is in the range from 0% to 100%, with a spatial resolution of 10 m.

b) S2: The average sealing intensity level describes the average sealing level for sealed areas relative to the spatial unit. The indicator does not include non-sealed areas such as agricultural land and green areas, so it more accurately depicts the actual level of sealing within urban areas. The amount of the indicator identifies the scale of the problem but does not define the phenomenon of sealing in a quantitative context. The indicator does not differentiate areas by land use type, so the selection of solutions specific to dense urban development and areas covered with impervious surfaces must be further differentiated by indicator S4 and is dependent on the predominant type of development within a given area. The average sealing intensity level was calculated according to the formula for S1, except that it applies to areas with soil sealing ranging from 1% to 100%, with a spatial resolution of 10 m.

c) S3: The ratio of sealed areas with a value of more than 60% in relation to the area of the auxiliary unit informs about the coverage of the area of this unit with a sealed area with low and very low levels of rainwater infiltration. A high level of the indicator indicates a large amount of impervious surfaces in the unit or dense development, so it determines the quantitative range of measures needed to be implemented. Like the average sealing intensity index, the index does not distinguish between land developed with buildings and land sealed by impervious pavement. This information is supplemented by indicator S4.

The ratio of sealed areas with a value of more than 60% in relation to the area of the spatial unit, is calculated according to the formula: Sealed surface area > 60% [m2]Spatial unit surface area [m2]×100 [%] {{{\rm{Sealed}}\,{\rm{surface}}\,{\rm{area}}\, > \,60\% \,[{{\rm{m}}^2}]} \over {{\rm{Spatial}}\,{\rm{unit}}\,{\rm{surface}}\,{\rm{area}}\,[{{\rm{m}}^2}]}} \times 100\,[\% ]

d) S4: The ratio of built-up areas to sealed areas with a value of more than 60% in relation to the area of the auxiliary unit differentiates areas developed with buildings from areas covered with impervious surfaces, therefore complementing the other indicators related to sealing. The indicator allows directional differentiation of solutions for dense development and for sealed soil covered with impervious surfaces. A high value of the indicator indicates a high density of development, while a low value of the indicator indicates the predominance of impervious surfaces that are distinctive for the infrastructure accompanying the built-up areas and for communication areas.

The ratio of built-up area to sealed areas with a value of more than 60% in relation to the area of the spatial unit [S4], is calculated according to the formula: Build-up surface area [m2]Sealed surface area > 60% [m2]×100 [%] {{{\rm{Build}} - {\rm{up}}\,{\rm{surface}}\,{\rm{area}}\,\,[{{\rm{m}}^2}]} \over {{\rm{Sealed}}\,{\rm{surface}}\,{\rm{area}}\, > \,60\% \,[{{\rm{m}}^2}]}} \times 100\,[\% ]

Indicators S1, S2, S3, and S4 depicting the level of sealing in auxiliary units complement each other, so they should be considered together. Indicator S1 forms the basic ranking for units, prioritizing actions. Activities in units with the highest value of sealing rates should be undertaken first. In these areas, there is a tendency for soil sealing hazards to deepen most rapidly. Units with a high S2 indicator require intensified and efficient measures in terms of rainwater infiltration, including spot measures targeting areas with the highest sealing. In units characterized by a higher S3 index value, point and systemic measures should be taken in parallel, since there is a sealing problem which is widespread and applies to areas of significant size.

Indicator S4 distinguishes whether buildings or other impervious surfaces predominate in a sealed area. A value S4 > 50 % signifies that buildings occupy more than half of the impervious area, whereas S4 < 50 % indicates that pavements, roads, or other non building surfaces dominate. Because no universal threshold exists for this indicator, the 50 % cut off was chosen pragmatically, guided by related research. For example, Tang et al. [2025] analysed the cooling effects of 40 urban green spaces and found that heat mitigation benefits varied with the “core proportion” of green space; cooling performance plateaued when the core proportion was around 20%–55 % and increased markedly beyond 55 %. This suggests that mid range thresholds around 50 % can mark shifts in thermal performance. Similarly, Mehrotra et al. [2019] used the Built Coverage Ratio (BCR) to explain urban heat island intensity; they reported that increases in built up fractions (particularly when built ratios exceeded 50 %) were associated with heightened microclimate stress in heterogeneous urban forms (see their geographically weighted regression results). Xu et al. [2023] compared satellite and GIS based local climate zone (LCZ) classifications for Guangzhou; LCZ schemes differentiate compact and open building forms using density thresholds that often hinge on or exceed 50%, distinguishing heavily built up zones from green or low density areas. Collectively, these studies justify using a midrange (around 50%) benchmark to indicate when buildings rather than other impervious surfaces dominate.

The high value of the S4 indicator should target solutions to those using building elements to create green roofs and walls, as well as the use of green vertical elements overgrown with climbing plants in public spaces, especially in the space of road lanes. The low value of the S4 index requires intensive greening and unsealing of the land, e.g., by replacing surfaces to allow full or partial infiltration of rainwater. The high level of sealing indexes, combined with the presence of a SUHI, requires immediate action to reduce the temperature in public spaces.

As a result of the presence of surface sealing and buildings accumulating thermal energy, the temperature in specific areas in the city is higher than in suburban and rural areas, causing the so-called effect of a UHI [Zhou 2015]. An increase of the temperature of the ground surface has significant implications for the city's ecological system and causes changes in the structure of this system. Among other things, it affects the hydrological situation, soil properties, and the health of citizens [Yang et al. 2016]. High temperatures also affect the functionality and usability of public spaces, especially those intended for pedestrians, which are more often and more willingly used if they provide thermal comfort to the users [Kleerekoper et al. 2012]. The indicator determining the occurrence of SUHIs in the city was defined as the ratio of the SUHI area to the area of the spatial unit. The extent of the heat island was increased by a 100-m buffer around the designated high-temperature areas. The indicator identifies spatial units, particularly affected by higher temperatures, where it is necessary to implement systemic solutions to reduce the amount of accumulated thermal energy. A high SUHI correlates with high sealing rates, so the key objective of the solutions used for high SUHI values should address solar reflection or shading, as well as cooling by the evapotranspiration process. This approach requires the use of appropriate qualitative selection of vegetation (species, habit, height) and the selection or replacement of structural and material solutions for canopies of large-scale facilities and building roofs.

Urgent implementation actions are required in areas where a high level of sealing-related risk and a low number of BAAs are observed simultaneously. The shortage of BAAs contributes to exacerbating the negative consequences of soil sealing. Reducing the negative effects of rapid urbanization and consequent urban sealing requires solutions that emulate natural hydrological processes, especially through urban greening and sustainable stormwater management [Artmann 2016]. Urban greening is significant due to numerous ecosystem benefits and its positive impact on residents' well-being and health [Jim et al. 2004]. The BAA index was used to determine the quantitative status of areas providing plant vegetation in the city. This indicator illustrates the ratio of such areas to the total area of a spatial unit. A high level of the indicator points to a significant proportion of biologically active land in the spatial unit. BAAs are crucial for determining the city's adaptability level and sensitivity to climate change and involve a wide range of ESS. Greening practices are aimed not only at using rainwater in rainfall places, but also at purifying the air, lowering the temperature, and creating a good quality living environment. It is therefore important to select solutions that make wide use of GBI elements. These can be spot actions called green acupuncture in the existing urban fabric or the designation of new green areas in the city. Actions should achieve goals such as increasing biodiversity and integrating fragmented habitats resulting from urban structural development. The BAA index should be considered in conjunction with the NDVI, used to detect vegetation coverage of the terrain and assess its condition. Values of the calculated NDVI index range from −1 to 1. Values of the index for vegetation-covered areas have positive values between 0.2 and 1 [Bhandari et al. 2012]. The NDVI was calculated as the ratio of the area where NDVI exceeds 0.2 to the area of the spatial unit, allowing the indication of vegetation cover at least in good condition. The interpretation of the NDVI combined with the biologically active land area index served as a determinant of the risk associated with low vegetation quality caused by factors such as urban drought and low biodiversity.

The last of the indicators identifying potential threats is the green-space AV index. Accessibility to green spaces is currently one of the determinants of the so-called 15-minute city, indicating a functional living environment in the city. Contact with nature, located directly adjacent to residential areas, is essential for maintaining the mental well-being and mitigating environmental stress of residents [Gidlöf-Gunnarsson, Öhrström 2007]. The accessibility of green spaces in cities is now a key area of planning considerations. In many cities, recommendations are emerging regarding the minimum distance between residential areas and green spaces [Kabisch et al. 2016]. The green spaces AV index was calculated as the ratio of the area of residential zones not covered by a 300-m accessibility buffer to the total area of residential zones. A high AV index indicates a large area of residential areas located more than 300 m away from green spaces. Low accessibility requires organizational actions to designate new urban green spaces or transform existing areas towards greening. If the area is heavily urbanized, green acupuncture should be intensified so that small-scale solutions such as pocket parks can supplement missing ESS.

In addition to the threat indicators, additional auxiliary indicators defining local conditions were determined. To diagnose constraints and potentials, the dominant type of development for each spatial unit was identified, as well as the area of parking spaces and the share of historic buildings in the built-up area.

Parking area (P) measures the share of parking in the spatial unit. The United Nations' “Five Principles” for sustainable neighbourhood planning note that efficient street networks and parking typically require 20%–30% of urban land, rising to 40%–60% in commercial centres [United Nations Human Settlements Programme (UN Habitat). A New Strategy of Sustainable Neighbourhood Planning: Five Principles. UN Habitat, 2014]. So, a provisional threshold of P > 20% was adopted.

Indicator HB expresses the proportion of buildings subject to heritage or conservation protection. Given the wide variation in protection criteria and inventories across countries and districts, an illustrative threshold of HB > 10% is applied. In Trondheim, Norway, heritage protected buildings make up about 10% of the building stock, but this is not typical across Europe [Bjelland et al., 2025]. A 2022 report for the Concerted Action on the Energy Performance of Buildings Directive notes that approximately 35% of EU buildings are over 50 years old and that member states lack consistent data on heritage buildings. Therefore, any quantitative threshold should be adapted to local and national contexts, considering both the age of the building stock and the extent of conservation regulations.

The existing development of urbanized space can significantly limit or disable the application of specific GBI solutions within a given spatial unit. The dominance of compact urban development indicates the need for small-scale but systemic solutions that have a small footprint, allowing for the integration of diverse functional and transportation needs. It is also necessary to use vertical elements and fragments of buildings or structures for greening. In such spaces, small-scale solutions with a wide range of ESS or linear actions that connect and allow movement between larger patches of green spaces work best. Residential multifamily housing poses challenges in shaping public and semi-private green spaces with strong public support and advocacy. The solutions can cover a larger area and contribute more to improving biodiversity in the city. At the same time, GBI elements in housing estates should create coherent systems enabling efficient management, including stormwater management. Single-family development requires solutions tailored to the individual needs of private property owners. Regardless of the type of development, education and financial support as well as the dissemination of good practices are necessary, along with linear solutions used in road corridors.

Some areas in the city indicate potential opportunities for systemic action, such as paved parking lots, which can be gradually transformed to achieve a better level of rainwater infiltration. Multihectare parking spaces serve as a surface resource for retaining water in underground tanks. At the same time, through permeable and semi-permeable surfaces, surface runoff of rainwater can be significantly reduced.

The city of Olsztyn is located in the northeastern part of Poland. It covers an area of 8832 hectares as a county-level city and serves as the capital of the Warmian-Masurian Voivodeship. It is the central city of the Olsztyn agglomeration. The population for 2022 was 168,21 thousand [BDL 2022]. Population density is 1904.6 people per km² [BDL 2022]. Olsztyn is dominated by built-up and urbanized land (42.34% of the area) [POŚ, 2020], but there are many green areas located within the city, including 103.05 hectares of walking and recreational parks and 1351.64 hectares of communal forest land, which shapes the ratio: 80.4 m² of communal forest land per capita [BDL 2022]. A significant part of biologically active land is occupied by land under water: 9.62% [POŚ 2020]. Olsztyn is divided into 23 districts (i.e., local communities of residents). For the city of Olsztyn, indicators were calculated according to the methodology's specified division into five groups: sealing (indicators S1, S2, S3, S4), SUHI, BAA, NDVI, and AV. The purpose of using the indicators is to identify potential threats, which are the basis for determining the ESS that are missing in the area and need to be introduced. The types of threats and missing ESS, in turn, form the basis for determining strategic directions of action, within which subsequent implementations of GBI should be carried out.

The average sealing level S1 in the city of Olsztyn is 5.3%– 55.4% and is much lower than the average sealing intensity level S2, which is 37.6%–71%. The difference in the value of S1 and S2 indicators reflects the occurrence of spatial units within the city with a significant number of green areas and the need to differentiate the intensification of activities in the individual areas. A statistical outlier value of the U2 indicator shows the occurrence of a spatial unit where a particular threat associated with strong sealing has been diagnosed. At the same time, the strong sealing density (> 60%) of most spatial units does not reach 30% of their area. The S4 index at 45%–65% for most spatial units shows the predominance of built-up area over sealed surface. GBI solutions should therefore include the adaptation of facilities and building elements as an important support for the GBI system. The outliers for the SUHI index reveal the presence of two spatial units under the strong influence of the SUHI. The index of the amount of BAA in the spatial unit represents a wide range of values, which indicates the significant diversity of land use in the city area. The irregular distribution of values of the AV index indicates unequal access to green space for residential areas in the individual spatial units. A statistical chart of indicator values for the city of Olsztyn is illustrated in Figure 1. A summary of the indicators for each spatial unit is illustrated in Figure 2 and Table 3.

Figure 1.

Statistical chart of indicator values for the city of Olsztyn. The graph shows the distribution of data in quartiles, distinguishing the mean (−), median (x), and outliers (single dot). Vertical lines mark the range of values from minimum to maximum, and any point that lies outside the lines is an outlier.

Figure 2.

Map of Olsztyn divided into spatial units. Based on the BDOT10k database

Table 3.

Olsztyn's spatial units are variable in terms of the intensity of risks occurring (Table 3). In the city space, there are units with a high degree of threat identified using all or most of the applied indicators (SU: 6, 7, 8, 10, 18, 20, 22). Some spatial units show moderate threat levels (SU: 3, 4, 14, 16) or low threat levels (SU: 1, 2, 5, 9, 11, 12, 13, 15, 17, 19, 21, 23). High threat levels concern certain areas with a predominance of neighbourhood (B) development, compact, or other types of development such as industrial. High levels of indicators are not achieved for areas with predominantly single-family housing. Sealing indicators S1, S2, and S3 are diverse concerning each other: when selecting detailed solutions, they should be interpreted together with indicator S4 determining the predominance of buildings (S4 > 50%) in the sealed area or the predominance of tight surfaces in the sealed area (S4 < 50%). Targeted solutions should also be diversified based on the potential surface area of spaces that can be unsealed, such as parking areas (P > 20%), or due to the presence of a large number of buildings under conservation protection (HB > 10%). All high-hazard spatial units are characterized by two or more high sealing index values. High threat level due to the presence of the SUHI correlates with strong soil sealing but does not apply to all heavily sealed areas. A high S4 indicator particularly concerns areas with predominantly single-family housing. A U4 indicator value below 50% indicates extensive infrastructure accompanying urban development that needs to be unsealed.

The districts subject to the highest threat are characterized by high values of most indicators; however, the intensity of the problems varies. The method adopted in the work assumes a combined interpretation of indicators within the city's spatial units. This not only leads to setting priorities for action but above all to diversifying solutions and determining strategic directions that differentiate actions.