Accessibility to green area qualities in the Stockholm region and their possible correlations to property values

IN THE FIELD OF TECHNOLOGY

DEGREE PROJECT

ENERGY AND ENVIRONMENT

AND THE MAIN FIELD OF STUDY

ENVIRONMENTAL ENGINEERING,

SECOND CYCLE, 30 CREDITS

,

STOCKHOLM SWEDEN 2020

Accessibility to green area

qualities in the Stockholm region

and their possible correlations to

property values

A GIS-based network analysis

ANNA TONNER

Accessibility to green area

qualities in the Stockholm

region and their possible

correlations to property values

A GIS-based network analysis

ANNA TONNER

Supervisor

Ulla Mörtberg

Examiner

Ulla Mörtberg

Degree Project in Environmental Engineering and Sustainable Infrastructure KTH Royal Institute of Technology

School of Architecture and Built Environment

Summary in Swedish

Konceptet tillgänglighet har under de senaste åren använts mer frekvent inom stadsplanering, där tillgänglighet till urban service och attraktiva platser, så som grönområden, är önskvärt. I en snabbt växande stadsregion som Stockholm är tillgänglighet extra viktigt för att bidra till en attraktiv och hållbar region. Ett flertal studier stödjer att det finns en positiv korrelation mellan grönområden, mänskligt välmående och att upprätthålla en god livskvalitet. Det är dock snarare kvaliteterna ett grönområde besitter som är attraktivt och inte nödvändigtvis själva ytan. Följaktligen kan tillgänglighet till grönområden anses påverka fastighetsvärdet och kan vidare ge en indikation om vilka grönområdeskvaliteter som efterfrågas, då det kan variera för olika grönområdestyper med olika kvaliteter. Trots att studier angående tillgänglighet till grönområden och korrelationen till fastighetsvärden redan har utförts i viss utsträckning i Stockholm, saknas kombinationen med grönområdeskvaliteter på en regional nivå med ett nätverksanalytiskt angreppssätt.

Syftet med denna studie är att mäta tillgängligheten till grönområdeskvaliteter i Stockholmsregionen och vidare att bedöma om det uppmätta avståndet mellan fastigheter och grönområden korrelerar med fastighetsvärden i regionen. I studien väljs fyra grönområdeskvaliteter baserat på upplevelsevärden, vilka är: Rymd, Rofylldhet, Park och Skyddade områden för biodiversitet. Dessa kvaliteter antas bidra till ökad livskvalitet, men dess användbarhet inom regional planering skulle kunna vara mer attraktiv om de kombineras med andra kvaliteter i analysen. Tillgängligheten mäts från varje bostad i Stockholmsregionen till dessa grönområdeskvaliteter, via gångvägsnätet, för olika geografiska områden inom regionen, vilka är: Skärgård, Tätortsnära landsbygd, Landsbygd och Tätort. Detta utförs genom att använda en GIS-baserad nätverksanalys. Korrelationen mellan det uppmätta avståndet och fastighetsvärden per areal bestäms genom att använda Pearsons korrelationsmetod.

Resultatet från tillgänglighetanalysen visar på att tillgängligheten till rymliga grönområden generellt sett är bra eller mycket bra i hela regionen, medan tillgängligheten till rofylldhet och skyddade områden för biodiversitet generellt sett är dålig. Tillgängligheten till parker är endast uppmätt för urbana områden på grund av begränsad kartering och resultatet visar på att majoriteten har bra eller mycket bra tillgänglighet. Följaktligen indikerar resultaten att tillgängligheten skiljer sig för de olika geografiska områdena inom regionen och det kan konstateras att urbana områden generellt sett har sämre tillgänglighet än landsbygdsområden. Dessutom indikerar resultaten att tillgängligheten varierar mellan de olika grönområdeskvaliteterna. Användbarheten av tillgänglighetsresultaten antas vara relevant för framtida regionala planerare som ämnar fokusera på dessa specifika kvaliteter. För att få ett mer heltäckande perspektiv angående befolkningens livskvalitet i regionen behövs dock även andra grundläggande kvaliteter inkluderas i tillgänglighetsanalysen. Kvaliteterna i denna studie kan dock bidra som en del i bedömningen av regional tillgänglighet med fokus på att upprätthålla en god livskvalitet.

Den statistiska analysen visar att alla olika dataset har en positiv korrelation mellan det uppmätta avståndet och fastighetsvärdet per areal, vilket innebär att fastighetsvärdet ökar för bostäder som ligger långt ifrån dessa grönområdeskvaliteter. Undantagen för detta är: Park, i urbana områden, och Skyddade områden för biodiversitet, i tätortsnära landsbygd, som istället har en negativ korrelation. Detta innebär att fastighetsvärdet ökar för bostäder som ligger nära parker i urbana områden och skyddade områden för biodiversitet i tätortsnära landsbygd. Studien visar även på att korrelationerna mellan samtliga nämnda variabler är svaga. Dessa svaga korrelationer antas bero på inverkan av andra samverkande lägesfaktorer, såsom tillgänglighet till stadskärnan, kollektivtrafik och annan urban service, som möjligtvis har direkt motsatt korrelation till dessa grönområdeskvaliteter. Samtliga korrelationer är dock statistiskt signifikanta, vilket innebär att slutsatsen kan dras att det finns en sann korrelation, även om den är svag. För att få det statistiska resultatet mer användbart för regional planering kan även andra lägesfaktorer analyseras för att få en bättre förståelse för det statistiska resultatet. Slutligen kan resultaten från både tillgänglighetsanalysen och den statistiska analysen användas som ett underlag för framtida planering och som en del av ett rumsligt beslutsstöd för att upprätthålla en god livskvalitet inom Stockholmsregionen.

Abstract

The concept of accessibility has in recent years been more used in urban planning, where access to urban services and attractive places, such as green areas, is desirable. In a rapidly growing city region like Stockholm, accessibility is especially important in order to provide an attractive and sustainable region. Several studies have supported the positive correlation between green areas, human well-being and sustaining a good quality of life. However, it is rather the qualities possessed by a green area that is attractive and not necessarily the space itself. Moreover, access to green areas is considered to affect the property values and can further indicate whether a green area quality is demanded or not since it may differ depending on the green area type and its quality. Although studies concerning accessibility to green areas and the correlation to property values already have been conducted to some extent for Stockholm, the combination of qualitative green areas have not been extensively researched for the entire region with a network analysis approach.

The purpose of this study is to measure the accessibility to green area qualities within the Stockholm region and further assess whether the measured distances correlate with the property values within the region. In this thesis, four green area qualities are selected, based on experience values, which are: Spacious, Quietness, Parks and Protected areas for biodiversity. These green area qualities are considered to contribute to individual’s quality of life, but for their usefulness from a regional planning perspective, a combination of different qualities could have been more attractive. The access is further measured from each dwelling in the Stockholm region to these green area qualities, via the pedestrian road network, for different geographical divisions within the region. These four divisions are: Archipelago areas, Urban countryside areas, Countryside areas and Urban areas. The accessibility analysis is conducted by using a GIS-based network analysis. The correlation between the measured distances and the property values per area is determined by using the Pearson correlation method. The results show that the access to spacious green areas generally is at least good in the entire region, while the access to quietness and protected areas for biodiversity generally is poor. The access to parks is only measured for Urban areas, due to limited mapping, and the results show that a majority have at least good access. Also, the results show that dwellings generally have best access to spacious green areas, while they have poorest access to quietness within the entire region. Dwellings located in the outer parts of the Urban areas had poorer access to Parks than in the more central areas. Moreover, the results indicate that there are differences in access for different geographical divisions within the region and it can be concluded that Urban areas have generally poorer access than the more rural areas.

It is found that all different datasets have a positive correlation between the measured distance and the property values per areal, which means that dwellings located far away from these green area qualities have higher property values. The exceptions were for the green area quality: Parks, in Urban areas, and the green area quality: Protected areas for biodiversity, in Urban countryside areas, which instead had a negative correlation. Basically, this means that these dwellings located close to parks in Urban areas and close to protected areas for biodiversity in Urban countryside areas have higher property values. The study further shows that the correlation between all datasets is weak. These weak correlations are however assumed to be affected by other location factors that may have directly opposite correlations to these green area qualities, such as access to the city center, public transport and other urban services. However, all correlations in this study are found to be statistically significant, which mean it can be concluded that a true correlation exists, although it is weak. For the statistical results to be more useful from a regional planning perspective, other location factors could be analyzed as well, in order to get a better understanding of the statistical results.

Finally, the results from both the accessibility analysis and the statistical analysis can be used as a basis for future planning and as a spatial decision support to sustain a good quality of life within the Stockholm region.

Key words: Accessibility, Network analysis, GIS, Green area, Green area quality, Experience values,

Acknowledgements

Table of contents

3 STUDY AREA: STOCKHOLM COUNTY ... 6

3.1 THE COMPACT CITY ... 7

3.2 GOOD QUALITY OF LIFE ... 8

3.3 GREEN AREAS IN STOCKHOLM ... 9

3.4 GEOGRAPHICAL DIVISION ... 9

4 QUALITY MEASURES FOR GREEN AREAS ... 11

4.1 QUALITY APPROACH ... 12

4.2 EXPERIENCE VALUES ... 12

4.3 GREEN AREA QUALITIES ... 13

4.3.1 Spacious ... 13

4.3.2 Quietness ... 13

4.3.3 Parks ... 14

4.3.4 Protected areas for biodiversity ... 14

5 PROPERTY VALUES RELATED TO GREEN AREAS ... 15

6 METHODOLOGY... 16

6.1 LITERATURE REVIEW ... 16

6.2 GIS-BASED ANALYSIS ... 16

6.2.1 Determination of geographical divisions ... 17

6.2.2 Spatial data collection ... 18

6.2.3 Preparation and processing of spatial data ... 21

6.2.4 Network analysis model ... 26

6.3 STATISTICAL ANALYSIS ... 27

7 RESULTS & ANALYSIS ... 29

7.1 ACCESSIBILITY TO GREEN AREA QUALITIES ... 29

7.2 STATISTICAL RESULTS ... 44

8 DISCUSSION ... 48

8.1 GENERAL DISCUSSION ... 48

8.2 ERRORS AND UNCERTAINTIES ... 49

8.3 USEFULNESS AND FURTHER APPLICATIONS ... 50

9 CONCLUSIONS ... 52

REFERENCES ... 53

APPENDIX A. LAND COVER CLASSES CORINE ... 61

APPENDIX B. SPATIAL DATA PROTECTED AREAS ... 61

APPENDIX C. GIS OPERATIONS ... 62

APPENDIX D. ACCESS POINTS ... 63

List of figures

FIGURE 1.THE STUDY AREA:STOCKHOLM COUNTY AND ITS MUNICIPALITIES.SPATIAL DATA ©STATISTICS

SWEDEN AND ©ENVIRONMENTAL PROTECTION AGENCY. ... 7

FIGURE 2.GEOGRAPHICAL DIVISIONS FOR STOCKHOLM COUNTY.SPATIAL DATA ©STOCKHOLM COUNTY COUNCIL. ... 18

FIGURE 3.POLYGONS FOR THE GREEN AREA QUALITY:SPACIOUS.SPATIAL DATA ©LANTMÄTERIET. ... 22

FIGURE 4.POLYGONS FOR THE GREEN AREA QUALITY:QUIETNESS.SPATIAL DATA ©COUNTY ADMINISTRATIVE BOARD. ... 23

FIGURE 5.POLYGONS FOR THE GREEN AREA QUALITY:PARKS.SPATIAL DATA ©TRF. ... 24

FIGURE 6.POLYGONS FOR THE GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY.SPATIAL DATA ©ENVIRONMENTAL PROTECTION AGENCY. ... 25

FIGURE 7.INTERSECTION BETWEEN ROAD NETWORK AND A GREEN AREA. ... 26

FIGURE 8.DISTANCE TO THE NEAREST GREEN AREA QUALITY:SPACIOUS, FOR DWELLINGS LOCATED IN THE ARCHIPELAGO AREAS. ... 30

FIGURE 9.DISTANCE TO THE NEAREST GREEN AREA QUALITY:SPACIOUS, FOR DWELLINGS LOCATED IN THE URBAN COUNTRYSIDE AREAS... 31

FIGURE 10.DISTANCE TO THE NEAREST GREEN AREA QUALITY:SPACIOUS, FOR DWELLINGS LOCATED IN THE COUNTRYSIDE AREAS... 32

FIGURE 11.DISTANCE TO THE NEAREST GREEN AREA QUALITY:SPACIOUS, FOR DWELLINGS LOCATED IN THE URBAN AREAS. ... 33

FIGURE 12.DISTANCE TO THE NEAREST GREEN AREA QUALITY:QUIETNESS, FOR DWELLINGS LOCATED IN THE ARCHIPELAGO AREAS. ... 34

FIGURE 13.DISTANCE TO THE NEAREST GREEN AREA QUALITY:QUIETNESS, FOR DWELLINGS LOCATED IN THE URBAN COUNTRYSIDE AREAS. ... 35

FIGURE 14.DISTANCE TO THE NEAREST GREEN AREA QUALITY:QUIETNESS, FOR DWELLINGS LOCATED IN THE COUNTRYSIDE AREAS. ... 36

FIGURE 15.DISTANCE TO THE NEAREST GREEN AREA QUALITY:QUIETNESS, FOR DWELLINGS LOCATED IN THE URBAN AREAS. ... 37

FIGURE 16.DISTANCE TO THE NEAREST GREEN AREA QUALITY:PARKS, FOR DWELLINGS LOCATED IN THE URBAN AREAS. ... 38

FIGURE 17.DISTANCE TO THE NEAREST GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY, FOR DWELLINGS LOCATED IN THE ARCHIPELAGO AREAS. ... 40

FIGURE 18.DISTANCE TO THE NEAREST GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY, FOR DWELLINGS LOCATED IN THE URBAN COUNTRYSIDE AREAS. ... 41

FIGURE 19.DISTANCE TO THE NEAREST GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY, FOR DWELLINGS LOCATED IN THE COUNTRYSIDE AREAS. ... 42

FIGURE 20.DISTANCE TO THE NEAREST GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY, FOR DWELLINGS LOCATED IN THE URBAN AREAS. ... 43

FIGURE 21.SCATTERPLOT DISPLAYING THE DISTRIBUTION OF THE DATA POINTS REPRESENTING PROPERTY VALUE PER AREAL AND DISTANCE TO PARKS IN URBAN AREAS. ... 46

FIGURE 22.SCATTERPLOT DISPLAYING THE DISTRIBUTION OF THE DATA POINTS REPRESENTING PROPERTY VALUE PER AREAL AND DISTANCE TO QUIETNESS IN URBAN COUNTRYSIDE AREAS. ... 47

FIGURE 23.THE PERFORMED GIS OPERATIONS IN ORDER TO GET THE USED GREEN AREA QUALITY POLYGONS. ... 62

FIGURE 24.ACCESS POINTS FOR THE GREEN AREA QUALITY:SPACIOUS. ... 63

FIGURE 25.ACCESS POINTS FOR THE GREEN AREA QUALITY:QUIETNESS. ... 64

FIGURE 26.ACCESS POINTS FOR THE GREEN AREA QUALITY:PARKS. ... 65

List of tables

TABLE 1.ASSUMPTIONS FOR THE ACCESSIBILITY ANALYSIS BASED ON THE LITERATURE REVIEW. ... 17

TABLE 2.LIST OF AVAILABLE DATA THAT WERE USED IN THE SPATIAL ANALYSIS. ... 21

TABLE 3.GUIDELINES FOR THE PEARSON CORRELATION COEFFICIENT. ... 28

TABLE 4.PERCENTAGE OF DWELLINGS FOR EACH ACCESSIBILITY INTERVAL BASED ON THE GEOGRAPHICAL

AREAS FOR THE GREEN AREA QUALITY:SPACIOUS. ... 34

TABLE 5.PERCENTAGE OF DWELLINGS FOR EACH ACCESSIBILITY INTERVAL BASED ON THE GEOGRAPHICAL

AREAS FOR THE GREEN AREA QUALITY:QUIETNESS. ... 38

TABLE 6.PERCENTAGE OF DWELLING WITH ACCESS, BASED ON THE ACCESSIBILITY INTERVALS FOR THE

GREEN AREA QUALITY:PARKS. ... 39

TABLE 7.PERCENTAGE OF DWELLINGS FOR EACH ACCESSIBILITY INTERVAL, BASED ON THE GEOGRAPHICAL

AREAS FOR THE GREEN AREA QUALITY:PROTECTED AREAS FOR BIODIVERSITY. ... 44

TABLE 8.THE PEARSON CORRELATION COEFFICIENT, R, THE LEVEL OF SIGNIFICANCE, THE P-VALUE, THE

MEAN VALUE AND THE MEDIAN VALUE FOR THE TWO VARIABLES.THESE ARE SHOWN FOR THE GREEN

AREA QUALITIES AND THE DIFFERENT GEOGRAPHICAL AREAS. ... 45

TABLE 9.THE EXTRACTED LAND COVER CLASSES FROM THE CORINE DATASET. ... 61

TABLE 10.LIST OF AVAILABLE SPATIAL DATA THAT WERE USED FOR THE PROTECTED AREAS. ... 61

TABLE 11.THE SAMPLE SIZE, COEFFICIENT OF DETERMINATION AND THE LINE OF BEST FIT FOR THE

CORRELATION.THESE ARE SHOWN FOR THE GREEN AREA QUALITIES AND THE DIFFERENT

Nomenclature

Existence value – the willingness to pay for an environmental resource existence without using it on-site (Brännlund and Kriström, 2015).

Experience value – humans expected experience from visiting a specific place (EPA, 2016). Green area – includes the concept of green space which refers to all green elements in the urban area. A green area can also refer to connected green spaces with no general size restrictions. More specifically, green areas are further defined based on the chosen quality.

Green structure – all green areas within the built environment such as parks, gardens, urban nature and green spaces (RTK, 2008).

Grouping of properties – properties which refers the same type of property, the same owner and is located in the same municipality. Grouping of properties have the same ID-number.

Gröntypologi – the concept of “Gröntypologi” describes categories of public green spaces (<0,25 ha), their land characteristics and landowners (ORP, 2010).

Sociotope – a place for human’s experiences and activities (Ståhle, 2006).

1 Introduction

Urbanization is a global trend where people move towards cities, which has accelerated during the past decades. According to UN (2014), 66% of the of the world's population is expected to live in cities by 2050. Around 85% of the Swedish citizens already live in cities and urban areas today, which is expected to increase (Goldenberg et al., 2018). Also, the rapid rate of expanding urban areas are generally contributing to a significant increase of land-uptake. The competition of land resources increases due to interest of both ecosystem services (ES) and construction in cities and urban areas (Mörtberg, et al., 2017).

ES can be defined as the ecosystem’s contribution to human well-being, which include all benefits people obtain from ecosystems (BISE, 2019). More specifically, these can be divided into four categories: provisioning, habitat, regulating and cultural. While provisioning services are the energy or material output from ecosystems (e.g. fresh water and raw material), habitat services function as support and can also be defined as supporting services for other ES (e.g. photosynthesis). Regulating services provide regulating functions (e.g. flood control and climate regulation) and the cultural services provide nonmaterial recreational values from ecosystems to enhance human well-being (e.g. recreational green areas) (BISE, 2019; TEEB, 2019). Humans are both benefiting from and are reliant on these services, which are produced in a natural manner by ecosystems and their associated organisms (ORP, 2019). In urban environments, green areas are potential carrier of such services (EPA, 2019a). Generally, green areas obtain vital ES with high recreational values and important regulating functions. These services are also of high importance in urban environments in order to promote attractive and liveable cities. Moreover, ES are necessary for the extensive urban growth (Chang et al., 2017; Stockholm Resilience Centre, 2019).

Stockholm is considered to be one of the fastest growing regions in Europe and the substantial growth within the region leads to great challenges, where sustainable urban planning is needed (Stockholm City Council 2019; ORP, 2016). In order to achieve sustainable urban development in a greater extent, planning for compact cities has been an increased focus within urban policy and planning. To make compact cities more sustainable, both densification and integration of ES need to be included. In particular, ES need to be integrated within the dense urban areas and the cities need to have access to public green areas and public transport (Mörtberg et al., 2017). Solely densification may contribute to negative effects, such as air pollution, reduced availability of daylight and lack of green areas (OECD, 2012; Steemers, 2003). Additionally, it is the qualitative attributes that arise within a dense city that are coveted by people, not the density itself. An attribute that arises in dense cities is the accessibility to different user values (e.g. urban services and green areas). Accessibility to such values is considered as one of the most important factors for Stockholm attractiveness and character as a city (National Board of Housing, Building and Planning, 2016; Stockholm City Council, 2010). Accessibility is a commonly used concept in urban planning and there is an ambition within the Stockholm region to provide an accessible region by 2050. Furthermore, 95% of the new housing within the region should be located accessible to attractive locations, such as green areas (ORP, 2016). The concept of accessibility is also related to the UN’s Sustainable Development Goals concerning equality justice and access to green public areas (UN, 2019a; UN, 2019b). Several studies have in recent years focused on accessibility to green areas, where high accessibility is desirable (Balfors et al., 2016; Fan et al., 2017; Goldenberg et al., 2018) Moreover, several studies have supported the positive relation between green areas, increased health effects and a good quality of life (Fan et al., 2017; Goldenberg et al., 2018; Grahn & Stigsdotter, 2003). However, according to Stessens et al. (2017) and Ståhle (2010), the proximity is less important than the actual quality possessed by the green area. This is further supported by several sources, which claim that it is the access to the qualities possessed by a green area, and not the space itself, that is attractive (RTK, 2008; Wood et al., 2018).

accessibility to green areas and the correlation to property values already have been conducted for Stockholm, the combination of qualitative green areas have not been extensively researched for the entire region. Hence, a research gap is identified regarding planning for accessibility to qualitative green areas on a regional scale. The scope of this thesis and its relevance is motivated in more details in the following chapters.

1.1 Aim & Objectives

This study aims to measure the accessibility to green area qualities within the Stockholm region, using Geographic Information System (GIS). Furthermore, this study aims to assess whether access to qualitative green areas correlate with the property values within the region. The results intend to be used as a basis for future planning and as a spatial decision support to sustain a good quality of life within the Stockholm region. More specifically, the aim is based on achieving the following objectives:

- Identify relevant green area qualities in the Stockholm region, concerning sustaining a good quality of life.

- Measure the accessibility to green area qualities in the Stockholm region and localize areas with potentially high and low accessibility towards these qualities.

2 Accessibility measures

The concept of accessibility can be defined in several ways and in different dimensions and contexts. Generally, the concept refers to measuring the ability to access a specific destination or activity. El-Geneidy & Levinson (2006) consider it as the ease of reaching a destination. The concept involves two fundamental parts: an origin in relation to a specific destination and the associated route between the origin and the destination (Vickerman, 1974). Depending on the accessibility approach, both spatial and spatial factors can have an impact. Perceived accessibility depends on spatial as well as non-spatial factors, which take cultural and social aspects into consideration regarding how the access is perceived. Physical accessibility, also called geographical accessibility, is a spatial based measure, which also is the most commonly used approach (Haugen, 2012). The different approaches are explained in more detail in the two following subchapters. Moreover, the accessibility approach requires definitions of the accessibility distance, the chosen distance measure, address points and the access points (for the origins and destinations), which are defined in the latter subchapter.

2.1 Perceived accessibility

The perceived accessibility approach has both spatial and non-spatial characteristics, where social and cultural dimensions are included for the non-spatial ones. The approach focuses on an individual perspective, where each person's perception is taken into consideration (Lättman et al., 2018). In the study conducted by Lättman et al. (2016), perceived accessibility is defined as using the transport system in order to live in satisfaction. More specifically, this definition refers to the perceived possibility of each individual to live independently by their choice, in correlation to using the transport system. Furthermore, Ståhle (2010) claims that an integration of the axial line concept in accessibility affects how it is perceived. The axial line concept refers to measuring the distance with the least direction changes, where a shorter distance is perceived when there are few changes. Also, Lättman et al. (2018) state that both a general and an individual perspective is needed when measuring accessibility to avoid social exclusion. In order to measure the perceived accessibility however, detailed information is required, since the approach is based on individuals with different social and cultural characteristics (e.g. age, income level & physical conditions) (Lättman et al., 2016). Regarding access to green areas, the physical distance has been the most commonly used factor in order to measure the accessibility, even when factors for perceived accessibility are included (Higgs et al., 2012; Van Herzele & Wiedemann, 2003; Wang et al., 2015). The study conducted by Wang, et al. (2015) concerns the factors influencing the perceived access to urban parks and they conclude that the physical distance is the most important factor. Furthermore, in Van Herzele & Wiedemanns (2003) study, concerning the provision of accessible green areas, they claim the physical distance factor to affect the utilization of green area the most. Also, Higgs et al. (2012) state that the positive outcomes of green areas, such as human well-being, are predominantly correlated to the physical distance. Accordingly, perceived accessibility is complex to measure and includes a variability of factors, where the physical distance is the one with the most influence. Hence, this study only considered physical accessibility, which also is the most used factor according to previously mentioned literature. For the remainder of the report, only the physical approach is included in the term accessibility.

2.2 Physical accessibility

2.2.1 Accessibility distance

According to EEA (2009), walking as a transportation mode is one of the key components for a city to be accessible and attractive. Also, cities are considered to be healthier when walking is encouraged, especially to open green areas (Spacescape, 2019). Moreover, the walking distance is postulated to be the most appropriate measure regarding accessibility to green areas. The reason for this is that people tend to walk when visiting green areas rather than using other transportation modes, such as cars or public transport (Higgs et al., 2012). Furthermore, several studies use walking as transportation mode and the walking distance as the accessibility distance (de Jong et al., 2011; Goldenberg et al., 2018; Grunewald et al., 2017; Mörtberg et al., 2017).

Concerning people’s willingness to walk, they generally tend to visit their nearest green area (Barbosa et al., 2007; Higgs, et al., 2012). In the study conducted by Van Herzele & Wiedemann (2003), it is claimed that the use of green areas is mostly affected by the proximity from people’s homes. This is further supported by EPA (2017), who states that the distance from people’s home to the closest green area have a great influence on the visitor frequency. However, other factors than the nearest distance may have an influence on the visitor frequency. Green areas that possess qualities, which people add value to, may also be a contributing factor to their visit (Björk et al., 2008; Ståhle, 2010). In the scope of this study, people are assumed to visit their nearest green area, assuming the green area fulfills a certain quality. This study's approach toward values and qualities of green areas is further explained in chapter 4.

Moreover, there are several different distances used in research and literature concerning the maximum distance a person is willing to walk to visit a green area. Also, several cities have guidelines for the maximum recommended distance for each resident between their home and a park or a green area. According to the Swedish standard, suggested by the National Board of Housing, Building and Planning, they recommend a maximum distance of 300 m with no restrictions for the size of the green area (National Board of Housing, Building and Planning, 2007a). In the park programme for the Stockholm municipality, they mention 1000 m as the maximum distance people are willing to walk to visit a green area (Stockholm City Council, 2006). However, Ståhle (2010) argues for a negative correlation between people's actual need and the existing standards and guidelines.

There are several different distances used in research and literature concerning the maximum distance a person is willing to walk to visit a green area, where both 300 m and 1000 m are commonly used. A distance of 300 m between residents and green areas is recommended to increase visitor frequency (Barbosa et al., 2007; Björk et al., 2008; Dadvand et al., 2014; Salat et al., 2014). According to Hörnsten & Fredman (2000), 1000 m is the maximum distance for a Swede to walk to a green area or other recreational areas. In the same study, they also claim that the utilization of a green area has a significant decrease with larger distances. Also, in several Swedish studies and documents 1000 m is used as maximum distance, with reference to Hörnsten & Fredman (2000) study (Helsingsborg municipality, 2013; Spacescape & Evidens, 2011; Ståhle, 2010).

According to EPA (2017), a green area used on a daily basis should be within a distance of 1000 m. However, to obtain a good quality of life they also claim that an even shorter distance of 300 m is needed. Moreover, two recent studies Grunewald et al. (2017) and Grunewald et al. (2019) support that 1000 m is an accurate measure for larger green spaces, while 300 m should be used for nearby nature with no particular size. In a study conducted by Sotoudehnia & Comber (2011), they use three distance intervals for measuring the accessibility distance. The used constraints for distances are: 0-300 m – which is associated with a good access, 0-300-1000 m – which indicates intermediate access and a distance >1000 m – which is associated with poor access. Within the scope for this study, these distance intervals were used.

2.2.2 Network analysis

are frequently used when assessing the accessibility to green areas (Higgs, 2004; Higgs et al., 2012; Nicholls 2001). The euclidean distance method simply calculates the shortest distance by the straight-line distance without considering the possibility for actual travel along the path, while the network analysis method uses a network of paths and roads to find the shortest distance or travel time between the origin and the destination.

In general, people travel along networks. A network system consists of elements, which are interconnected, such as connecting junctions and edges (points and lines). These elements are further representing possible routes one can take from a location to another. Several sources claim that a network analysis provides a more accurate measure of the accessibility to green areas than the euclidean distance method (Gupta et al., 2016; Comber et al., 2008; Sotoudehnia & Comber, 2011). In addition, the network analysis assumes travel patterns along the predefined roads, which according to Nicholls (2001) is a better estimation than traveling along a straight-line for visiting green areas. She further states that the euclidean distance method in general underestimates the actual distance and is not accounting for barriers along the straight-line. In most research and literature today, the GIS-based network analysis is used for the least cost path analysis related to green areas. Based on this, the network analysis approach was used in this study.

2.2.3 Access points

Within the concept of accessibility, address points (origins) and access points (destinations) are two of the most fundamentals parts. An origin refers to a start point with a geographical location for an individual, while a destination refers to an end point with a geographical location that the individual wants to access. Regardless whether the euclidean distance or the network analysis method is used to measure the distance, these points need to be decided. According to Higgs et al. (2012), either individual population data, based on a household-level, or more approximated population data can be used as input for the creation of the origins. Regarding the destination points, there are several different ways to define when a visitor actually reaches a green area. Higgs et al. (2012) points out three approaches to measure the distance to a green area, via a road network, using access points: green space centroid, nearest boundary point and nearest access point. The green space centroid represents the main destination within the green area, which is collated in the center while the nearest boundary point refers to the closest point on the green area. The nearest access point refers to the closest point accessed via the road network, which often is defined as the main entrance.

3 Study area: Stockholm County

The study area for this thesis is the Stockholm County, which includes Stockholm City and its surroundings. Stockholm County is located in the middle part of Sweden on the eastern coast, where it is connected to the Baltic Sea (Figure 1). It consists of 26 municipalities and covers a total land area of 6 500 km2_{, where 630 km}2 _{consists of waterbodies such as streams and lakes. The area covered by}

the Baltic Sea is 9 400 km2_{. Moreover, a population of 2.3 million and 1.05 million dwellings were}

recorded during 2018 within the County, which is expected to increase within the near future (ORP, 2018). However, there are great variations in population density, land use and accessibility to urban services between the inner city of Stockholm and the more sparsely populated areas within the County (Stockholm County Council, 2011).

According to the regional development plans for Stockholm, RUFS 2050, the vision is to develop the most attractive city in Europe by 2050, where cultural and regulating ES are essential (ORP, 2018). In addition, sustainable development has an important function and characterizes the whole regional plan. Accessibility has been mentioned as an important concept in order to obtain sustainable development and an attractive city (Mörtberg et al., 2017; ORP, 2018). In the regional development plans, there is a goal to develop Stockholm towards an accessible region with a good living environment. More specifically, 95% or more of new buildings should be located within the most accessible locations of the region. Also, it is stated in the regional development plans that a region with availability of social services, public transport, parks and green areas is an accessible region, which offers a good living environment for the citizens. An ambition for the region is to have good accessibility to attractive green areas since the access affects the use of them. As previously mentioned however, it is not only the access affecting the use of green areas, since their quality affects the use as well (RTK, 2008; ORP, 2018).

Figure 1. The study area: Stockholm County and its municipalities. Spatial data © Statistics Sweden and © Environmental Protection Agency.

3.1 The compact city

a greater extent. Furthermore, the compact city promotes less uptake of land, which enables the preservation of green areas with important functions. The compact city also offers higher accessibility to urban services and high valued locations, which makes the city more livable and attractive (Mörtberg et al., 2017).

However, the planning approach and development regarding the compact city are important as well. Implementation of a monocentric dense area is expected to result in loss of green areas, since this compact city approach mainly focuses on energy and transport efficiency, while the protection and preserving of green areas are mostly excluded (Mörtberg et al., 2017). In correlation to EU’s spatial development policy, a polycentric planning approach has been implemented in European cities during the last decade (European Commission, 1999). According to ORP (2018), the Stockholm region aims to implement a polycentric planning approach, which includes several regional city cores. This planning approach has been implemented to avoid both the inefficiency with urban sprawl and the disadvantages with a dense monocentric approach. Moreover, the focus on a polycentric dense urban form entails efficient use of resources, improved accessibility, protection of green areas with high values and protection of ES (Mörtberg et al., 2017).

In order to make compact cities more sustainable, both densification and integration of ES need to be included. A polycentric planning approach includes both densification and preservation of valuable green areas that include ES. In particular, ES need to be integrated within the dense urban areas and the cities need to have access to public green areas and public transport (Mörtberg, et al., 2017). However, solely densification may contribute to negative effects, such as air pollution, reduced availability of daylight and lack of green areas (Khoshkar et al., 2018; National Board of Housing, Building and Planning, 2016; OECD, 2012; Steemers, 2003). Furthermore, the densification itself does not contribute to the attractiveness of the city. Instead, the qualities arise when dense areas are developed, with high accessibility to public transport, urban services and attractive places (e.g. green areas) (ORP, 2018).

3.2 Good quality of life

In order for cities to be sustainable, they should contain desirable attributes that are attractive and promote a good quality of life with livable landscapes (Mörtberg et al., 2017). Several sources claim that a good quality of life in cities is highly related with access to green areas and recreational values (Mörtberg et al., 2017; Stessens et al., 2017; Wang et al., 2015). In addition, people living in urban areas perceive access to green areas, and their associated ES, as extra important for their quality of life (Wang et al., 2015). Furthermore, the need for integrating accessibility to attractive places (e.g. green areas) is especially important to sustain a good quality of life in growing regions, such as Stockholm (Ekologigruppen, 2017). A growing region pressures the urban development plans and according to the County Administrative Board Stockholm (2019a), the proximity to green areas increases the attractiveness and the quality of life within the region. Another important factor for urban green areas in order to provide a good quality of life is that they are available for the public, free of charge (Spacescape, 2019).

3.3 Green areas in Stockholm

Generally, Stockholm is a green city region with more than 50 % of the land surface covered by general vegetation and forest. From an international perspective, Stockholm is unique in the sense of green area presence and waterbodies in the proximity to the inner-city core. The high presence of green areas both depends on natural conditions and the historical focus in regional planning during the recent centuries. Green areas and the planting of trees have been implemented in the regional city planning since the 17th _{century, when people realized the health benefits, as well as the function of}

preventing the spread of fires in the urban areas. During the early 19th _{century, green areas were also}

integrated in the urban planning with the purpose of creating reactional values in the proximity to residents. The proximity was considered to be important since it was assumed that people were limited to travel far when visiting a green area (County Administrative Board Stockholm, 2019a). Moreover, the city planning department in Stockholm have traditionally had a strong regulating power, which they also have today. This enables the implementation of more public facilities, such as parks, but also the preservation of green areas within the city. In Stockholm today, the largest proportion of land is owned by the government, which enables a management in line with the regional city plans. It is also common that authorities in Sweden buy large areas of privately-owned land, where they implement public facilities such as parks (Spacescape, 2019).

The region presently contains many green areas with both high nature values and cultural values. These values are of high importance for the region to be considered attractive. Furthermore, the green areas in Stockholm County are significantly correlated to the attractiveness and the identity of the entire region (RTK, 2008). The regional development plans for Stockholm focus a lot on the “Stockholm green wedges”, which are connected green areas that form a green structure from the urban fringe into the center of Stockholm. Accessibility to the green wedges have been the main focus in the regional development plans for Stockholm, since their qualities and connections are considered to be of high importance (Stockholm County Council, 2019a). However, important qualities exist in other green areas within the region as well, which motives the study area in this thesis.

There are several ways of defining what a green area is and what it includes. Statistics Sweden (2015) and ORP (2010) have come up with two definitions which have been used previously for accessibility measures in Stockholm. According to Statistics Sweden (2015), a green area is defined as a public area of connecting green areas with a minimum size of 0.5 ha. However, this definition has a geographical boundary, which only includes green areas in urban areas or within a distance of 3 km from an urban area. ORP (2010) defines a green area as a public green area covered by at least 50 % vegetation, with a minimum size of 0.25 ha, based on their concept of “Gröntypologi”. In this study however, the definition of a green area will vary corresponding to the different qualities described in chapter 4 and in chapter 6.

3.4 Geographical division

Geographical divisions have for a long time been necessary in order to provide statistics over regions and enhance the urban planning (Statistics Sweden, 2005; Stockholm County Council, 2019b). In addition, geographical divisions are interesting in relation to accessibility within the Stockholm region, since there are great differences between different geographical areas. The conditions in the inner-city core and the rural areas may differ a lot. More specifically, urban areas and the inner city of Stockholm have a variety of urban services and more sociocultural experiences than the rural areas. However, the more sparsely populated areas generally have more nature and green areas in their neighborhood (TRF, 2018). According to Soga et al. (2015), people in rural areas use green areas more frequently for recreation than people in urban areas. However, Grahn & Stigsdotter (2003) and Ståhle (2010) states that there is no significant difference in demand for public green areas between different geographical areas.

4 Quality measures for green areas

There are several benefits and health effects related to green areas and there are good reasons for people wanting to have good access to them. They provide several important ecological functions, as well as recreational ones (Cetin, 2015). Furthermore, urban green areas are important from a social perspective, since they can function as meeting spots without costing anything in monetary values (County Administrative Board Stockholm, 2003; County Administrative Board Stockholm, 2019a). The attractiveness of urban green areas is not only dependent on the accessibility in terms of proximity, but also which values that specific piece of green area provides the inhabitants (Ståhle, 2010; Van Herzele and Wiedemann, 2003). The values of green areas can be described as the qualities a green area possesses. The qualities can further consist of the attributes and features a green area has, which attracts people to visit and stay there. Also, some sources claim the proximity itself to be less important than the values of the green area (Stessens, et al., 2017; Ståhle, 2010). It can be considered that it is the accessibility to the qualities and not necessary the space itself that is attractive. Generally, providing green areas with high quality has been challenging when planning for compact cities (Khoshkar et al., 2018). Moreover, the National Board of Housing, Building and Planning (2016) expresses a need to include different values and qualities of green areas, especially in urban regions when approaching a compact city.

It is already established in research that accessibility to green areas is important for human well-being and people’s quality of life. However, the qualities a green area possesses are also important factors concerning the attractiveness and use of green areas to improve neighborhood satisfaction, the overall human well-being and quality of life (Björk et al., 2008; Ekkel & de Vries, 2017; Jansson, 2014; Zhang et al., 2017). In general, people tend to be healthier when living with good access to green areas with qualities and people also consider themselves to be healthier when having good access to these qualitative green areas (Jansson, 2014). This importance of green area qualities has been reflected in research during the recent years, where several studies have had focus on values and qualities provided by green areas. For example, Björk (2008) concludes that people living in the vicinity to green areas with high qualities and recreational values visit the green area more frequently, which is further supported by Kemperman & Timmermans (2006). However, Ekkel & de Vries (2017) claim that a high density of visitors to a green area causes crowdedness, which may be experienced as less relaxing and less beneficial, while others mention the presence of other people to increase the perceived safety and add social values (Giles-Corti et al., 2005). For this thesis, the potential overcrowding of people was not taking into account.

There are several ways of defining what a green area quality is in relation to accessibility. Some sources (Jansson, 2014; Zhang et al., 2017) claim the accessibility (proximity) itself to be a quality while others do not (Ekkel & de Vries, 2017: Van Herzele & Wiedemann, 2003). According to Ekkel & de Vries (2017), the proximity is defined as a quantity, while a quality refers to something physical within the green area (e.g. animal species or diversity in vegetation). Van Herzele & Wiedemann (2003) mentions the proximity to be a precondition of use, which refers to whether one visits a green area or not, while they claim that the quality describes the amount of time people want to stay there. They also mention the availability in terms of open public use as a precondition of use. In this thesis, the proximity and the availability are referred to preconditions of use and not qualities.

In summary, several sources point out that the quality of a green area should be more focused on in the research field concerning accessibility to green areas. In order to include qualities, the characteristics of a green area quality, as well as its importance, need to be defined. The following subchapters explicate the relevance of qualities in green areas and some important concepts, which have been used for determining the green area quality in previous studies and literature.

in subchapter 4.2. In the last subchapter (4.3), the used qualities, as well as their importance, are introduced.

4.1 Quality approach

In recent years, several studies have had focus on values and qualities provided by green areas (Fan et al., 2017; Ståhle, 2010; Van Herzele & Wiedemann, 2003). However, there are different approaches in order to measure the quality of a green area: perceived quality and spatial quality. The perceived quality approach refers to site-specific and survey-based studies, while the spatial quality approach refers to observation-based or expert-based quality assessments (de Jong et al., 2012). Previously conducted studies have attained mixed results whether there is an agreement between spatial and perceived measures of green areas. Zhang et al. (2017) found an alignment between the physical availability, of accessible green areas with usable qualities, and how people perceive the quality of the green areas in their neighborhood, which is correlated to use characteristics. Examples of such use characteristics are mentioned to be: amenities, natural features, accessibility, facilities and maintenance. Also, Zhang et al. (2017) suggest that assessments of qualities of a green area that are spatial and expert-based can be a fair proximation of people’s subjective perceptions. However, according to de Jong, et al., (2012) and Ståhle (2010), the spatial assessed qualities do not necessarily correlate with the actual experience of a specific location. Also, some studies claim that green area qualities are individual based and require site-specific knowledge and extensive surveys in terms of citizen dialogues (Ståhle, 2010; Wang et al., 2015). However, such surveys are complex, require a lot of resources and are heavily time consuming (EPA, 2016). Hence, they are therefore outside the scope of this study. In this thesis, a spatial quality approach was used, where the quality parameters are based on people’s experiences of green areas.

4.2 Experience values

Regarding qualities and green areas, there has been intensive research on the concept of experience values in Sweden during the last two decades (National Board of Housing, Building and Planning, 2007c). More specifically, experience values can be defined as humans expected experience from a visit in nature (EPA, 2016). There is generally sufficient knowledge concerning ecological and cultural environments related to green areas, while the overall knowledge about their social values are lacking, especially on a regional level. In order to include and map social values in green areas in the regional planning, experience values are commonly used (County Administrative Board Stockholm, 2003). Moreover, experience values and the nature in general contribute to increased quality of life and human well-being (RTK, 2004a). Patrik Grahns research concerning urban green areas and parks presents eight characteristics that Swedes in general wish to experience when visiting a green area. These characteristics are: “Spacious”, “Wilderness”, “Rich variety of species”, “Play inspiring”, “Cultural”, “Peacefulness”, “Square” and “Festive” (Berggren-Bärring & Grahn, 1995). “Spacious” refers to a large place where feelings like privacy and freedom can take place and “Wilderness” refers to a quiet, spacious areas with untouched nature. An area which is mentioned as “Rich variety of species” has a naturalistic environment with a high biodiversity, while “Play inspiring” refers to an environment suitable for kids to play, preferably with nature characteristics. A “Cultural” area includes heritage or something with a cultural context and “Peacefulness” refers to a spacious area that is quiet, without noise pollution or lively activities. “Square” refers to an active place close to the city and “Festive” refers to a cultural place, which functions as a social meeting spot that is lively (Berggren-Bärring & Grahn, 1995).

2004a). There are mixed opinions regarding the ranking of the importance for different experience values. However, several Swedish studies promote “Peacefulness”, “Rich variety of species”, “Festive” and “Spacious” as values of high importance that are strongly related to increased well-being and increased quality of life (Grahn & Stigsdotter, 2009; National Board of Housing, Building and Planning, 2007b; Public Health Institute, 2009). These values are further used as a basis for the used qualities in this study, although they are modified to some extent. In the proceeding parts in this study, “Peacefulness” is evaluated as Quietness, “Festive” as Parks, “Rich variety of species” as Protected areas for biodiversity and “Spacious” as Spacious. Moreover, the choice of these qualities was partly based on the availability of spatial data. Hence, the importance of other qualities is not neglected, although they are outside the scope of this thesis.

4.3 Green area qualities

In the following four subchapters, 4.3.1, 4.3.2, 4.3.3 and 4.3.4, the used qualities and their importance are introduced.

4.3.1 Spacious

The shape, function and size of a green area is considered to have great influence on the sense of spaciousness and the possibility to utilize a green area is considered to increase with its size (EPA, 2016; RTK, 2004a; National Board of Housing, Building and Planning, 2007a). According to Berggren-Bärring & Grahn (1995), the sense of spaciousness is the most prioritized character by all experience values, which is further supported by Fan et al. (2017). They also mention the size, in terms of area, to be one of the most important indicators for green area quality in urban landscapes. According to Ekkel & de Vries (2017), larger green areas provide restorative functions and have in general better conditions for people to perform physical activates than smaller green areas. Also, they maycontribute to a more stress relief environment. However, there are mixed opinions concerning which dimension a large green area is referring to. According to Kristianstad Municipality (2016) and Public Health Institute (2009), a large green area should be between 10-50 ha, while Helsingborg Municipality (2013) mentions nature areas greater than 15 ha to be large. Van Herzele & Wiedemann (2003), used 10 ha as a minimum size when measuring accessibility to green areas qualities, while a study conducted for the Stockholm region claims green areas greater than 6 ha to result in a sense of space (RTK, 2004a).

In this study, the choice of size for a green area to be considered spacious is decided based on the used spatial data, which is further presented in chapter 6. Furthermore, in larger green areas, qualities such as peacefulness, quietness and rich variety of species can be more easily reached. The size of a green area affects to which extent a green area contributes to ES and favorable habitats for animal and plant species, which in turn contributes to even more quality (County Administrative Board Stockholm, 2019a).

4.3.2 Quietness

noise levels below 40 dBA to be desirable and Helsingborg municipality (2013) supports that the noise level should not exceed 45 dBA in a recreational green area. The choice of appropriate noise level for this study is further decided based on the used spatial data, which is presented in chapter 6.

4.3.3 Parks

Generally, parks contain features, facilities and provide several opportunities for activities that are of high importance and provide quality to people. Park facilities (e.g. benches & playgrounds) can be a crucial aspect for people to stay in the green areas (Reyes et al., 2014; Van Herzele & Wiedemann, 2003). Moreover, parks contribute to urban quality and function as meeting spots, which is beneficial to human health in a social context (Giles-Corti et al., 2005; Public Health Institute, 2009). According to Van Herzele & Wiedemann (2003), parks are especially important in the inner city, since they may be the only contact with green areas in some people’s everyday life. Furthermore, maintenance and safety are important aspects that provide values for people and are generally related to parks in a green area context (National Board of Housing, Building and Planning, 2007b). Parks can also be associated with several experience values. However, the opportunities that entail for social interaction is the main motivation for this quality.

4.3.4 Protected areas for biodiversity

5 Property values related to green areas

Within the real estate industry, it is generally expressed that “location, location, location” is what creates value for a property (WSP, 2017). Location factors, which influence property values, contribute to the attractiveness of the city and are mentioned as city qualities by Spacescape & Evidens (2016). In general, cities with good accessibility to different location factors with urban values are also highly valued in monetary terms. Moreover, there are several urban values affecting the property value (e.g. city center, public transport, water and green areas) (Li, 2018; Spacescape & Evidens, 2011; Spacescape & Evidens, 2016). In the study conducted by Spacescape & Evidens (2011) regarding valuation of city qualities in Stockholm, accessibility to parks and green areas are mentioned as contributors to higher property values in monetary terms. They further claim that property values indicate whether a quality is demanded or not, which might give an indication of the attractiveness of a green area.

6 Methodology

The methodology for the master thesis project was divided into four main parts. The first part included collection of two types of data, where the first type was collected through a literature review. This was done in order to identify and choose relevant green area qualities. These qualities were determined with input from TRF and their expert knowledge within regional planning. Also, this was done to ensure the relevance for the qualities in a regional context. Data related to the accessibility concept, planning approaches for the study area and attractors for property values related to green areas were collected as well. The results of the literature review are presented in more detail in the previous chapters of this report. Also, this first part included determination and collection of spatial data, and the modification of it. In the second part of the methodology, the accessibility was measured in GIS by assessing the walking distance, using the pedestrian road network between dwellings and the access points for qualitative green areas. The third part involved a statistical analysis to examine whether there were any correlations between the property value and the accessibility distance to green areas for different geographical divisions. The last part was to analyze and visualize the outputs from the accessibility analysis and the statistics. In the following subchapters, each part is described in more detail.

6.1 Literature review

The literature review was an essential part for better understanding and for relevance and motivation of this thesis. The results of the literature review are presented in the previous chapters of this report. In short, it included:

- Understand the accessibility concept with its different approaches, perceived accessibility & physical accessibility. Also, to understand and choose measurement methods with their associated parameters.

- Understand the importance for accessibility to green areas concerning quality of life and human well-being.

- Study area determination and understanding for Stockholm’s regional planning approaches and its relationship to Stockholm’s green areas.

- Understand the variety of conditions and the importance of urban services and qualities for different geographical areas.

- Identification and review of qualities that have been frequently used in previous studies related to green areas, especially related to a good quality of life and experience values. - Choice of four green area qualities: Spacious, Quietness, Parks and Protected areas for

biodiversity.

- Understand the attractors for property values and its relationship to green areas, especially for different geographical areas.

6.2 GIS-based analysis

Table 1. Assumptions for the accessibility analysis based on the literature review.

Accessibility refers to the physical accessibility. Thus, it refers to the distance. The distance refers to the walking distance.

The maximum walking distance was based on people’s willingness to walk to the nearest green area in the vicinity of their homes. The following distance interval were used: 0 – 300 m (refers to very good access), 300 – 1000 m (refers to good access) and > 1000 m (refers to poor access.) The accessibility measure refers to the least cost path via the road network (network analysis) The intersections between the road network and the green areas were used as access points. Spatial measured qualities are assumed to correspond to the actual qualities of green areas. Spacious, Quietness, Parks and Protected areas for biodiversity were chosen as the green area qualities.

6.2.1 Determination of geographical divisions

Figure 2. Geographical divisions for Stockholm County. Spatial data © Stockholm County Council.

6.2.2 Spatial data collection

which refers to where the road network coincides with the polygons for each green area quality. The creation of these access points is further explained in the next subchapter. Moreover, the start points were created from a dataset displaying dwellings per property. The dwellings that belong to the same property were merged into multipart polygons instead of separate polygons. This was done in order for them to keep the same ID-number in the attribute table. This dataset was further based on the FastPak-point dataset provided by Statistics Sweden (2019) and the Property Map provided by Lantmäteriet (2017). The FastPak-point dataset was crucial, since it provided each property location, contained census data and the value of each property and its areal. Furthermore, the used road network consisted of walkable roads, where traffic barriers such as highways were excluded. The dataset was created based on the Property map provided by Lantmäteriet (2017). These roads are further based on NVDB (national road database of Sweden), which is supplemented by Lantmäteriet with aerial photo interpretation and municipal information regarding walking paths, hiking trails and floodlit trails.

As briefly mentioned in the previous paragraph, polygons for each green area quality coincides with the road network. In order to display or create these polygons, spatial input data were needed. For the first quality: Spacious, green areas were extracted from CORINE Land Cover (CLC) 2018 dataset (Lantmäteriet, 2018). CLC is coordinated by the European Environment Agency (EEA) and is a European programme, which aims to offer consistent information regarding land cover and its changes within Europe. The CLC dataset is further based on photointerpretation, on a national level, of satellite images for each cooperating country (Copernicus, 2018). The dataset consists of 44 land cover classes, which are not all used for Stockholm. The extracted land cover classes are presented in more detail in Appendix A. Furthermore, CLC uses 25 ha for areal phenomena as the minimum mapping unit (MMU). Therefore, areas that are smaller than 25 ha are generalized to the largest or most similar neighbor class (Copernicus, 2018).

As mentioned in the literature review regarding defining a proper size for a large green area, the Public Health Institute refers a large green area to be within 10-50 ha (Public Health Institute, 2009), which is suitable for the CLC dataset with an MMU of 25 ha. Regarding the resolution of the data, some dwellings were located within the green areas and were therefore receiving a false distance to the access point. However, these dwellings were manually assigned a distance of 0 m afterwards. The same procedure was done for the other qualities if the dwelling points overlap the green area quality polygons.

For the quality: Quietness, polygons were created based on LST Noise study 2016, where the dataset provides an overall picture of the noise intensity for all counties in Sweden. The dataset was further provided by the County Administrative Board (2016). The noise intensity is presented by an integer raster with a spatial resolution of 25x25 m, which represent the intensity of noise conditions. The noise intensity is further divided into five intervals: [0.00 – 0.10], [0.10 – 0.25], [0.25 – 0.50], [0.50 – 0.75], [0.75 – 1.00], which only includes areas greater than 1 ha. Moreover, these intervals are representing the noise intensity in a relative noise scale instead of using the standard measure, decibel (dB). The relative scale aims to be more descriptive for visualization. The first interval: [0.00 – 0.10] represent a quiet area with noise levels which corresponds to noises provided in nature such as bird song, which were further used for creation of polygons representing quiet areas. The second interval: [0.10 – 0.25] refers to the noise from a normal conversation while the third interval: [0.25 – 0.50] refers to a noise level similar to a sawmill or a foundry. The fourth interval: [0.50 – 0.75] refers to noise from a country road, a power plant or general industry noise and the firth interval: [0.75 – 1.00] refers to a very loud noise, which can come from heavy traffic.