Multiscale GIS based Analysis of Urban Green Spaces (UGS)

Accessibility: Case Study of Sisak (Croatia)

Silvija Šiljeg

, Rina Milošević

and Edita Vilić

University of Zadar, Department of Geography, Trg kneza Višeslava 9, 23000 Zadar, Croatia

Keywords: Urban Green Spaces (UGS), ANGst, Accessibility, Sisak.

Abstract: By the year 2050, two-thirds of the world population will live in urban areas. Therefore the quality of life in

cities has become the object of numerous research papers. One of the basic elements of satisfying the

quality of life is the accessibility of urban green spaces (UGS). In this paper accessibility of UGS for the

city of Sisak (Croatia) has been analysed. Based on the fact that Sisak is traditionally an industrial type of

town, the optimal distribution of UGS has the potential to ease the negative effects of urbanization and

industrialization. Accessibility analysis was performed according to guidelines of ANGst (Accessible

Natural Greenspace Standard) methodology. The research is conducted at a multiscale level (based on GIS

analysis). The primary spatial database of UGS for the city of Sisak was created using the supervised

classification method of Sentinel-2A images and vectorization of high-resolution digital orthophoto (DOP).

Accessibility zones were generated using the Network Analyst extension. Results show that the basic

ANGst standard of UGS accessibility is not satisfactory throughout the city. To get more detailed results we

suggest using the very high-resolution satellite imagery or aerial photogrammetry.

https://orcid.org/0000-0002-5473-2579

https://orcid.org/0000-0002-2302-7738

1 INTRODUCTION

Urbanization can be considered as a rapid converter

of natural environments to impervious surfaces

(Frick and Tervooren, 2019). City population

constantly grows and according to the United

Nations’ population projections, by the year 2050,

approximately two-thirds of the world population

will live in urban areas (UN, 2018).

As a consequence of urbanization, the quality of

life in cities has become the object of research to

numerous authors (Franklin, 2001, Amao, 2012,

Balestra et al., 2013, Pacione, 1986).

Accessibility of urban green spaces (UGS) is an

integral element of satisfying the quality of life

(Šiljeg et al., 2018). Accessibility can be defined as

“relative ease” of approach to a specific location, in

this case, UGS (Mak et al., 2017). It usually

represents the non-linear distance traveled without

using the means of transportation. According to the

European Urban Green Belt project, urban green

space is any public or private open property covered

with vegetation, directly or indirectly accessible to

users (Šiljeg et al., 2018).

Access to green areas provides the potential to

reduce health inequalities, improve well-being, and

aid in the treatment of mental illness (WHO, 2019).

WHO (World Health Organization) quotes that

physical inactivity, linked to the lack of access to

recreational zones accounts for 3,3% of global

deaths. From the environmental aspect, besides

producing oxygen, green spaces are a sufficient filter

for air pollution and have an impact on moderating

temperatures.

At the Rio+20 conference on sustainable

development (2012) is concluded that the square

meter per capita of urban green space is one of the

health indicators of sustainable cities. Studies

conducted in several cities confirmed that green

spaces are more accessible to high-income residents

(Hoffimann et al. 2017). Therefore, the lack of

green spaces in cities can be an indicator of

inequality and marginalization among the

population.

240

Šiljeg, S., Miloševi

c, R. and Vili

c, E.

Multiscale GIS based Analysis of Urban Green Spaces (UGS) Accessibility: Case Study of Sisak (Croatia).

DOI: 10.5220/0009470802400245

In Proceedings of the 6th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2020), pages 240-245

ISBN: 978-989-758-425-1

All of the abovementioned contribute that

studies about UGS are becoming one of the key

elements in urban planning. Green volume data can

be used as an input for modeling (urban climate,

water balance) as well as for evaluation processes

(Frick and Tervooren, 2019).

In this research, the case study is the city of

Sisak. Sisak is a traditionally industrial city and

there are two industrial zones within the boundaries

of the settlement. The optimal spatial distribution of

UGS in the city of Sisak is important to keep the

environmental balance of the urban landscape and to

mitigate negative industrial effects.

The aim of this research is to analyse the

coverage and accessibility of green spaces according

to European standards (ANGst, WHO) at the

multiscale level. The research questions are:

 Are the green spaces equally accessible to all

residents?

 Are there enough green spaces unto the

population?

2 STUDY AREA

Sisak (32 km²) is the administrative center of the

Sisak-Moslavina county (Fig. 1.). Development of

Sisak as an industrial center started in the second

part of 20 century and it became one of the most

developed industrial cities in Croatia (Slukan - Altić,

2003). At the latest census (2011), the total

population of Sisak was 33,322 (DZS). The city of

Sisak consists of 28 statistical circles – the smallest

official statistical unit in Croatia. They were created

in 1959 and have been revised for each population

census. They represent a permanent network of

spatial units, which covers the entire mainland area

of Croatia (Šiljeg et al., 2018).

Figure 1: Study area - City of Sisak.

3 MATERIALS AND METHODS

The database of UGS was created based on the

analysis of multispectral satellite images (Earth

Explorer, 2019) and high-resolution (0,5 meters)

digital orthophoto (DOP). Analysis has been

conducted in Erdas Imagine (2018) and ESRI Arc

GIS 10.1.

3.1 Satellite Remote Sensing Data

The land cover model of Sisak is generated in Erdas

Imagine by supervised classification using the

Maximum Likelihood algorithm (pixel-based

approach). Classes of the UGS were initially derived

from Sentinel 2A satellite imageries analyses. They

were modified and corrected using the data acquired

by the manual vectorization method of high-

resolution DOP. The spatial resolution of Sentinel

2A varies from 10m to 60m depending on the

spectral band (Drusch et al., 2012).

Classification is made according to the

methodology of the Urban Green Belt project (URL

urban green belts (UB) project wpt 1 activity) and

following the ANGst standards.

3.2 ANGst Methodology and

Accessibility

Accessibility was analysed following the ANGst

methodology (English Nature, 2003), which

recommends that everyone should have access to the

natural greenspace. The main criteria followed in

this paper were:

 Minimum 2 hectares in size, no more than 300

meters (5 minutes walk) from home

Accessibility model was generated based on

UGS, which are publicly accessible (with or without

entry fee) and bigger than 2 hectares. That eight

classes of UGS encompass; urban forest, public

green spaces (parks, promenades, lawns, city

gardens, greenery by the buildings and institutions,

playgrounds), green spaces by the river, abandoned

areas, sport green surfaces, green spaces by the train

rails and cemetery.

UGS accessibility zones were determinate using

the Network Analyst extension, specifically the New

Service area tool in which the following parameters

were used;

a) Default breaks 5 minutes

b) Restriction – disabled

c) Polygon type – Detailed, trim polygons 50

meters

Multiscale GIS based Analysis of Urban Green Spaces (UGS) Accessibility: Case Study of Sisak (Croatia)

241

d) Multiple facilities options – overlapping

e) Overlap type – discs

3.3 Data Quality Assessment

Topology correction was performed on the traffic

data following the rules: must not overlap, must not

intersect, must not self-overlap, must not self-

intersect, must be a single part. As a cost attributes

walking time (minutes) was calculated. Assumed

walking speed is 5 km/hour. After the topology

correction Network dataset has been created.

The classification was based on satellite imagery

with a spatial resolution of 10 meters. Accuracy of

the classification is evaluated based on the field

validation. On the public green spaces bigger than 2

hectares, 30 control points were selected.

3.4 Indicators of Green Space

Accessibility

Indicators of green space accessibility, in general,

takes into account the distribution of the population

(statistical circle, settlement) in terms of their

proximity to green space (WHO, 2019). The most

widely used indicator to assess green spaces is their

total area with respect to the total population

(m²/inhabitant) (Taylor et al., 2011; Van Herzele and

Wiedemann, 2003; Caspersen et al., 2006; Kabisch

and Haase, 2013; ISO, 2014). But this indicator is

too general and does not give us information about

the actual distribution of UGS and population

throughout the city. That is why this research is

conducted at three levels; macro, meso and micro:

• At the macro level of the research, UGS

accessibility is expressed as a percentage of the total

settlement area, which has accessible UGS.

• At the meso level of the research, UGS

accessibility is expressed as a percentage of the total

statistical circle area that has accessible UGS.

• At the micro-level of the research, UGS

accessibility is expressed as a percentage of the total

housing units within the UGS service area.

In this paper, the presented results are only for

representative statistical circles, those with the

highest percentage and with the lowest percentage.

Total housing units data are the result of fusing the

data downloaded from Geofabrik and data from the

State Geodetic Administration. Missing units were

manually vectorized using the high-resolution DOP.

The results of these indicators are relevant for

comparison with other European cities (Zadar,

Leicester, Scheffiled).

4 RESULTS AND DISCUSSION

4.1 UGS Database of Sisak City

In the city of Sisak 11 distinctive classes (public and

private) of UGS were identified (Fig. 2). Excluding

the private green spaces, UGS encompasses slightly

less than 31%. Forest is the most widespread class of

UGS (bigger than 2h) with 12% in the urban area.

The second is public green spaces with 9%.

Figure 2: UGS classes (private and public) and other

classes in the city of Sisak.

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

242

4.2 Macro Level

Within 5 minutes UGS is accessible to nearly 30%

of the city area. On average there is 297m² of UGS

per capita. The optimum value suggested by the

WHO is 10 to 15 m² of UGS per capita, while the

minimum is 9 m² (Karayannis, 2014, Šiljeg et al.,

2018). Sisak with 297 UGS per m² is far beyond the

optimal distribution.

On the other hand, ANGst states that the entire

population of the city should have access to UGS.

Considering that, Sisak does not meet the standard

(Fig. 3). If we compare UGS per capita in other

towns (Zadar, Leicester), Sisak has much better

results. Each resident in the settlement of Zadar has

around 114 m² per capita of green space (Šiljeg et

al., 2018). Leicester has 30 m² per capita of UGS

(Comber et al 2008).

Figure 3: UGS accessibility map.

4.3 Meso Level

The results at meso level highlight the diversity

among statistical circles. This was expected and

corresponds with the results of other authors.

Circles located near the city centrum have

mostly more percentage of UGS accessible within

five minutes than those closer to the outskirts of

town. In the city centrum, there are a lot of green

areas intended for recreation and rest. On the

outskirt of the town, green areas are predominantly

transferred into private gardens and agricultural land

which are excluded from the analysis. Three out of

28 statistical circles do not have access to UGS.

(Fig. 4). Following the WHO suggestion (10 to 15

m² per capita), all statistical circle meets the

standard.

Figure 4: Percentage of statistical circle areas with

accessible UGS (bigger than 2 ha) within 5 minutes.

4.4 Micro Level

The result at the micro-level has shown the unequal

distribution of access to UGS. UGS is accessible

within 5 minutes to 79,9% of the housing area.

There are three statistical circles with 100%

accessibility to UGS. These circles are closer to the

city center. The six from 28 circles (20%) don’t have

access to UGS but there is not a lot of residential

buildings in these circles. For comparison, in the

Leicester city, UGS was accessible to 10,3% (Kuta

et al., 2014) of the population, in the Sheffield

36,5% (Barbosa et al., 2007) and in Zadar 38,9%

(Šiljeg et al., 2018). Therefore, Sisak with 79,9% of

housing units with accessible UGS is above

average.

Multiscale GIS based Analysis of Urban Green Spaces (UGS) Accessibility: Case Study of Sisak (Croatia)

243

Figure 5: Percentage of housing units with accessible UGS

(bigger than 2 ha) within 5 minutes.

4.5 Data Accuracy

Data quality assessment was performed by field

validation on 30 control points. Data accuracy is

96,66%. Maximal accuracy was expected because

the object of research was areas bigger than 2

hectares.

Figure 6: Field validation of data.

5 CONCLUSION

Multiscale GIS analysis of UGS accessibility

reflected differences among results depending on the

level of research. At the macro-level (297 m² per

capita) Sisak meets the WHO standards of UGS

accessibility. On the other hand, only 30% of the

urban area meets the ANGst standard. So answers to

research questions are; there are enough m² UGS per

capita, but they are not equally accessible to all

residents. In comparison with other cities (Zadar,

Sheffield, Leicester) in which the same methodology

has been applied, Sisak has average results.

At the meso level, there are differences among

statistical areas, depending on their location in the

city. Few statistical circles are completely (100%)

within the UGS access area. However, these

statistical units are not only encompassing the

housing area but are also referring to unpopulated

areas.

Therefore, analysis of the micro-level of research

is performed. It takes into account the spatial

distribution of the housing units within statistical

circles. At the Micro level, 79,9% of the population

have five-minute access to UGS bigger than 2

hectares. The result is above average in comparison

with other towns. Possibly, there are plenty of UGS

in the city center because of industrial zones, to

make a balance in the ecology system. Still,

considering UGS benefits, every resident should

have standardized access to green spaces. As long as

some important measures don’t get applied to certain

city areas, residents may feel marginalized or

discriminated. In Sisak, a big part of UGS is

abandoned and is not reaching its full potential.

Taking care of these areas may bring equality to

certain areas in the city. Using high-resolution

multispectral images is suggested to generate a

better quality model of UGS.

ACKNOWLEDGEMENTS

This work has been supported in part by Croatian

Science Foundation under the project UIP-2017-05-

2694 and National Park „Krka“.

REFERENCES

Amao, L. F. (2012). Housing Quality in Informal

Settlements and Urban Upgrading in Ibadan, Nigeria,

Developing Country Studies 2 (10), 68-80.

Balestra, C., Sultan, J. (2013). Home Sweet Home: The

Determinants of Residential Satisfaction and its

Relation with Well-being, OECD Statistics Working

Papers, 5, 1-42.

Barbosa O., Tratalos J. A., Armsworth P. R., Davies R. G.,

Fuller R. A., Johnson, P., K. J. Gaston (2007). Who

benefits from access to green space? A case study

from Sheffield, UK, Landscape and Urban Planning

83 (2):187-195.

Caspersen, O.H., Konijnendijk, C.C., Olafsson, A.S.

(2006). Green space planning and land use: an

assessment of urban regional and green structure

GISTAM 2020 - 6th International Conference on Geographical Information Systems Theory, Applications and Management

244

planning in Greater Copenhagen. Geogr. Tidsskr.-

Dan. J. Geogr. 106 (2), 7–20.

Comber, A., Brunsdon, C., E. Green (2008). Using a GIS-

based network analysis to determine urban greenspace

accessibility for different ethnic and religious groups,

Landscape and Urban Planning 86 (1):103-114.

De la Barrera, F., Reyes-Paecke, S., and Banzhaf, E.

(2016). Indicators for green spaces in contrasting

urban settings. Ecological indicators, 62, 212-219.

Drusch, M., Del Bello, U., Carlier, S., Colin, O.,

Fernandez, V., Gascon, F. and Meygret, A. (2012).

Sentinel-2: ESA's optical high-resolution mission for

GMES operational services. Remote sensing of

Environment, 120, 25-36.

Državna geodetska uprava 2017. Digitalna ortofoto

karta – listovi naselja Sisak, Državna geodetska

uprava, Zagreb., retrived from (URL 7

https://geoportal.dgu.hr/) Accessed 05 January 2020.

Državni zavod za statisku (2011). Population data, retrived

from (URL 1 www.dzs.hr ) Accessed at 04 January

2020

Earth Explorer (2019). Satellite images, retrived from:

(URL 6 https://earthexplorer.usgs.gov/)

English Nature (2003). Providing accessible natural

greenspaces in towns and cities: a practical guide to

assessing the resources and implementing local

standards for provision, UK.

Environmental health indicators for the Rio+20 UN

Conference on Sustainable Development (2012).

Background paper for WHO/NIEHS Expert

consultation, Geneva, National Institute of

Environmental Health Sciences, 17-18 May.

Franklin, B. (2001). Discourses of Design: Perspectives on

the Meaning of Housing Quality and Good Housing

Design, Housing, Theory and Society, 18 (1-2), 79-92.

Frick, A., and Tervooren, S. (2019). A framework for the

long-term monitoring of urban green volume based on

multi-temporal and multi-sensoral remote sensing

data. Journal of Geovisualization and Spatial

Analysis, 3(1), 6.

Geofabrik (2019). retrieved from (URL 2 http://www.

geofabrik.de/) Accessed at 02 January 2020

Hoffimann, E., Barros, H., Ribeiro, A., I. (2017).

''Socioeconomic inequalities in Green Space quality

and accessibility – evidence from a Southern European

City'', Int J Environ Res Public Health, 14(8): 916.

International Standards Organization (ISO), (2014). ISO

37120: Sustainable Development of Communities –

Indicators for City Services and Quality of Life.

Kabisch, N., Haase, D., (2013). Green spaces of European

cities revisited for 1990–2006. Landsc. Urban Plan.

110, 113–122.

Karayannis, G., (2014). Dissecting ISO 37120: Why shady

planning is good for smart cities, SmartCitiesCouncil,

retrieved from ( URL 3 http://smartcitiescouncil.com/

article/dissecting-iso-37120-why-shady-planning-

good-smart-cities), Accessed at 30 March 2019.

Kuta, A. A., Odumosu, J. O., Ajayi, O. G., Zitta, N.,

Samail-Ija, H. A., E.A. Adesina (2014). Using a GIS-

Based Network Analysis to Determine Urban

Greenspace Accessibility for Different Socio-

Economic Groups, Specifically Related to Deprivation

in Leicester, UK, Civil and Environmental Research 6

(9): 12-20.

Mak, C., Scholz, M., P. James (2017): Sustainable

drainage system site assessment method using urban

ecosystem services, Urban Ecosystems 20 (2): 293-

307.

Pacione, M. (1986). Quality of life in Glasgow: an applied

geographical analysis, Environment and Planning A:

Economy and Space, 18 (11), 1499-1520.

Selmi, W., Weber, C., Riviere, E., Blond, N., Mehdi, L.,

Nowak, D. (2016). ''Air pollution removal by trees in

public green spaces in Strasbourg, France'', Urban

Forestry & Urban Greening, 17 (2016) 192–201.

Šiljeg, S., Marić I., NIkolić G., Šiljeg A. (2018): “Analiza

dostupnosti urbanih zelenih površina u naselju Zadar”,

Šumarski list, Vol. 142 No.9-10, 2018.

Slukan-Altić M. (2003.). Povijesni atlas gradova,

II.svezak; Sisak, Zagreb.

Taylor, B., Fernando, P., Bauman, A., Williamson, A.,

Craig, J., Redman, S. (2011). Measuring the quality of

public open space using Google Earth. Am. J. Prev.

Med. 40 (2), 105–112.

United Nations (2018). World population to reach 9.7

billion by 2050 - UN, retrieved from (URL 4

https://www.dw.com/en/world-population-to-reach-

97-billion-by-2050-un/a-49241728) Accessed at 02

January 2020.

Van Herzele, A., Wiedemann, T. (2003). A monitoring

tool for the provision of accessible and attractive urban

green spaces. Landsc. Urban Plan. 63, 109–126.

World Health Organization (2020). Urban green spaces,

retrieved from (URL 5 who.int/sustainable-

development/cities/health-risk/urban-green-space/en/)

Accessed at 15 January 2020.

Zadar Nova, (2016). Urban Green Belts (UGB) project

WPT 1 Activity 1.1, Common methodology for local

assessment and analysis of urban green space (UGS)

(http://www.zadra.hr/wpcontent/uploads/2016/10/Met

odologija.pdf), Accessed 25 March 2019.)

Multiscale GIS based Analysis of Urban Green Spaces (UGS) Accessibility: Case Study of Sisak (Croatia)

245