Expansion of the built-up areas in Finnish city regions – The approach of travel-related urban zones

Applied Geography 101 (2018) 1–13

Contents lists available at ScienceDirect

Applied Geography journal homepage: www.elsevier.com/locate/apgeog

Expansion of the built-up areas in Finnish city regions – The approach of travel-related urban zones

T

Maija Tiitu Finnish Environment Institute SYKE, PL 140, FI-00251, Helsinki, Finland

A R T I C LE I N FO

A B S T R A C T

Keywords: Land cover changes Urbanisation Urban zones

Finnish city regions have expanded rapidly due to urbanisation and urban sprawl, most recently in the 2000s. In this study, an approach of travel-related urban zones is introduced to enhance the interpretation of urban expansion in relation to previous land cover and urban form. The urban expansion of 34 Finnish city regions was studied by combining national CORINE Land Cover (CLC) data with socio-economic data using GIS. The regions were analysed by dividing them into zones: pedestrian, public transport, and car zones. The approach is applied from the theory of three urban fabrics, where the fabrics have distinct spatial characteristics and have been shaped by the evolution of transport systems. The aim was to determine which land cover classes were the most exposed to urban expansion during 2000–2012, and how large a share of the new residential floor space was located in previously developed areas in each zone. All the regions expanded the most into former forest areas in the fringe of the city, i.e., in car zones. On average, 77% of the area of urban expansion was previously classified as green spaces suitable for recreation. In the majority of the regions, the largest share of new residential floor space was concentrated in car-oriented locations that were not previously developed. The built-up areas grew even within regions with a declining population, exceeding the population growth altogether by a factor of 1.6. The study supports the previous findings on the sprawl-type urban expansion of Finnish cities, and implies insufficient steering of the urban development with negative environmental impacts in the 2000s. However, for the largest cities, the analyses also show signs of shifting towards infill development policies in public transport zones.

1. Introduction Urbanisation is one of the megatrends of this century world-wide, which implies that the single most important factor changing the urban land cover is the expansion of the built-up areas. It is often linked to urban sprawl development near growing cities (Luck & Wu, 2002; Van Eetvelde & Antrop, 2004). Urbanisation is also accelerating in the rather recently urbanised Finland, where approximately 70% of the population lived in the 34 largest city regions in 2014. The share has risen by 10 percentage points since the year 1990 (YKR, 2018). Urban sprawl has many negative environmental impacts, such as higher energy consumption, greater air pollution, reduced regional open space, loss of farmland, reduced diversity of species, and ecosystem fragmentation (Johnson, 2001). However, it is important that green spaces suitable for recreation are preserved in city regions, since creation of large parks, green belts, and local recreation areas improves the mental and physical well-being of city dwellers (von Hertzen, Hanski, & Haahtela, 2011). According to Colding et al. (2003), urban green spaces and recreation areas are also important habitats for urban biodiversity because nature

conservation areas often cover only a small area in cities. Finnish city regions have undergone a vast expansion during the last decades. In addition, the dispersion of urban areas and the land uptake per person is high compared to other European countries (EEA, 2016). The urbanisation process accelerated in the post-war period (Tervo, 2010). For instance, in the Helsinki metropolitan area, yearly land uptake was 3.9% in 1950–1960 (EEA, 2006). According to recent findings, the population of the Finnish urban fringe areas was on the increase throughout the economic expansion in the 2000s, but started to decrease during the economic recession towards the end of the decade (Ristimäki et al., 2017). Along with the housing areas, it has been reported that a lot of services (including retail), moved out of the city centres and subcentres into more car-oriented areas during the 2000s (Rehunen et al., 2014), indicating urban sprawl development. The spatial expansion and sprawl of cities as a whole have been widely studied, particularly in European (Alphan, 2003; Salvati & Sabbi, 2011; Solon, 2009), Asian (Bagan & Yamagata, 2012; Dewan & Yamaguchi, 2009; Lv, Dai, & Sun, 2012), and American (Alig, Kline, & Lichtenstein, 2004; Hasse & Lathrop, 2003; Kaza, 2013; Theobald,

E-mail address: maija.tiitu@ymparisto.fi. https://doi.org/10.1016/j.apgeog.2018.10.001 Received 25 August 2017; Received in revised form 14 August 2018; Accepted 15 October 2018 0143-6228/ © 2018 Elsevier Ltd. All rights reserved.

Applied Geography 101 (2018) 1–13

M. Tiitu

1996). In these studies, the approach has been helped to recognise, classify and assess the automobile dependency of cities that results from urban sprawl. The approach has been further applied in analysing urban dynamics in the form of GIS-based travel-related zones (Ristimäki, Kalenoja, & Tiitu, 2011; Ristimäki, Tiitu, Kalenoja, Helminen, & Söderström, 2013). Since then, several European case studies have found significant differences between urban zones by applying the method in studying urban phenomena. Topics cover urban form and travel behaviour (Boitor, Antov, Iliescu, Antso, & Mäe, 2015; Mäe, Antov, Antso, & Kalenoja, 2013; Söderström, Schulman, & Ristimäki, 2015), urban well-being (Ala-Mantila, Heinonen, Junnila, & Saarsalmi, 2018), and physical activity (Mäki-Opas et al., 2016). In this study, the urban expansion in 34 Finnish city regions was examined by using GIS and the approach of travel-related urban zones. Based on previous studies, the research hypothesis was to find urban sprawl type expansion but also significant differences between different urban zones. Research objectives of this study were (1) to determine which land cover classes and urban zones have been the most impacted by the increase in the built-up areas in Finnish city regions during 2000–2012, and (2) to compare the development of different city regions and urban zones in terms of urban sprawl and its presumed impacts.

2001) cities, and in the global context (Angel, Parent, Civco, Blei, & Potere, 2011). Studies on a more detailed scale often include landscape metrics (Jaeger & Schwick, 2014; Jat, Garg, & Khare, 2008; Uuemaa, Mander, & Marja, 2013; J. G. Wu, Jenerette, Buyantuyev, & Redman, 2011). Administrative borders, gradient analyses, or concentric ring partitioning are often used for spatial grouping of the results (Jiao, 2015; Weng, 2007; J. G. Wu et al., 2011). However, there are fewer studies in which the cities have been spatially classified according to their urban form (Nong, Fox, Miura, & Saksena, 2015) or transport system. Data based on administrative borders is often easily available. However, city boundaries may change over time, and they do not divide city regions by their structure but rather divide them arbitrarily into areas of different sizes. This makes comparing city regions difficult. In gradient or moving window analysis, certain transect or landscape windows are selected and analysed according to different distances in the urban–rural gradient (Weng, 2007; Q. Wu et al., 2006; Yang, Zhou, Gong, & Wang, 2010). This is a good method for interpreting land cover changes in relation to the urban–rural gradient, but it does not spatially cover the whole region, which may become a problem for studies aiming at full spatial coverage. Gradient analysis can be extended to cover the whole city region with concentric ring partitioning. It is a method where several concentric buffer zones at different distances surrounding the city are studied (Jiao, 2015; Schneider & Woodcock, 2008). Using this method, the figures describing individual city regions are easily comparable. However, the method alone does not take into account different geographical characteristics, such as population density, occurring at the same distance from the city centre. For example, urban expansion often occurs along the road corridors, while large unbuilt areas remain in between (Antrop, 2004). From the perspective of sustainable urban growth, it is important to separate urban expansion in the bus or rail-based public transport corridors from that of the car-oriented areas at the same distance—these two types of urban expansion have different environmental impacts. In this study, the expansion of the built-up areas in Finnish city regions is studied by dividing the regions into three travel-related urban zones: pedestrian zones, public transport zones, and car zones. The theoretical framework for the zones is the theory of three urban fabrics: walking, transit, and automobile urban fabrics (Newman, Kosonen, & Kenworthy, 2016), which represent cities' historically developed urban structures that are related to emergence of different transport systems. This means that instead of one system, cities are composed of three overlapping systems, or urban fabrics, based on their land use and transportation facilities. Since the fabrics have different characteristics concerning, e.g., density, mobility, and lifestyles, their problems related to sustainability are different. Therefore, in order to plan sustainable cities, the fabrics need to be recognised and the regenerative actions need to be tailored accordingly. In addition, by monitoring cities' development in each of the fabrics, it is possible to assess their sustainability more accurately than by analysing the city as a whole. For example, the detection of urban sprawl is linked to the growing area of automobile city fabric. The concept of urban fabrics has been applied both in practical urban planning and in urban research. In the City of Kuopio, Finland, the approach has been developed in local master planning for over 20 years (Kosonen, 2015). It started in the 1990s as a new way of thinking: the city was viewed and planned through the recognition of three fabrics. It led to a new focus for guiding urban change, and new urban planning practices which promote sustainable development and a healthy city (Kosonen, 2015; Mäntysalo & Kanninen, 2013). The planning practices are reflected by the compact urban form of the region compared to other Finnish mid-sized city regions. In 2016, 76% of the Kuopio region inhabitants lived in areas with a density of at least 20 inhabitants per hectare (YKR, 2018). The concept of walking, transit, and automobile city has also been used in comparative city analyses in the 1980s and 1990s (Newman & Hogan, 1987; Newman & Kenworthy,

2. Material and methods 2.1. Study area Finland is a rather sparsely populated northern European country with 5.5 million inhabitants (Official Statistics of Finland, 2018). However, the population is heavily concentrated in city regions. This study focuses on all the city regions with at least 15,000 inhabitants in their central locality. The location of the 34 city regions of Finland are presented in Fig. 1. In 2012, their population range was from 1.27 million inhabitants in the Helsinki city region to 18,200 inhabitants in the Heinola region (YKR, 2018). The second and third largest city regions after Helsinki are Tampere (336,000 inhabitants) and Turku (301,000 inhabitants). The urban population in Finland is thus concentrated in the three largest city regions of Helsinki, Tampere and Turku, and 77% of the urban population live in these regions. In addition, 13 of the other regions lost population during 2000–2012. In addition to the size of the population, the Helsinki region differs in many ways from the other regions. The Helsinki city region has a large rail-based transport system available both in the downtown area (metro and tram network) and in peri-urban areas (commuter train connections). The residents of the Helsinki city region also have the longest commuting distances among the city regions. The structure of the city region is polycentric with many large subcentres. The majority of the subcentres are connected to the CBD by rail. In mid-sized regions public transport is much less used (Ristimäki et al., 2013), and it is based on bus routes. In three of the small city regions (Raahe, Pietarsaari, and Forssa), there is no viable local public transport system, apart from school transport or need-based mobility services. The city regions in Finland are concentrated in southern Finland, and most of them have a coastline either by the Baltic Sea or lakes. The waterbodies have also shaped the morphology of the city regions; they have limited urban expansion in certain directions. Due to these geographical constraints, the urban form in different city regions is fairly different from one another (Fig. 1). In general, the distances between individual regions are rather long, and the dominant land cover surrounding Finnish city regions in 2012 was forest (YKR, 2018). Croplands are more common in the southern and southwestern Finland due to more favourable climate and soil for agriculture compared to the East and North. The spatial dimension of the study areas—the 34 city regions—is based on a delineation of urban regions provided by the Finnish Environment Institute (Finnish Environment Institute, 2010). The 2

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 1. Location of the study areas. The study area was 34 Finnish city regions, shown here with the borders of administrative regions.

regions.

spatial coverage of the dataset is independent of administrative borders. Urban regions include localities with more than 15,000 inhabitants and smaller neighbouring localities defined with certain spatial rules. According to the criteria, urban regions comprise a central locality with at least 15,000 residents, with a fringe area reaching up to 5 km from the border of central locality and 3 km from the borders of other localities of the region. In order to aggregate results for different city regions, they were grouped into four groups according to the number of inhabitants and jobs in the central locality. The Helsinki region (with a division into urban core and peri-urban areas) was analysed as such. Large city regions were defined by at least 200,000 inhabitants and 100,000 jobs in the central locality (Turku and Tampere). Mid-sized regions were defined by at least 40,000 inhabitants and 20,000 jobs in the central locality. All the remaining city regions were classified as small city

2.2. Land cover data Urban land cover changes were calculated by combining the Corine Land Cover (CLC) data describing the year 2000 with CLC data describing the year 2012 containing only the built-up land cover classes. CLC data is provided throughout Europe with a spatial resolution of 100 m. Since this study focuses on Finland, a more detailed national raster version with a spatial resolution of 25 m of CLC was used to avoid some of the disadvantages reported of the rather coarse European-level data (Kabisch & Haase, 2013). The national CLC data is provided by the Finnish Environment Institute; it is based on an automated interpretation of satellite images and data integration with digital map data, such as the Topographic Database of Finland, Digiroad (digital road database 3

Applied Geography 101 (2018) 1–13

M. Tiitu

highest number of buildings was defined as the centre of the city region. Central pedestrian zones are located at a maximum Euclidean distance of one kilometre from the centre. The fringe of pedestrian zone reaches 2.5 km from the city centre (5 km for Helsinki). Subcentres have been defined according to focal statistics analysis using YKR data on population, jobs, retail and public transport supply (Söderström et al., 2015). Since the structure of Helsinki city region is polycentric, the largest (in terms of population) peri-urban subcentres were provided with a central pedestrian zone and a fringe zone as well (Fig. 3). Public transport zones are distinguished from the car zones by a decent supply of public transport during rush hour: public transport zone comprises areas with a public transport service every 30 min during rush hour (in Helsinki, the criterion is 15 min). For the intensive public transport zone, the limit is 15 min (Turku, Tampere 10 min, Helsinki 5 min). In addition to these service level criteria, acceptable walking distance is demanded for areas classified as public transport zones: 250 m for bus stops and 400 m for railway stations. All remaining areas of the localities are defined as car zones (Söderström et al., 2015). Distances are calculated for each grid cell by buffering the transit stops according to the distance criteria, and the grids are classified into zones according to their centre point. In the Helsinki city region, travel-related zones are dealt with separately for urban core areas and peri-urban areas, since the city region is much larger both in terms of population and area. In addition, the public transport system of the area is more complex, as described in section 2.1. The spatial extent of the core area was discussed and defined in cooperation with local transport and land use planning experts. The urban core area used in this study included areas up to 10–15 km ground distance from the centre of Helsinki. In the area of impact of the main railway line and the coastal railway line, the distance was extended to 20–25 km from the city centre. This urban core area used in this study is a combination of “urban core” and “outer urban areas” used by Söderström et al. (2015). The remaining areas of Helsinki city region are considered peri-urban in this study.

of Finland) and the SLICES land-use database (Finnish Environment Institute, 2000). The built-up land cover classes of CLC are based on data of existing buildings and other developments by the Population Register Centre. The resolution of the official CLC 2012 data is different (20 m) from the CLC 2000 data (25 m). To retain comparability with CLC 2000, a 25-m raster layer describing the urban land cover classes of 2012 was used. The dataset was additionally provided by the Finnish Environment Institute. The Finnish CLC data has a hierarchical classification system for land cover at four different levels. The level 1 classification has five classes, and the level 4 has a total of 30 classes that occur in the Finnish city regions. The number of classes in the level 1 classification was considered good for a city region scale study. However, the actual classification was not optimal to study land conversion and infill development in city regions, since the level 1 class “built-up areas” contains very different types land cover. For example, the level 4 class 1410 “green urban areas” belongs to “built-up areas” in the level 1 classification. Consequently, the level 4 classes were reclassified to be better suited to analysing changes in built-up areas by separating the core functions (residential, industrial and commercial areas) from the more extensive built-up areas that have potential for urban infill development (Fig. 2). In practice, all the 30 level 4 classes were reorganised into five slightly different classes compared to original level 1 classification. The main difference is dividing the built-up areas into two different classes based on their infill development potential. Summer cottages are classified here as “residential, commercial and industrial areas”, since according to a recent study, second homes in Finland are becoming more equipped and do not often differ from permanently inhabited detached houses in the countryside (Adamiak et al., 2015). Describing the year 2012, only the classes belonging to residential, industrial, and commercial areas were used. They were spatially compared with previous land cover in 2000 using GIS. The level 4 classes of CLC data were additionally reclassified according to whether they are to be considered green spaces suitable for recreation. The reclassification was done according to a classification used in a set of national indicators that are developed for planning sustainable urban regions and utilize CLC data (Kopperoinen et al., 2012).

2.4. Socio-economic data Socio-economic data was also added to further analyse the land cover changes. Data on residential floor space of developments was first combined with CLC data. The floor space data is a part of national Building and Dwelling Register (BDR) provided by the Finnish Population Register Centre (Population Register Centre, 2018). This point dataset covers each building in Finland. For this study, all developments classified as residential and constructed between 2000 and 2012 were extracted and joined with the CLC data of 2000. The analysis was done to find out how large a share of the floor space of the new residential developments built during 2000–2012 was constructed on land cover that was previously built-up. One way of studying the character of urban expansion is by comparing the growth of the built-up areas to the population growth in the same time period (Bagan & Yamagata, 2012; Nong et al., 2015; Subasinghe, Estoque, & Murayama, 2016). This indicator is selected as one of the Green Growth Indicators by OECD (OECD, 2017). The population change (%) was calculated from the YKR database. The population data of YKR represents the permanent population as a cross section describing the end of each year. It includes all the citizens according to their place of residence, if it is known by the authorities. In this study, the total population of the year 2000 was compared to the one in 2012. The growth of the built-up areas was calculated by comparing the area of CLC classes belonging to the class “residential, industrial, and commercial areas” (Fig. 2) in 2000 and 2012. The built-up area change and the population change were calculated both for each city region and for each zone inside each of the city regions. Problems related to different spatial resolution (25 m for CLC, 250 m for YKR) were handled using the travel-related urban zones as the smallest spatial unit in the population analysis. The travel-related zones data is

2.3. Travel-related urban zones Travel-related urban zones, describing the year 2010, were used as a spatial grouping of the results. The data is grid-based with a spatial resolution of 250 m (Fig. 3). It is provided by the Finnish Environment Institute SYKE. The main criteria for the zones are the distance to the city centres and public transport supply (Söderström et al., 2015). The urban zone classification gives an overview of mobility possibilities for different locations that is based on qualities of prevalent urban form and transport system. The zone definitions are based on available options rather than actual travel behaviour, but surveys conducted in urban regions show that the classification is also consistent with the actual travel behaviour (Ristimäki et al., 2011, 2013). Even though urban fabrics, as defined by Newman et al. (2016), are overlapping systems, travel-related zones applied in this study are not. For simplicity, each grid cell is classified into only one of the zones. Walking urban fabric defined by Newman et al. (2016) is mainly represented by different pedestrian zones. The distance criteria for the pedestrian zones are calculated from a point that describes the centre for each city region. City centres were defined by the number of buildings in the industries that usually gravitate towards the centre (shopping, restaurant and office), based on the national database—the Monitoring System of Spatial Structure and Urban Form (YKR). The data is grid-based with a spatial resolution of 250 m, and it is provided by the Finnish Environment Institute and Statistics Finland. Focal statistics, i.e., summing figures of neighbouring cells were used as a method for locating centres. The centre point of the grid cell with the 4

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 2. The reclassification of CLC 2000 data. The CLC data was reclassified into five classes from the original level 4 classes. Classes marked with an asterisk (*) are the classes interpreted as green spaces suitable for recreation.

Travel-related urban zones show the differing intensity of land use. For example, on average more than 80% of the land area in the pedestrian zones of Finnish cities is developed (YKR, 2018). For car-oriented areas the share is 50%. Outside of densely populated localities, over 90% of the land area consists of natural or semi-natural areas. These statistics are supported by the findings on the Helsinki metropolitan region by Ristimäki et al. (2011), who found that not only spatial characteristics (such as population density, household size) vary considerably depending on the travel-related zone but also travel behaviour. For example, the use of public transport was much lower in car-oriented zones than in pedestrian or public transport zones. Furthermore, Mäkiopas et al. (2016) found out that commuting physical activity of people living in car zone is significantly weaker compared to people living in pedestrian zones. Since these structural and behavioural differences between zones exist, studying land use changes of the zones may provide important information on the expansion of

spatially compatible with the YKR data with the same spatial resolution. 3. The theory of urban fabrics in relation to green spaces Land use and transport are interconnected. Land use determines the need for spatial interaction, i.e. transport; however transport, by the accessibility it provides, also determines spatial development (Wegener, 2004). According to Antrop (2000), transportation infrastructure and accessibility have a dominant role in land cover changes both in urban and rural areas. Consequently, it is convenient to study urban expansion using a spatial delineation that takes into account both the transport system and the urban form. This kind of holistic spatial classification also considers the varying intensity of land use between different urban forms, since transport systems require certain levels of residential densities to be economically sustainable (Newman & Kenworthy, 2006). 5

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 3. Travel-related urban zones data. Example of the zone data in the Helsinki city region, Southern Finland. Smaller regions of Hyvinkää-Riihimäki, Lohja, and Porvoo are located in the vicinity of the Helsinki city region.

on the geographical qualities of land, large agricultural areas are often located in the outer fringe of the city. Located on the edge of the city regions, it is often the agricultural areas that start to shrink when cities are growing (Alig et al., 2004; Brown, Johnson, Loveland, & Theobald, 2005; Deng, Qiu, Wang, Yang, & Shi, 2011; Skinner, Kuhn, & Joseph, 2001). However, the Helsinki region has been reported to have more forest loss compared to other European cities (Kasanko et al., 2006).

Finnish city regions. Söderström et al. (2015) compared the development of urban form in the city-regions of Helsinki and Stockholm using travel-related urban zones. One of their main results was that pedestrian zones of Helsinki had been losing jobs to more car-oriented locations, whereas in Stockholm, the workplaces had been concentrated in rail-based subcentres. The recognition and delineation of travel-related urban zones were utilised in revealing crucial differences in the urban sprawl of the two regions. There are many spatial configurations of urban development, like edge-expansion on the urban fringe, infill development, and outlying urban sprawl in the peripheral areas (Lv et al., 2012). The theory of urban fabrics can also be applied to studying these kinds of land cover changes and their impacts (Fig. 4), since the varying density and structure of the fabrics set different preconditions for urban planning (Newman et al., 2016). In walking cities, which are located in the vicinity of city centres, green spaces are often intensively maintained parks and other small green spaces most often threatened by infill development. With a transit city fabric, larger green areas are more likely to exist between the road corridors due to the "star-shaped urban sprawl" along the traffic lanes (Antrop, 2004). These green wedges are also exposed to infill policies, since infill development is preferably practiced in locations with good accessibility. In contrast, automobile city fabric, where the urban and rural areas meet, is characterized by large farmlands and forests, often highly fragmented by dispersed housing (Weng, 2007). Even though the distribution of land cover classes generally depends

4. Results 4.1. The expansion of residential, industrial and commercial areas The expansion of residential, industrial and commercial areas was examined by comparing land cover data from 2000 to the data from 2012. According to the results, forests were the most common land cover type to be displaced by the expansion (Fig. 5). In the pedestrian zones, approximately 20% of the expansion happened on previously developed land cover classes, particularly with industrial units expanding into previously existing port areas. In contrast, in public transport and car zones, the share of building in previously developed areas was only 10%. Developing agricultural land is far more common in public transport and car zones than in pedestrian zones. Obviously, this is due to the original land cover distribution on the area—there is more agricultural land in the fringe of the city. Urban expansion to agricultural areas is most widespread in southwestern Finland. In contrast, expansion of the built-up areas into former forests seems to have been about as common 6

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 4. Illustration on the approach of three urban fabrics in relation to green spaces and urban expansion. Examples of the types of urban expansion in transit urban fabric and in automobile urban fabric. A. Infill development near Kauklahti railway station in the Helsinki city region. B. Scattered housing in a car-dependent location in the Jyväskylä city region. Sources: Illustration; Finnish Environment Institute SYKE. GIS data; SYKE (partly METLA, MMM, MML, VRK), Digiroad. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

4.2. Construction of residential floor space in relation to previous land cover

in all the zones and cities. Forests have been developed not only in the fringe of the city, but also in denser inner city areas. Urban forests form the basis of the urban green network for recreation purposes. The impact of urban expansion in these areas that were previously classified as green spaces suitable for recreation was studied. On average 77% of the area of urban expansion in all the 34 regions was previously classified as green spaces suitable for recreation. However, inside each city region, there are great differences in total developed areas between urban zones. In addition to the differences between different zones, in Helsinki metropolitan region the expansion in the urban core area has been very different when compared to the surrounding peri-urban area. In the urban core area, most of the expansion has occurred in the public transport and car zone. In contrast, the peri-urban area has the most expansion outside localities (Fig. 6).

In the above results, the area of industrial and commercial units was examined together with the residential area. In addition to this, it is important to separate non-residential areas—often covering large areas of land—from residential construction in order to interpret the sprawl and densification of housing in the city regions. For this reason, the distribution of new residential floor space into former land cover classes was studied. The majority of the floor space of new residential developments seems to have been located on previously built-up areas. However, the travel-related zone approach gives a whole new perspective to the results. In Table 1, city regions are divided both into zones and according to the previous land cover. Thus, the table shows into which zones housing has concentrated and whether the area was already developed in 2000. In the largest city regions, Helsinki and Tampere, approximately a 7

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 5. Distribution of land cover classes that were displaced by the expansion of residential, industrial and commercial areas in 2000–2012. The shares represent all of the 34 city regions.

third of the residential floor space was located in the public transport zones that had already been developed. In the next largest regions, Turku and Oulu, much more of the new residential floor space was concentrated in the car zone areas that were not previously developed. This is also the case in the majority of the mid-sized and small city regions. As an exception, in the city region of Kuopio, the main housing zone was the public transport zone that was not previously developed. This is related to a construction of a rather large new neighbourhood of Saaristokaupunki ("the Town of Islands") connected to the city centre by a new road and a bus line across Lake Kallavesi. In some of the mid-sized and small regions much of the residential floor space was located in the pedestrian zones. This is mainly because in rather small cities the central pedestrian zone and its fringe already cover a large proportion of the whole locality. Since the zones of Table 1 have been generalised by combining the numbers of these both zones, much of the housing activities have been located inside this area as a result. Actually, many areas of the fringe of pedestrian zones in

small cities are rather car-oriented (Ristimäki et al., 2013), and thus represent an automobile urban fabric rather than a walking urban fabric. 4.3. Population growth in relation to urban expansion The growth of the built-up areas in Finnish city regions during 2000–2012 was compared to the population growth in different travelrelated urban zones (Fig. 7). When dealing with all the zones, the builtup areas increased at a much larger pace than the population in almost all of the regions, apart from the large city regions of Helsinki, Tampere and Oulu. The built-up areas exceeded the population growth altogether by a factor of 1.6 (calculated by percentage of built-up area change divided by the percentage of population change). However, the relationship between the variables is very different, depending on the zone. There is no significant relationship between population change and growth of built-up area in the pedestrian zones. The relationship is

Fig. 6. Developing green spaces in the urban zones of Helsinki region during 2000–2012. The darker bars represent the area of new residential, industrial or commercial areas in hectares that is constructed on previous green spaces suitable for recreation during 2000–2012. The "other land cover" class comprises of both developed and semi-natural areas. 8

Applied Geography 101 (2018) 1–13

M. Tiitu

Table 1 The distribution of residential floor space of developments built during 2000–2012 into travel-related urban zones and into developed and undeveloped areas in 2000 (%). The largest share is bold for each region. The land cover is defined according to the centre of the buildings. The cities are grouped according to the population in their central locality. Pedestrian zonesa

City region

METROPOLITAN REGION

LARGE CITY REGIONS MID-SIZED CITY REGIONS

SMALL CITY REGIONS

a b c

Helsinki - urban core area - peri-urban area Tampere Turku Oulu Lahti Jyväskylä Pori Kuopio Hyvinkää-Riihimäki Vaasa Joensuu Kotka-Hamina Lappeenranta Kouvola Rovaniemi Hämeenlinna Seinäjoki Kemi-Tornio Porvoo Mikkeli Kokkola Lohja Rauma Kajaani Salo Imatra Savonlinna Forssa Pietarsaari Varkaus Raahe Heinola Valkeakoski Iisalmi

Public transport zonesb

Car zonesc

Developed

Not developed

Developed

Not developed

Developed

Not developed

18 16 22 10 12 10 14 13 21 18 20 14 13 34 29 21 22 15 16 29 20 11 29 16 20 40 16 24 27 31 23 22 23 20 26 33

4 2 9 6 2 5 3 3 5 1 10 5 3 8 3 8 2 3 13 11 22 6 9 3 15 8 1 21 0 12 18 10 18 20 8 10

33 41 12 27 19 17 27 25 9 20 12 14 13 10 13 7 8 16 3 1 15 15 0 3 4 7 12 3 7 0 0 4 0 11 4 7

13 16 6 14 14 18 12 11 4 27 19 5 12 6 7 5 5 18 1 2 7 3 0 1 3 5 8 2 2 0 0 1 0 4 4 2

19 16 25 18 23 18 17 16 30 8 15 30 15 23 15 27 26 17 18 27 15 21 24 36 25 16 14 27 24 24 28 32 29 18 15 16

13 8 25 25 30 31 29 32 30 25 25 32 44 20 33 32 36 33 48 30 22 43 38 40 32 25 49 24 39 33 31 32 31 27 42 32

Pedestrian zone, fringe of pedestrian zone, subcentres Intensive public transport zone, public transport zone including the fringe of pedestrian zone with public transport supply Car zone and the areas outside localities

stronger in public transport zones (R2 = 0.296) and the strongest in car zones, where population change explains almost 40% of the variation in the growth of built-up area (R2 = 0.390). In pedestrian zones, the land use is more mixed, and new developments are often built denser and higher, which decreases the need for land conversion despite a growing population. This is one of the key reasons for the different relationship between population change and built-up area growth in the zones. The built-up area grew even in the areas of negative population growth. This is mainly related to a trend of decreasing household size (Liu, Daily, Ehrlich, & Luck, 2003). The declining Finnish cities suffer from rapid aging due to heavy emigration to larger cities, which tends to decrease the household sizes even more. These household dynamics lead to a situation, where elderly people move to the city centre to get closer to services, but their old apartments become vacant, since their location and the condition do not attract younger residents and families. Thus, new developments are built in spite of the declining population. In remote locations, it is typical to have many empty apartments even within cities (Sikiö, Pitkänen, & Rehunen, 2014). A statistical significance (p < 0.001) was found for differences in population change (%) between travel-related zones using the Kruskal–Wallis test. In addition, the differences in built-up area growth (%) between zones were significant (p < 0.001). Table 2 presents the growth of the built-up areas and change of population in different travel-related zones and city region groups. The high growth of the

built-up areas compared to population growth relates to very low or negative population growth in the public transport zones of small and mid-sized city regions. The built-up areas of these zones have still grown, and the biggest growth concerns the cities' car zones. This reflects the findings on planning strategies in small and mid-sized Finnish cities, which imply that only a few of the cities' main new development projects concern public transport oriented neighbourhoods. These neighbourhoods are mostly built in the 1960s and 1970s, and they have suffered from depopulation. The plans of the cities contain developments into either pedestrian zones or more car-oriented locations (Ristimäki et al., 2017). 5. Discussion In this study, the expansion of the built-up areas in 34 Finnish city regions during 2000–2012 was analysed. The key methodology was combining land cover data with a spatial classification of travel-related urban zones. Since land use and transport are interconnected, it was hypothesized as beneficial to use a spatial classification that takes into account both urban form and transport systems. The objectives were to determine which land cover classes and zones have been the most impacted by the increase of the built-up areas and to compare the development of different city regions and zones in terms of urban sprawl and its presumed impacts. 9

Applied Geography 101 (2018) 1–13

M. Tiitu

Fig. 7. Linear regression between growth of residential, industrial and commercial areas and population change during 2000–2012 in travel-related urban zones. Apart from Helsinki, each of the observations represents one city region. Helsinki is presented by two observations that describe the urban core area and the periurban area. All zones: p < 0.01, R2 = 0.271; Pedestrian zones: p > 0.05, R2 = 0.0281; Public transport zones: p < 0.01, R2 = 0.296; Car zones: p < 0.001, R2 = 0.390.

According to the results, most of the new built-up areas were previously different kinds of forests. In the European context, Finland is an exception—in many central European cities, most of the urban

expansion is concentrated on former farmlands (EEA, 2006). The results describe the land cover characteristics of Finnish city regions: many of the cities are isolated from each other, and they are surrounded by

Table 2 Change in population (POP, %) and in the area of residential, industrial and commercial developments (BUILT-UP, %) during 2000–2012. Changes are grouped by urban zones and by city regions of different size presented in Table 1. City region

Helsinki metropolitan region - urban core area - peri-urban area Large city regions Mid-sized city regions Small city regions All city regions

Pedestrian zones

Public transport zones

Car zones

All zones

BUILT-UP %

POP %

BUILT-UP %

POP %

BUILT-UP %

POP %

BUILT-UP %

POP %

9 7 10 8 9 9 9

13 12 14 13 8 −1 8

7 6 10 10 11 7 9

11 10 19 7 4 −6 7

15 9 18 21 23 20 20

21 18 28 22 18 6 17

11 7 16 17 18 16 16

13 12 20 13 10 0 10

10

Applied Geography 101 (2018) 1–13

M. Tiitu

areas of the Helsinki city region, residential areas are located far from workplaces and services, which increases commuting distances and leads to the larger amount of carbon dioxide emissions of mobility (Ristimäki et al., 2011). The area of detailed planning only covers less than one fifth of the area of the Finnish city regions (YKR, 2018). Thus, there might be a need for stricter governance in order to steer the expansion of cities. For example, there has not been any official greenbelt designation policies in use in Finland, though the approach applied for example in the UK has crucially decreased development in greenfield areas (Baing, 2010). Decreasing population has not had an effect on the expansion of the built-up areas in certain city regions. Urban expansion, regardless of negative population growth, has been reported in many other European cities as well (Kabisch & Haase, 2013; Kroll & Haase, 2010). On average, the growth of built-up areas in the regions exceeded the growth of population by a factor of 1.6, which is rather high for cities that have already gone through the period of their fastest urban growth. For reference, the corresponding factor was 2.6 in the Tokyo metropolitan region during its great expansion from the 1970s to the present (Bagan & Yamagata, 2012). Significant differences between urban zones were found in the relationship of population and urban area change. Previously, differences in urban zones have been detected in at least travel behaviour and physical activity (Mäki-Opas et al., 2016; Ristimäki et al., 2011). The classification of travel-related urban zones can thus be seen as relevant for recognising different spatial and behavioural patterns in city regions. In addition, the approach has already been applied in developing urban planning practices in Finland (e.g. Kosonen, 2015) and elsewhere. For example, Boitor et al. (2015) have used the methodology to identify planning problems and solutions tackling car dependency in the city of Cluj-Napoca, Romania. The need for such integrated tools that serve both research and planning is likely to increase, as urbanisation is accelerating worldwide. For making conclusions on the neighbourhood scale, the resolution of CLC data is rather coarse. This means that, for example, areas defined as already developed actually comprise many green spaces such as pocket parks, yards and roadsides. Thus, the results of this study often do not represent brownfield development, but rather the building of new developments near existing neighbourhoods. As a result, the spatial delineation of the built-up areas in this study is a little bit overestimated, which means that even more of the developments have been constructed into areas that were not previously developed. When interpreting the land cover changes at a more detailed scale, data more accurate than CLC should be used. For instance, remote sensing technologies, such as laser scanning (LiDAR), provide high-resolution land cover data that could be applied in analysing the expansion of the builtup areas (Niemeyer, Rottensteiner, & Soergel, 2014).

boreal forests. Agricultural land was developed the most in growing mid-sized city regions of southern Finland, mainly due to their geographical conditions suitable for agriculture. In the population-gaining areas, selling parcels of land soon becomes more profitable for the landowners compared to agriculture (EEA, 2006). At the same time, the recreational value of agricultural land increases near cities (Buciega, Pitarch, & Esparcia, 2009); and there are also environmental and ethical issues involved (Deng et al., 2011). Therefore, developing agricultural areas near cities deserves careful consideration from many perspectives. However, in the Finnish context, developing agricultural land is not considered cost-effective to start with because piles for buildings have to be driven deep through the thick layers of clay that are also exposed to frost during wintertime. Housing activities have been concentrating in the car zone and outside localities in many of the Finnish city regions, indicating urban sprawl. Urban sprawl has also been reported for previous decades in European cities using a variety of indicators (Kasanko et al., 2006; Salvati, Sateriano, & Bajocco, 2013). It has also been detected that the population density of Finnish city regions has been on the decrease throughout the 2000s (Ristimäki et al., 2013). The same trend of declining density has been reported globally for built-up areas (Angel et al., 2011). According to Söderström et al. (2015), 39% of the population growth in the Helsinki region in 2000–2010 was concentrated in car zones (pedestrian zones 31%, public transport zones 30%). Results of this study also showed that the built-up areas increased the most (15%) in car zones compared to other zones in Helsinki. However, the largest share of the residential floor space was concentrated in public transport zones, indicating a more transit-oriented planning compared to other cities. The difference reflects Fig. 7: land uptake in public transport zones is lower compared to car zones, where more of the developments are detached houses. Even though most of the residential floor space was concentrated in public transport zones, the urban area increased the most in car zones. The urban expansion in car zones reported in this study is also reflected by the motorisation rate in Finland that has been increasing steeply throughout the 2000s, in contrast to Sweden for example (Antso, Antov, & Mäe, 2013). One reason for the urban sprawl outside densely populated areas in Finnish city regions is the loose regulation of residential land use in Finland. The frequency of constructing new developments with exceptional permits was studied as part of the assessment on the effectiveness of Finnish Land Use and Building Act (Ministry of the Environment, 2014). During 2000–2011, an exceptional permit application was submitted for a total of 32,500 construction sites, of which only 5% were rejected by the municipality. According to the report, building without planning is concentrated in the largest cities and their surrounding periurban municipalities. This relates to another problem accelerating urban sprawl in Finland—the competition for tax payers between municipalities (Hytönen et al., 2016), which could be tackled for example by rearranging the municipality structure into bigger units. This would also help the integrated planning of land use and transport system within the regions. The state has created incentives for the municipalities of the city regions to cooperate in urban planning. In addition, agreements on land use, housing and transport between the state and municipalities are made in order to steer urban development. Nevertheless, these efforts have not been able to solve the problems related to the lack of integrated, inter-municipal planning (Hytönen et al., 2016). Constructing without a conventional planning process lacks proper impact assessment. For example, an urban green network is in danger of severe fragmentation if urban areas expand in an unplanned manner (EEA, 2011). In almost every city region studied, over half of the area of new residential, industrial or commercial units was located in car zone locations that were previously green spaces suitable for recreation. This kind of development is likely to cause both conflicts between different land uses and fragmentation of green spaces used for recreational activities. In addition to green space fragmentation, urban sprawl causes transport-related environmental problems as well. In the peri-urban

6. Conclusions The main trend of urban expansion in Finnish city regions has been developing the previously natural areas in car-oriented areas in the fringes of the cities. This kind of development has many negative environmental impacts. However, the trend of infill development can already be seen in the development of the core areas of the growing city regions of Helsinki and Tampere. The differences were revealed by using the approach of travel-related urban zones, dividing the regions based on their urban form and transport system. The built-up areas have not reflected the declining population of some of the regions, but have continued to grow. When cities are increasingly shifting into infill development policies, it becomes more and more important to reasonably integrate infill development and urban green spaces. The travel-related urban zones provide one tool for structuring the city for that purpose. The resolution of the land cover data needs to be taken into account in analysing cities. Research on the expansion of built-up areas at a scale 11

Applied Geography 101 (2018) 1–13

M. Tiitu

of individual neighbourhoods will become even more important as urban infill development proceeds.

European Urban and Regional Studies, 23(1), 40–55. https://doi.org/10.1177/ 0969776413490424. Jaeger, J. A. G., & Schwick, C. (2014). Improving the measurement of urban sprawl: Weighted Urban Proliferation (WUP) and its application to Switzerland. Ecological Indicators, 38, 294–308. https://doi.org/10.1016/j.ecolind.2013.11.022. Jat, M. K., Garg, P. K., & Khare, D. (2008). Monitoring and modelling of urban sprawl using remote sensing and GIS techniques. International Journal of Applied Earth Observation and Geoinformation, 10(1), 26–43. https://doi.org/10.1016/j.jag.2007. 04.002. Jiao, L. M. (2015). Urban land density function: A new method to characterize urban expansion. Landscape and Urban Planning, 139, 26–39. https://doi.org/10.1016/j. landurbplan.2015.02.017. Johnson, M. P. (2001). Environmental impacts of urban Sprawl: A survey of the literature and proposed research agenda. Environment and Planning A: Economy and Space, 33(4), 717–735. https://doi.org/10.1068/a3327. Kabisch, N., & Haase, D. (2013). Green spaces of European cities revisited for 1990-2006. Landscape and Urban Planning, 110, 113–122. https://doi.org/10.1016/j.landurbplan. 2012.10.017. Kasanko, M., Barredo, J. I., Lavalle, C., McCormick, N., Demicheli, L., Sagris, V., et al. (2006). Are European cities becoming dispersed? A comparative analysis of 15 European urban areas. Landscape and Urban Planning, 77(1–2), 111–130. https://doi. org/10.1016/j.landurbplan.2005.02.003. Kaza, N. (2013). The changing urban landscape of the continental United States. Landscape and Urban Planning, 110, 74–86. https://doi.org/10.1016/j.landurbplan. 2012.10.015. Kopperoinen, L., Eerola, K., Shemeikka, P., Väre, S., Söderman, T., & Saarela, S.-R. (2012). Criteria and indicators for planning of sustainable urban regions – description of GIS-methods (in Finnish). Reports of the Finnish Environment Institute, Vol. 28Helsinki: Finnish Environment Institute 2012. Kosonen, L. (2015). The three fabrics strategy in Finland. In H. Barton, S. Thompson, S. Burgess, & M. Grant (Eds.). The routledge handbook of planning for health and wellbeing: Shaping a sustainable and healthy future (pp. 521–539). (1st ed.). London: Routledge. Kroll, F., & Haase, D. (2010). Does demographic change affect land use patterns?: A case study from Germany. Land Use Policy, 27, 726–737. https://doi.org/10.1016/j. landusepol.2009.10.001. Liu, J., Daily, G. C., Ehrlich, P. R., & Luck, G. W. (2003). Effects of household dynamics on resource consumption and biodiversity. Nature, 421(6922), 530–533. https://doi. org/10.1038/nature01359. Luck, M., & Wu, J. G. (2002). A gradient analysis of urban landscape pattern: A case study from the phoenix metropolitan region, Arizona, USA. Landscape Ecology, 17(4), 327–339. https://doi.org/10.1023/a:1020512723753. Lv, Z. Q., Dai, F. Q., & Sun, C. (2012). Evaluation of urban sprawl and urban landscape pattern in a rapidly developing region. Environmental Monitoring and Assessment, 184(10), 6437–6448. https://doi.org/10.1007/s10661-011-2431-x. Mäe, R., Antov, D., Antso, I., & Kalenoja, H. (2013). Determining the travel-related zones in urban and suburban areas of Tallinn. Paper presented at the the XXVIII International Baltic Road Conference (Vilnius, Lithuania). Mäki-Opas, T. E., Borodulin, K., Valkeinen, H., Stenholm, S., Kunst, A. E., Abel, T., & Koskinen, S. (2016). The contribution of travel-related urban zones, cycling and pedestrian networks and green space to commuting physical activity among adults – a cross-sectional population-based study using geographical information systems. BMC Public Health, 16(1), 14. https://doi.org/10.1186/s12889-016-3264-x. Mäntysalo, R., & Kanninen, V. (2013). Trading between land use and transportation planning: The Kuopio model. In A. Balducci, & R. Mäntysalo (Vol. Eds.), Urban and landscape perspectives: . Vol. 13. Urban planning as a trading zone (pp. 57–73). Ministry of the Environment (2014). Assessment of the effectiveness of the land use and building Act in 2013 (in Finnish). Finnish environment, 1/2014. Helsinki: Ministry of the Environment. Newman, P., & Hogan, T. (1987). Urban density and transport: A simple model based on three city types, 1/87, Murdoch: Murdoch University37. Newman, P., & Kenworthy, J. (1996). The land use—transport connection: An overview. Land Use Policy, 13(1), 1–22. https://doi.org/10.1016/0264-8377(95)00027-5. Newman, P., & Kenworthy, J. (2006). Urban design to reduce automobile dependence. Opolis, 2(1), 35–52. Newman, P., Kosonen, L., & Kenworthy, J. (2016). Theory of urban fabrics: Planning the walking, transit/public transport and automobile/motor car cities for reduced car dependency. Town Planning Review, 87(4), 429–458. https://doi.org/10.3828/tpr. 2016.28. Niemeyer, J., Rottensteiner, F., & Soergel, U. (2014). Contextual classification of lidar data and building object detection in urban areas. ISPRS Journal of Photogrammetry and Remote Sensing, 87, 152–165,. https://doi.org/10.1016/j.isprsjprs.2013.11.001. Nong, D. H., Fox, J., Miura, T., & Saksena, S. (2015). Built-up area change analysis in Hanoi using support vector machine classification of landsat multi-temporal image stacks and population data. Land, 4(4), 1213–1231. https://doi.org/10.3390/ land4041213. OECD (2017). Built-up area growth surpassed population growth in most countries: OECD and G20 OECD green growth studies, green growth indicators 2017. Paris: OECD. Official Statistics of Finland (2018). Population structure. e-publication. Retrieved 11.5.2018, from Statistics Finland http://www.stat.fi/til/vaerak/index_en.html. Population Register Centre (2018). Building information. Retrieved 15.5.2018, from Population Register Centre http://vrk.fi/en/building-information. Rehunen, A., Helminen, V., Kosonen, P., Viinikka, A., Ahonen, O., & Käyhkö, H. (2014). City centres and retail services in urban regions (in Finnish). Reports of the Ministry of the Environment, Vol. 8Helsinki: Ministry of the Environment 2014. Ristimäki, M., Kalenoja, H., & Tiitu, M. (2011). Travel-related zones of urban form – zone

Acknowledgements Writing of this article has been funded by Strategic Research Council at Academy of Finland for the project “Beyond MALPE-coordination: Integrative Envisioning (BeMInE, project number 303556) and Finnish Ministry of the Environment. I would like to thank the two anonymous reviewers and Mr. Antti Rehunen for the valuable comments on the manuscript. References Adamiak, C., Vepsäläinen, M., Strandell, A., Hiltunen, M., Pitkänen, K., Hall, M., & Åkerlund, U. (2015). Second home tourism in Finland - perceptions of citizens and municipalities on the state and development of second home tourism. Reports of the Finnish environment Institute: Vol. 22Helsinki: Finnish Environment Institute en/2015. Ala-Mantila, S., Heinonen, J., Junnila, S., & Saarsalmi, P. (2018). Spatial nature of urban well-being. Regional Studies, 52(7), 959–973. https://doi.org/10.1080/00343404. 2017.1360485. Alig, R. J., Kline, J. D., & Lichtenstein, M. (2004). Urbanization on the US landscape: Looking ahead in the 21st century. Landscape and Urban Planning, 69(2–3), 219–234. https://doi.org/10.1016/j.landurbplan.2003.07.004. Alphan, H. (2003). Land-use change and urbanization of Adana, Turkey. Land Degradation & Development, 14(6), 575–586. https://doi.org/10.1002/ldr.581. Angel, S., Parent, J., Civco, D. L., Blei, A., & Potere, D. (2011). The dimensions of global urban expansion: Estimates and projections for all countries, 2000-2050. Progress in Planning, 75, 53–107. https://doi.org/10.1016/j.progress.2011.04.001. Antrop, M. (2000). Changing patterns in the urbanized countryside of Western Europe. Landscape Ecology, 15(3), 257–270. https://doi.org/10.1023/A:1008151109252. Antrop, M. (2004). Landscape change and the urbanization process in Europe. Landscape and Urban Planning, 67(1–4), 9–26. https://doi.org/10.1016/s0169-2046(03) 00026-4. Antso, I., Antov, D., & Mäe, R. (2013). Settlement changes effect to mobility in suburban area of tallinn. Journal of International Scientific Publications: Ecology & Safety, 7(2), 4–19. Bagan, H., & Yamagata, Y. (2012). Landsat analysis of urban growth: How Tokyo became the world's largest megacity during the last 40 years. Remote Sensing of Environment, 127, 210–222. https://doi.org/10.1016/j.rse.2012.09.011. Baing, A. S. (2010). Containing urban Sprawl? Comparing brownfield reuse policies in England and Germany. International Planning Studies, 15(1), 25–35. https://doi.org/ 10.1080/13563471003736910. Boitor, M., Antov, D., Iliescu, M., Antso, I., & Mäe, R. (2015). Sustainable urban transport planning. Romanian Journal of Transport Infrastructure, 2(1)https://doi.org/10.1515/ rjti-2015-0010. Brown, D. G., Johnson, K. M., Loveland, T. R., & Theobald, D. M. (2005). Rural land-use trends in the conterminous United States, 1950-2000. Ecological Applications, 15(6), 1851–1863. https://doi.org/10.1890/03-5220. Buciega, A., Pitarch, M. D., & Esparcia, J. (2009). The context of rural-urban relationships in Finland, France, Hungary, The Netherlands and Spain. Journal of Environmental Policy and Planning, 11(1), 9–27. https://doi.org/10.1080/15239080902774929. Colding, J., Elmqvist, T., Lundberg, Ahrné, K., Andersson, E., & Barthel, S. (2003). The Stockholm Urban Assessment (SUA-Sweden) Millennium assessment sub-global summary report28 9.12.2003. Deng, J. S., Qiu, L. F., Wang, K., Yang, H., & Shi, Y. A. Y. (2011). An integrated analysis of urbanization-triggered cropland loss trajectory and implications for sustainable land management. Cities, 28(2), 127–137. https://doi.org/10.1016/j.cities.2010.09.005. Dewan, A. M., & Yamaguchi, Y. (2009). Land use and land cover change in Greater Dhaka, Bangladesh: Using remote sensing to promote sustainable urbanization. Applied Geography, 29(3), 390–401. https://doi.org/10.1016/j.apgeog.2008.12.005. EEA (2006). Urban sprawl in Europe - the ignored challenge. EEA Report, Vol. 10, European environment agency56. EEA (2011). Green infrastructure and territorial cohesion - the concept of green infrastructure and its integration into policies using monitoring systems. EEA Technical Report, Vol. 18, European Enviroment Agency138. EEA, & FOEN (2016). Urban sprawl in Europe. EEA ReportVol. 11, European Enviroment Agency EEA and Federal Office for the Environment135. Finnish Environment Institute (2000). Corine land cover 2000. Retrieved 16.5.2018 http://metatieto.ymparisto.fi:8080/geoportal/catalog/search/resource/details. page?uuid=%7BEEC35D81-765A-4A7B-BC9B-F89A476807EC%7D. Finnish Environment Institute (2010). Urban regions. Retrieved 17.10.2018 http:// metatieto.ymparisto.fi:8080/geoportal/catalog/search/resource/details.page? uuid=%7BD71ED5D0-9DD2-4C72-A53C-439E6CC5F7AF%7D. Hasse, J. E., & Lathrop, R. G. (2003). Land resource impact indicators of urban sprawl. Applied Geography, 23(2–3), 159–175. https://doi.org/10.1016/j.apjeog.2003.08. 002. von Hertzen, L., Hanski, I., & Haahtela, T. (2011). Natural immunity. EMBO Reports, 12(11), 1089–1093. https://doi.org/10.1038/embor.2011.195. Hytönen, J., Mäntysalo, R., Peltonen, L., Kanninen, V., Niemi, P., & Simanainen, M. (2016). Defensive routines in land use policy steering in Finnish urban regions.

12

Applied Geography 101 (2018) 1–13

M. Tiitu

growth using GIS and remote sensing: A case study of the Colombo metropolitan area, Sri Lanka. ISPRS International Journal of Geo-information, 5(11), https://doi.org/10. 3390/ijgi5110197. Tervo, H. (2010). Cities, hinterlands and agglomeration shadows Spatial developments in Finland during 1880-2004. Explorations in Economic History, 47(4), 476–486. https:// doi.org/10.1016/j.eeh.2010.05.002. Theobald, D. M. (2001). Land-use dynamics beyond the American urban fringes. Geographical Review, 91(3), 544–564. https://doi.org/10.2307/3594740. Uuemaa, E., Mander, U., & Marja, R. (2013). Trends in the use of landscape spatial metrics as landscape indicators: A review. Ecological Indicators, 28, 100–106. https://doi. org/10.1016/j.ecolind.2012.07.018. Van Eetvelde, V., & Antrop, M. (2004). Analyzing structural and functional changes of traditional landscapes - two examples from Southern France. Landscape and Urban Planning, 67(1–4), 79–95. https://doi.org/10.1016/s0169-2046(03)00030-6. Wegener, M. (2004). Overview of land use transport models. In D. A. Hensher, K. J. Button, K. E. Haynes, & P. R. Stopher (Vol. Eds.), Handbook of transport geography and spatial systems: Vol. 5, (pp. 127–146). Weng, Y. C. (2007). Spatiotemporal changes of landscape pattern in response to urbanization. Landscape and Urban Planning, 81(4), 341–353. https://doi.org/10.1016/j. landurbplan.2007.01.009. Wu, Q., Hu, D., Wang, R., Li, H., He, Y., Wang, M., et al. (2006). A GIS-based moving window analysis of landscape pattern in the Beijing metropolitan area, China. The International Journal of Sustainable Development and World Ecology, 13(5), 419–434. https://doi.org/10.1080/13504500609469691. Wu, J. G., Jenerette, G. D., Buyantuyev, A., & Redman, C. L. (2011). Quantifying spatiotemporal patterns of urbanization: The case of the two fastest growing metropolitan regions in the United States. Ecological Complexity, 8(1), 1–8. https://doi. org/10.1016/j.ecocom.2010.03.002. Yang, Y., Zhou, Q., Gong, J., & Wang, Y. (2010). Gradient analysis of landscape spatial and temporal pattern changes in Beijing metropolitan area. Science China Technological Sciences, 53(1), 91–98. https://doi.org/10.1007/s11431-010-3206-2. YKR (2018). Finnish monitoring system of spatial structure and urban form. from Finnish Environment Institute and Statistics Finland.

criteria, urban form statistical profiles, and travel habits (in Finnish). Publications of the Ministry of Transport and Communications, Vol. 15Helsinki: Ministry of Transport and Communications 2011. Ristimäki, M., Tiitu, M., Helminen, V., Nieminen, H., Rosengren, K., Vihanninjoki, V., ... Söderström, P. (2017). The future of urban form in urban regions – urban fabrics and urban zones (in Finnish). Reports of the Finnish Environment Institute, Vol. 4Helsinki: Finnish Environment Institute 2017. Ristimäki, M., Tiitu, M., Kalenoja, H., Helminen, V., & Söderström, P. (2013). Travelrelated Urban Zones in Finland – the development of pedestrian, transit and car-oriented zones from 1985 to 2010 (in Finnish). Reports of the Finnish Environment Institute, Vol. 32Helsinki: Finnish Environment Institute 2013. Salvati, L., & Sabbi, A. (2011). Exploring long-term land cover changes in an urban region of southern Europe. The International Journal of Sustainable Development and World Ecology, 18(4), 273–282. https://doi.org/10.1080/13504509.2011.560453. Salvati, L., Sateriano, A., & Bajocco, S. (2013). To grow or to sprawl? Land cover relationships in a Mediterranean city region and implications for land use management. Cities, 30, 113–121. https://doi.org/10.1016/j.cities.2012.01.007. Schneider, A., & Woodcock, C. E. (2008). Compact, dispersed, fragmented, extensive? A comparison of urban growth in twenty-five global cities using remotely sensed data, pattern metrics and census information. Urban Studies, 45(3), 659–692. https://doi. org/10.1177/0042098007087340. Sikiö, M., Pitkänen, K., & Rehunen, A. (2014). Empty residential buildings in rural areas (in Finnish). Maaseudun Uusi Aika (2/2014)43–55. Skinner, M. W., Kuhn, R. G., & Joseph, A. E. (2001). Agricultural land protection in China: A case study of local governance in Zhejiang province. Land Use Policy, 18(4), 329–340. https://doi.org/10.1016/s0264-8377(01)00026-6. Söderström, P., Schulman, H., & Ristimäki, M. (2015). Urban form in the Helsinki and Stockholm city regions - development of pedestrian, public transport and car zones. Reports of the Finnish environment Institute, Vol. 16Helsinki: Finnish Environment Institute 2015. Solon, J. (2009). Spatial context of urbanization: Landscape pattern and changes between 1950 and 1990 in the Warsaw metropolitan area, Poland. Landscape and Urban Planning, 93(3–4), 250–261. https://doi.org/10.1016/j.landurbplan.2009.07.012. Subasinghe, S., Estoque, R. C., & Murayama, Y. (2016). Spatiotemporal analysis of urban

13