A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city

A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city

Accepted Manuscript Title: A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city Authors: Laura Gr...

1MB Sizes 0 Downloads 33 Views

Accepted Manuscript Title: A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city Authors: Laura Grunwald, Jannik Heusinger, Stephan Weber PII: DOI: Reference:

S1618-8667(16)30243-6 http://dx.doi.org/doi:10.1016/j.ufug.2017.01.001 UFUG 25832

To appear in: Received date: Revised date: Accepted date:

8-6-2016 14-11-2016 4-1-2017

Please cite this article as: Grunwald, Laura, Heusinger, Jannik, Weber, Stephan, A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city.Urban Forestry and Urban Greening http://dx.doi.org/10.1016/j.ufug.2017.01.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

A GIS-based mapping methodology of urban green roof ecosystem services applied to a Central European city

Authors: Laura Grunwald* Climatology and Environmental Meteorology Institute of Geoecology Technische Universität Braunschweig Langer Kamp 19c 38106 Braunschweig, Germany E-mail: [email protected] Phone: +49 (0)531-391-5616

Jannik Heusinger Climatology and Environmental Meteorology Institute of Geoecology Technische Universität Braunschweig Langer Kamp 19c 38106 Braunschweig, Germany E-mail: [email protected] Phone: +49 (0)531-391-5630

Stephan Weber Climatology and Environmental Meteorology Institute of Geoecology Technische Universität Braunschweig Langer Kamp 19c 38106 Braunschweig, Germany E-mail: [email protected] Phone: +49 (0)531-391-5607

* Corresponding author

2 Research highlights:   



Urban green roof ecosystem services (UESS) were assessed in a GIS environment 3 km2 suitable roof area for greening was found for Braunschweig, Germany UESS evaluation based on thermal climate, air quality, stormwater retention, biodiversity Feasible approach to define priority areas for subvention of urban green roof installation

Abstract Green roofs provide a number of different urban ecosystem services (UESS), e.g. regulation of microclimate, support of air quality improvement, or stormwater retention. To estimate the spatial variation of green roof UESS across an urban area, a GIS-based mapping and spatial analysis methodology was established and applied to the city of Braunschweig, Germany. Based on the analysis of available geodata, in a first step, a quantity of 14,138 rooftops in the study area (14 % of all buildings) was found to be generally suitable for greening. This resulted in a green roof area of 3 km2. Based on criteria such as roof slope and minimum roof size, nearly two-thirds of these buildings (8596 buildings, 8.6 % of total number of buildings) were categorised ‘appropriate’ for greening and subject to green roof UESS analysis. The spatial distribution of green roof UESS was estimated based on the categories thermal urban climate, air quality, stormwater retention and biodiversity. Due to their potential benefits in the four UESS categories an overall assessment resulted in a number of 867 roofs (0.9 % of total number of buildings) categorised as ‘high benefit’ from rooftop greening. Another 3,550 buildings (3.5 %) and 4,179 buildings (4.2 %) were defined as ‘moderate benefit’ and ‘low benefit’, respectively. The inner city area of Braunschweig appears as a hot-spot of green roof UESS, i.e. higher percentage of ‘high benefit’ green roofs in comparison to residential areas. The proposed method is a simple but straightforward approach to analyse urban green roof UESS and their spatial distribution across a city but it is sensitive to the quality of the available input geodata.

3

1. Introduction Given the ongoing trend of global urbanization and the impacts of climate change on cities, there is an increased awareness and perception of different positive effects of urban vegetation, e.g. as a local climate adaptation measure (Seto et al., 2011, Rosenzweig et al., 2011; Larsen, 2015). A way to assess positive aspects of urban vegetation is the framework of urban ecosystem services (UESS), i.e. the benefits the urban population receives from ecosystems. This concept is increasingly applied in scientific studies (e.g. Gómez-Baggethun et al., 2013, Luederitz et al., 2015). UESS define provisioning (e.g. food), regulating (climate), supporting (habitat) and cultural (recreation) services of ecosystems or of specific components of ecosystems, i.e. trees, parks or street greenery (Luederitz et al., 2015). Green roofs are one specific type of vegetated urban ecosystems (Berardi et al., 2014; Sutton, 2015). The construction of green roofs concerning number and surface area of green roofs has been globally increasing during recent years (e.g. Charpentier, 2015). As an example for Germany, a leader in green roof construction, it is assumed that about 8 million m² of green roof area are installed annually (FBB, 2015). Green roofs are composed as either extensive or intensive roof vegetation systems (cf. Oberndorfer et al., 2007 for a detailed review). While the former have shallow substrate depths (2 – 20 cm) and primarily are composed of drought-tolerant sedum vegetation and mosses which require little maintenance, the latter have deeper substrates (> 20 cm), are more diverse, not limited to specific plant types, and require regular maintenance and irrigation (Oberndorfer et al., 2007; Pfoser et al., 2014). The implementation of a green roof depends on statical characteristics and on roof slope. Generally, green roofs can be installed at slopes between 0 – 30° (FLL, 2008; cf. section 2.2). Green roof ecosystems are characterised to provide a range of UESS, e.g. microclimate regulation, air quality improvement, stormwater retention, habitat for flora and fauna, and aesthetic values (Oberndorfer et al., 2007). The benefits of green roof ecosystems have been intensively reviewed in scientific literature (e.g. Oberndorfer et al., 2007; Rowe, 2011; Sutton, 2015) and will only be briefly summarised at this point. One of the most recognised environmental benefits of green roofs is the capacity for (local) thermal regulation. A couple of studies report a significant decrease of surface and air temperature above green roofs in comparison to conventional bitumen roofs (Gaffin et al., 2009; Teemusk and Mander, 2010; Jim and Peng, 2012; Heusinger and Weber, 2015). Additionally, green roofs were studied for their potential to mitigate air pollution (Getter et al., 2009; Rowe, 2011; Speak et al., 2012), and to reduce rainwater runoff (DeNardo et al., 2005, VanWoert et al., 2005, Mentens et al., 2006, Yang et al., 2015). Furthermore, the positive impact for urban biodiversity, e.g. as additional habitat for different animal species, was studied by a couple of researchers (Francis and Lorimer, 2011; Cook-Patton and Bauerle, 2012). These benefits, most of which are also related to other types of urban green infrastructure such as parks, forests or community gardens (Coutts and Hahn, 2015), are of specific importance especially in dense built inner city areas where the implementation of additional

4 green is limited due to space constraints, space competition and regulative aspects. Green roofs, however, can be implemented on roof area already in existence. To date relatively little is known about the existing surface area of green roofs in different cities, about potential rooftop areas suitable for future greening, or the spatial variability of UESS provided by green roofs. To foster climate friendly urban planning strategies that benefit from the effects of urban rooftop greening, it is important to assess the status-quo and potential of green roofs. A recent application of a remote sensing approach analysis (combining infrared and visible light orthophotos with building models and ground plan maps) documents the existing green roof area in the German cities of Munich, Stuttgart and Karlsruhe to amount to 1.5 m2 per inhabitant on average (Ansel et al., 2015). Another remote sensing approach was used to assess the rooftop potential for photovoltaic system installation, green roof implementation and the environmental benefits from green roofs (e.g. carbon sequestration) in a test area in Thessaloniki, Greece (Mallinis et al., 2014, Karteris et al., 2016). The studies were based on high spatial resolution ortho-imagery, digital surface models and geospatial vector data. In this study we define green roof potential area (GRPA) as surface area that is suitable for rooftop greening given roof dimension and constructional measures. The motivation of the present study is to assess and map the spatial variation of GRPA and their related UESS using a GIS-based methodology. We argue that the potential benefit from rooftop greening is higher in certain areas of a city, the more human well-being or health is limited due to the impact of environmental stressors, e.g. increased levels of air pollution or heat load. Consequently, UESS of green roofs are related to the characteristics and spatial extent of different urban environmental stressors (e.g. heat stress, air pollution, degree of surface sealing). Four green roof UESS were taken into account: thermal urban climate (regulative UESS), air quality (regulative UESS), stormwater retention (regulative UESS) and biodiversity (supporting UESS).

2. Material and methods 2.1. Study area and available geodata Braunschweig, situated in Northern Germany (52°16‘28’N, 10°30‘38’E), is the second largest city of Lower Saxony with a population of 253,000. The total city area of Braunschweig amounts to 192 km² (Fig. 1). Buildings take a total plan area of 12.8 km², which makes 7 % of the urban area.

>>> Fig. 1 should be placed here <<<

5 The study was performed using different geodata sources in a GIS environment (ArcMap Version 10, Software ArcGIS, ESRI). The geodata was available from the Environmental Agency of the city administration of Braunschweig (Fig. 2). The data basis consisted of a) a digital elevation model from airborne laser scanning with 2 m resolution and a height accuracy of 0.15 m including buildings and vegetation which was provided as point vector data (generated in 2003), b) a land use map which was provided as polygon vector data consisting of defined land use types based on a biotope type mapping from 2010, c) a building ground plan of Braunschweig which was provided as polygon vector data (generated in 2010), d) a traffic count map giving the annual average daily traffic intensity (AADT) of the urban road network as a projection for 2015 which was provided as line vector data of the major roads (generated in 2012), and e) a climate function map of Braunschweig generated in 2012 (Steinicke et al., 2012) which was provided as polygon vector data.

>>> Fig. 2 should be placed here <<<

2.2. Mapping GRPA In this study different applications of ArcGIS were used for mapping and spatial analysis, which will briefly be described in the following. GRPA were evaluated by considering the building ground plan and the digital elevation model of Braunschweig. Based on the elevation model, Triangulated Irregular Networks (TIN) were generated using ArcEsri’s 3D Analyst tools to represent surface morphology and calculate roof slopes. TINs represent the surface as a set of contiguous, non-overlapping triangular facets (Peucker et al, 1978). The point input features of the DEM were connected with a series of edges to form a network of triangles. For that purpose interpolation is needed, which in ArcGIS is done by the Delaunay triangulation method. The TIN method allows to preserve the precision of the input data, since the input features remain in the same position as the nodes and edges of the triangular facets. The resulting roof slopes were assigned to slope classes: (A) <1°, (B) 1° to <5°, and (C) >5°. The classes were defined based on green roof constructive and technical measures that need to be considered for different roof slopes (FLL, 2008). Roof coverings such as tiled roofs are generally not suitable for greening since specific constructional measures would be required (FLL, 2008). To prevent tiled roofs from being classified as suitable, only buildings in classes A and B were accepted, since most tiled roofs

6 were classified into class C. Buildings which did not meet this criteria were classified ‘not appropriate’. Due to roof obstructions such as chimneys, antennas, staircase and elevator shafts the calculated slopes can differ. A homogenous setup of the roof without many obstructions is preferred for greening (cf. Karteris et al., 2016). Hence, we termed buildings that have >=75 % of their roof area in slope classes A and/or B as ‘appropriate’. Building roof areas that fell into classes A or B with a percentage share <75 % were categorised ‘limited appropriate’. Furthermore, a minimum roof size of 10 m² for a specific building was defined. In the seven-year difference of publication of the digital elevation model (2003) and the building ground plan (2010) several new housing estates were developed and completed in Braunschweig. Hence, the building ground plan may list buildings while the elevation model still indicates undeveloped flat area, i.e. in recently developed city districts like ‘Broitzem’, ‘Mascherode’, ‘Lamme’ and ‘Volkmarode’ (Fig. 1). To prevent misclassification these buildings were excluded and assigned to the category ‘not appropriate’.

2.3. Mapping green roof UESS The overall assessment of green roof UESS was quantified neither by measurements nor by modelling, but was estimated by a ranking order based on the spatial variation of the four UESS categories under study. As to date, no data on the status-quo of existing green roof area is available for the study area. Thus, the present work considers potential areas, which are suitable for future greening. Our analysis will result in an overall assessment of green roof benefit to define priority green roofs (PRIOGR) that show the highest effect in terms of UESS. PRIOGR is a dimensionless, qualitative indicator based on the benefit in the green roof UESS categories (cf. section 2.3.5). Whereas mapping of GRPA considers both categories 'limited appropriate' and 'appropriate' (cf. section 2.2), the evaluation of PRIOGR is restricted to 'appropriate' roofs. This is done to confine the UESS analysis to the environmentally most relevant roofs, i.e. ‘appropriate’ roof areas.

2.3.1 Thermal urban climate The assessment of PRIOGR concerning thermal urban climate, e.g. heat stress reduction, is based on the climatope classification for residential areas (i.e. areas with similar microclimatic characteristics, e.g. Scherer et al., 1999), taken from the climate function map of Braunschweig. The classification into climatopes is defined by dominating land use and results of model calculations (i.e. air temperature and air humidity) available from a climate analysis of Braunschweig (Steinicke et al., 2012).

7 Built-up areas were classified into the climatopes inner city climate (high degree of sealed surface types, small green space density and high building density), city climate (moderate degree of sealed surface types, insufficient green space density, partly dense building development) and residential climate (low degree of sealed surface types, high green space density, low building density). The inner city climate is characterised by strong thermal stress conditions, the city climate by moderate and the settlement climate by low thermal stress (Steinicke et al., 2012). We defined three assessment categories for the different green roof UESS to evaluate benefits by roof greening: high benefit, moderate benefit and low benefit (Table 1). For the category thermal urban climate green roofs were rated with ‘high benefit’ in areas of high thermal stress, since highest green roof benefits due to evaporative cooling and modified albedo of green roofs can be expected. Buildings not located in any of the residential areas were labelled N/A (not available), which applied to 1517 buildings in the study area.

>>> Table 1 should be placed here <<<

2.3.2 Urban air quality Urban air quality is closely linked to gaseous and particulate pollutant emission from anthropogenic activity. The most important urban air pollutant sources are combustion emissions from road traffic, industry and district heating. Areas of high road traffic activity experience high levels of particulate and gaseous air pollution (e.g. Weber et al., 2013, von Bismarck-Osten et al., 2013). Green roofs, however, can function as a sink for pollutants through deposition on vegetation (Rowe, 2011, Speak et al., 2012, Janhäll, 2015). To assess PRIOGR in context of urban air quality the AADT values for 2015 were used. Other stationary or mobile pollutant emission data (industry, construction machinery) were not available for this study. A grid cell pattern with a horizontal cell dimension of 100 x 100 m was established, covering the whole city. The AADT values of the line input features were classified to the grid cells via spatial overlap and every cell is characterised by the sum of the AADT values from the streets within the respective cell (Table 1). Afterwards the “appropriate” buildings feature (polygon input) and the AADT grid cells were combined by spatial overlap.

2.3.3 Urban water retention Due to the ability of green roofs to considerably reduce rainwater runoff and to delay runoff peaks in comparison to conventional roofs (Mentens et al., 2006; Berndtsson, 2010) we assume a larger benefit of green roofs in areas with a high degree of surface sealing than in less sealed areas of the city. This, for instance, reduces flooding and the risk of sewage discharge into rivers after heavy precipitation

8 events (Bliss et al., 2009; Berndtsson, 2010). The evaluation of PRIOGR to enhance urban retention is based on the land use map of Braunschweig. Different degrees of sealing were assigned to the land use types according to Steinicke et al. (2012, Table 2). The degrees of sealing represent typical mean values (Table 1). Buildings not located in any of these land use types were labelled N/A (not available), which applied to 24 buildings in the study area.

>>> Table 2 should be placed here <<<

2.3.4 Urban biodiversity The assessment of the efficiency of green roofs to increase urban biodiversity is complex. The habitat prevalence of plant and animal species is variable as is the construction, layout and species mixture of urban green roofs (i.e. Brenneisen, 2006; Fernandez-Canero and Gonzalez-Redondo, 2010). Generally, larger green roofs might offer more design possibilities and food sources, which has a positive effect on biodiversity (Hui and Chan, 2011; Williams et al., 2014). In addition, the distance between green roofs and natural green areas in the urban area is important. The closer a green roof is located to a natural green area, the higher is the chance that different species and even low-mobility species will use them as additional habitat (Hui and Chan, 2011; Williams et al., 2014). These two factors, the size of the roof area suitable for greening and the distance to green spaces (i.e. allotments, meadows, agricultural areas, wood and forest) were defined to evaluate PRIOGR concerning their ability to enhance urban biodiversity. The evaluation of PRIOGR to enhance urban biodiversity is based on the land use map of Braunschweig and the area of ‘appropriate’ buildings. For each ‘appropriate’ building the total suitable greening area was divided in two classes: suitable greening area >100 m² and ≤100 m². To consider distance, green spaces were extracted from the land use map and a buffer was used which took into account all buildings within a 50 m distance. Appropriate buildings were classified into two classes: distance between building and green space ≤50 m and >50 m (Table 1).

2.3.5 Overall assessment of PRIOGR The results of the green roof UESS categories were integrated in an overall assessment of PRIOGR. The UESS categories were assessed with grades from 1 (high benefit) to 3 (low benefit). A total sum was calculated for every ‘appropriate’ building based on the four UESS grades. As a result, every building was assigned with a number between 3 and 12 to define PRIOGR. Finally, these values were ranked in three overall green roof benefit categories ‘high benefit’  3 to 5, ‘moderate benefit’  6 to 8 and ‘low benefit’  9 to 12.

9

3. Results 3.1 Spatial variation of GRPA The classification of GRPA resulted in a total horizontal roof area of 5.7 km² (14,138 buildings). This is equal to 3 % of the city area, and about 14 % of the total number of buildings in Braunschweig (n = 100,635, Fig. 3). GRPA were matched to the building plan area in GIS (Fig. 4). The area of buildings classified as ‘limited appropriate’ is 3.1 km² (5.5 % of buildings, 1.6 % of city area) while ‘appropriate’ roofs account for 2.6 km² (8.6 % of buildings, 1.4 % of city area).

>>> Fig. 3 should be placed here <<<

>>> Fig. 4 should be placed here <<<

The GRPA of classified roofs across the study area varies between 10 and 49,000 m2 with an average roof size of 404 m2 (Fig. 5a). However, only for a small fraction of buildings (3.9 %) the complete roof area is suitable for greening. For the majority of roofs the suitable area for greening is limited by roof obstructions (cf. section 2.2). Between 30 - 60 % of the roof area of most buildings is suitable for greening with an average of 44 % (Fig. 5b). Based on this roof area distribution the space that is actually suitable for greening results in 3 km2, evenly divided into 1.5 km2 in the categories ‘limited appropriate” and ‘appropriate”.

>>> Fig. 5 should be placed here <<<

In the overall assessment, the inner city area (densely built) indicates significant potential for green roofing. Both, ‘limited appropriate’ and ‘appropriate” buildings occur. Other residential districts, e.g. ‘Weststadt’, ‘Kanzlerfeld’, ‘Mascherode’ and the industrialized area ‘Veltenhof’ are also characterized by a high potential for urban green roofs (Fig 6a). In the following analysis of green roof UESS we will, however, not refer to the size of the roof area available for greening but rather to the percentage number of ‘appropriate’ roofs (normalized by the total number of buildings within the study area).

10 3.2 Spatial variation of green roof UESS Thermal urban climate As the city centre of Braunschweig is characterized by the highest probability for heat stress (climatope inner city climate, cf. section 2.3.4) all ‘appropriate’ buildings were classified with ‘high benefit’. Altogether 1.3 % of all buildings in Braunschweig were assigned to this category (Fig. 6b). Another 3.1 % of buildings were classified as ‘moderate benefit’, due to the location within the climatope city climate. In suburban areas buildings are classified as ‘low benefit’ (2.7 %).

>>> Fig. 6 should be placed here <<<

Urban air quality The PRIOGR due to air quality aspects strongly depend on the location of highly frequented main roads. Hence, buildings classified as ‘high benefit’ (1.8 %) are mainly located in the inner city area but also in short distance to highly frequented roads in suburban areas (Fig. 6c). 1.7 % of roofs in Braunschweig were classified as ‘moderate benefit’, another 5.1 % as ‘low benefit”.

Urban water retention A share of 2.7 % of the buildings in the study area were classified as ‘high benefit’ in terms of their probability to enhance urban retention. In Braunschweig the main fraction of buildings is located within densely built-up area of the city centre, the districts ‘Weststadt’ and ‘Gliesmarode’, or the industry sector of ‘Veltenhof’ (Fig. 6d). An additional share of 4.7 % of the buildings are characterised by ‘moderate benefit’. Buildings in that category are mainly located in the less built-up areas of the districts ‘Kanzlerfeld’, ‘Mascherode’ and other suburban areas. In the more rural areas of the city, characterized by a lower degree of surface sealing, most of the buildings show ‘low benefit’ (1.1 % in total).

Urban biodiversity The assessment of urban biodiversity indicates 1.3 % of buildings to have ‘high benefit’. These are mainly located in suburban areas in close vicinity to green spaces (Fig. 6e). In the industrially shaped district ‘Veltenhof’ situated in the NE of the city area, or in suburban areas ‘Weststadt’ and ‘Broitzem’ many buildings are classified into this category. The inner city area results in a more heterogeneous

11 distribution. Most buildings are classified as ‘moderate benefit’ and ‘low benefit’ with a percentage of 4.3 % and 2.9 %, respectively.

3.3 Overall assessment of PRIOGR Considering the four selected green roof UESS the overall assessment demonstrates 867 (0.9 %) of all buildings to have ‘high benefit’. PRIOGR are mainly located in the city centre, in the western suburban areas of the city and the industry sector ‘Veltenhof’ (Fig. 6f). A total of 3.5 % of buildings characterised by a homogeneous spatial distribution within the city boundaries were ranked as ‘moderate benefit’. In the suburban districts, most buildings show ‘low benefit’ (4.2 %). Table 3 summarizes the ranking of ‘appropriate’ buildings concerning the four UESS and the overall assessment.

>>> Table 3 should be placed here <<<

4. Discussion 4.1 Methodological aspects The present study proposes a geospatial analysis of green roof UESS in a GIS environment based. We did not perform specific measurements or modelling to evaluate the impact of green roofs within the study area. The present approach does not surpass experimental evidence or model calculations in terms of quality, accuracy and spatial resolution (cf. section 2). However, to date no adequate urban data basis is available that allows for reasonable extrapolation of green roof benefits from explicit measurements on green roofs. In addition, only few studies are available that quantify urban-scale green roof benefits on thermal climate, air quality, stormwater retention and biodiversity dynamics (i.e. Karteris et al., 2016; Sun et al., 2016). The present approach is a straightforward method to obtain an overview of the spatial variation of green roof UESS. This enables urban planners to determine areas where the positive impacts of green roofs are most effective. However, some restrictions need to be considered in the process of mapping GRPA. Chimneys and other obstructions on rooftops can result in TIN inaccuracies, given the relatively coarse resolution (2 m) of the elevation model (cf. section 2.2). Furthermore, static properties of roofs could not be considered in the evaluation process due to lack of data. However, static aspects are essential in green roof planning processes. Depending on green roof type, additional mass loadings of 20 – 190 kg m-² for extensive green roofs or up to 680 kg m-² for intensive green roofs must be considered (Pfoser et al., 2014).

12 Additionally, the method and the accuracy of the results depends on the quality of available geodata. However, the general concept is transferable. The geodata used in this study is sufficient to generate a first estimate for PRIOGR. With additional data and higher resolution of the input geodata even more accurate results could be achieved.

4.2 Green roof UESS Airborne dispersion of pollutants is complex and a function of the type of pollutant, emission strength and atmospheric conditions (i.e. Jerrett et al., 2005). In the present study, the estimation is based on spatial data intersection of traffic load and PRIOGR areas whereas the actual distribution of pollutants is not considered explicitly. In order to gain better estimates, data on the urban pollutant concentrations are needed. However, these can only be provided by dense networks of measuring stations or model simulations. Land use regression modelling that is increasingly applied in air pollution modelling might be one future possibility (e.g. Hoek et al., 2008, Ghassoun et al., 2015). Moreover, the present analysis is based solely on air pollutants from road transport and no further information about urban background pollution was included. The impact of green roofs in reducing street-level pollutant concentrations is limited. This is due to small air exchange rates between canopy layer and urban boundary layer (Pugh, 2012). Green roofs might provide more benefit in areas of lower building density or near open urban spaces, due to higher air exchange rates. Janhäll (2015) stated that vegetation close to emission source increases deposition capability because of higher pollutant concentration gradients. Taking this under consideration, building height could be a further criterion depending on the location of the source. The criteria applied to estimate potential roofs in terms of an increase of urban biodiversity in this study are a first attempt to attain a general assessment. Adding green roof height might be an additional asset to refine our assessment. According to Williams et al. (2014) green roofs on lower buildings are favoured by different animal species. Furthermore, the connection of green roofs (corridor) could be considered, since proximity of green roofs results in increased movement of individuals (Braaker et al., 2014). Depending on the studied species, additional criteria could be implemented since every species is characterised by specific source-sink metapopulation dynamics (Williams et al., 2014).

5. Summary and conclusions The present study applies a GIS-based mapping methodology to evaluate the spatial variation of urban green roof ecosystem services and priority green roofs in Braunschweig, Germany. Four different

13 categories of green roof UESS were taken into account, namely urban thermal climate, air quality, stormwater retention and biodiversity. Our analysis concludes that Braunschweig offers a GRPA of 3 km² on 14 % of the total number of buildings. A number of 8596 buildings (8.6 % of total number of buildings) were classified due to their benefits in the four UESS categories. However, the spatial distribution of green roof benefits (high, moderate, and low benefit) is dependent on the spatial distribution of suitable roof areas and environmental stressors. The inner city area of Braunschweig appears as a hot-spot of green roof potential, i.e. higher percentage of benefits in comparison to residential areas. This is of importance since one main asset of green roofs, in contrast to other types of green infrastructure, is that they can be implemented on roof area already in existence. Green roofs introduce additional vegetated areas within densely built cities where the impact of environmental stressors is usually high. The proposed approach is an operational method to map the potential of green roof UESS. The method is straightforward, simple to apply in a GIS environment, and depends on the quality of the available geodata. However, due to the lack of adequate urban-scale spatially resolved measurement or modelling data from green roofs the present approach is thought to give a reasonable overview of spatial differences of green roof UESS. The proposed methodology may be used by municipalities to define priority areas for promotion or subvention of green roof installation in urban areas.

Acknowledgements The geodata used in this study was kindly provided by the Environmental Protection Department of the city administration of Braunschweig, Germany. We thank Dipl.-Ing. M. Sc. Thomas Gekeler and Dipl.-Geogr. Andreas Bruchmann (Environmental Agency, City of Braunschweig) for providing the data. This study was partly funded in the research project “Analysis of the potential of green roofs for climate regulation in Braunschweig” funded by the city administration of Braunschweig.

14

6. References Ansel, W., Zeidler, J., and Esch, T. (2015). Fernerkundliche Identifizierung von Vegetationsflächen auf Dächern zur Entwicklung des für die Bereiche des Stadtklimas, der Stadtentwässerung und des Artenschutzes aktivierbaren Flächenpotenzials in den Städten. Abschlussbericht Az 30299, Deutsche Bundesstiftung Umwelt, 51 pp. Berardi, U., GhaffarianHoseini, A., and GhaffarianHoseini, A. (2014). State-of-the-art analysis of the environmental benefits of green roofs. Applied Energy, 115, 411-428. Bliss, D. J., Neufeld, R. D., and Ries, R. J. (2009). Storm water runoff mitigation using a green roof. Environmental Engineering Science, 26(2), 407-418. Berndtsson, J. C. (2010). Green roof performance towards management of runoff water quantity and quality: A review. Ecological Engineering, 36(4), 351-360. Braaker, S., Ghazoul, J., Obrist, M. K., and Moretti, M. (2014). Habitat connectivity shapes urban arthropod communities: the key role of green roofs. Ecology, 95(4), 1010-1021. Brenneisen, S. (2006). Space for Urban Wildlife: Designing Green Roofs as Habitats in Switzerland. Urban Habitats, 4(1), 27-36. Charpentier, S. (2015). Simulation of Water Regime and Sensible Heat Exchange Phenomena in Green Roof Substrates. Vadose Zone Journal, 14(6). Cook-Patton, S.C., and Bauerle, T.L. (2012). Potential benefits of plant diversity on vegetated roofs: A literature review. Journal of Environmental Management, 106, 85-92. Coutts, C., Hahn, M., 2015. Green Infrastructure, Ecosystem Services, and Human Health. International Journal of Environmental Research and Public Health 12. DeNardo, J. C., Jarrett, A. R., Manbeck, H. B., Beattie, D. J., Berghage, R.D. (2005). Stormwater mitigation and surface temperature reduction by green roofs. Trans. ASAE, 48(4), 1491–1496. FBB (2015). Bundesweite Strategie Gebäudegrün, Fachvereinigung Bauwerksbegrünung e.V. (FBB), Saarbrücken, 6 pp. FLL (2008). Richtlinien für die Planung, Ausführung und Pflege von Dachbegrünungen. Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau e.V. (FLL), Bonn, p. 118. Francis, R.A., and Lorimer, J. (2011). Urban reconciliation ecology: the potential of living roofs and walls. Journal of Environmental Management, 92, 1429-1437. Gaffin, S.R., Khanbilvardi, and R., Rosenzweig, C. (2009). Development of a green roof environmental monitoring and meteorological network in New York City. Sensors (Basel, Switzerland), 9(4), 2647– 2660. Getter, K. L., Rowe, D. B., Robertson, G. P., Cregg, B. M., and Andresen, J. A. (2009). Carbon sequestration potential of extensive green roofs. Environmental science and technology, 43(19), 7564– 7570.

15 Ghassoun, Y., Ruths, M., Löwner, M.O., and Weber, S. (2015). Intra-urban variation of ultrafine particles as evaluated by process related land use and pollutant driven regression modelling. Science of the Total Environment, 536, 150-160. Gómez-Baggethun, E., Gren, Å., Barton, D., Langemeyer, J., McPhearson, T., O’Farrell, P., Andersson, E., Hamstead, Z., and Kremer, P. (2013). Urban Ecosystem Services. In T. Elmqvist, M. Fragkias, J. Goodness, B. Güneralp, P. J. Marcotullio, R. I. McDonald, S. Parnell, M. Schewenius, M. Sendstad, K. C. Seto, and C. Wilkinson (Eds.), Urbanization, Biodiversity and Ecosystem Services: Challenges and Opportunities (pp. 175-251). Springer Netherlands. Heusinger, J., and Weber, S. (2015). Comparative microclimate and dewfall measurements at an urban green roof versus bitumen roof. Building and Environment, 92, 713-723. Hoek, G., Beelen, R., de Hoogh, K., Vienneau, D., Gulliver, J., Fischer, P., and Briggs, D. (2008). A review of land-use regression models to assess spatial variation of outdoor air pollution. Atmospheric Environment, 42, 7561-7578. Hui, S. C., and Chan, M. K. (2011). Biodiversity assessment of green roofs for green building design. In Proceedings of Joint Symposium, 1-11. Janhäll, S. (2015). Review on urban vegetation and particle air pollution–deposition and dispersion. Atmospheric Environment, 105, 130–137. Jerrett, M., Arain, A., Kanaroglou, P., Beckerman, B., Potoglou, D., Sahsuvaroglu, T., Morrison, J., and Giovis, C. (2005). A review and evaluation of intraurban air pollution exposure models. Journal of Exposure Science and Environmental Epidemiology, 15(2), 185-204. Jim, C.Y., and Peng, L. L. H. (2012). Weather effect on thermal and energy performance of an extensive tropical green roof. Urban Forestry and Urban Greening, 11(1), 73–85. Karteris, M., Theodoridou, I., Mallinis, G., Tsiros, E., and Karteris, A. (2016). Towards a green sustainable strategy for Mediterranean cities: Assessing the benefits of large-scale green roofs implementation in Thessaloniki, Northern Greece, using environmental modelling, GIS and very high spatial resolution remote sensing data. Renewable and Sustainable Energy Reviews, 58, 510-525. Larsen, L. (2015). Urban climate and adaptation strategies. Frontiers in Ecology and the Environment, 13(9), 486-492. Luederitz, C., Brink, E., Gralla, F., Hermelingmeier, V., Meyer, M., Niven, L., Panzer, L., Partelow, S., Rau, A.-L., Sasaki, R., Abson, D. J., Lang, D.J., Wamsler, C., and von Wehrden, H. (2015). A review of urban ecosystem services: six key challenges for future research. Ecosystem Services, 14, 98-112. Mallinis, G., Karteris, M., Theodoridou, I., Tsioukas, V., and Karteris, M. (2014). Development of a nationwide approach for large scale estimation of green roof retrofitting areas and roof-top solar energy potential using VHR natural colour orthoimagery and DSM data over Thessaloniki, Greece. Development, 5(6), 548-557. Mentens, J., Raes, D., and Hermy, M. (2006). Green roofs as a tool for solving the rainwater runoff problem in the urbanized 21st century? Landscape and Urban Planning, 77(3), 217–226. Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R.R., Doshi, H., Dunnett, N., Gaffin, S., Köhler, M., Liu, K.K.Y., and Rowe, B. (2007). Green Roofs as Urban Ecosystems: Ecological Structures, Functions, and Services. BioScience, 57, 823-833.

16

Peucker, T. K., Fowler, R. J., Little, J. J., & Mark, D. M. (1978). The triangulated irregular network. In Amer. Soc. Photogrammetry Proc. Digital Terrain Models Symposium (Vol. 516, p. 532). Pfoser, N., Henrich, J., Jenner, N., Schreiner, J., Kanashiro, C., Heusinger, J., Weber, S., Hegger, M., and Dettmar, J. (2014). Gebäude Begrünung Energie - Potenziale und Wechselwirkungen. Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau e.V. (FLL); Bundesinstitut für Bau-, Stadt- und Raumforschung (BBSR), FLL-Schriftenreihe, FV2014/01, Bonn, p. 305. Pugh, T. A., MacKenzie, A. R., Whyatt, J. D., and Hewitt, C. N. (2012). Effectiveness of green infrastructure for improvement of air quality in urban street canyons. Environmental science and technology, 46(14), 7692–7699. Rosenzweig, C., Solecki, W. D., Hammer, S. A., and Mehrotra, S. (2011). Climate Change and Cities: First Assessment Report of the Urban Climate Change Research Network. Cambridge University Press., Cambridge, 286 pp. Rowe, D. B. (2011). Green roofs as a means of pollution abatement. Environmental pollution, 159(8-9), 2100–2110. Scherer, D., Fehrenbach, U., Beha, H. D., and Parlow, E. (1999). Improved concepts and methods in analysis and evaluation of the urban climate for optimizing urban planning processes. Atmospheric Environment, 33, 4185-4193. Seto, K. C., Fragkias, M., Güneralp, B., and Reilly, M. K. (2011). A Meta-Analysis of Global Urban Land Expansion. PLoS ONE, 6, e23777. Speak, A. F., Rothwell, J. J., Lindley, S. J., and Smith, C. L. (2012). Urban particulate pollution reduction by four species of green roof vegetation in a UK city. Atmospheric Environment, 61, 283–293. Steinicke, W., Schwab, U., Röckle, R., Richter, C.-J., and Nielinger, J. (2012). Stadtklimaanalyse Braunschweig. On behalf of the Environmental Protection Department of the city administration of Braunschweig. Freiburg, p. 103. Sun, T., Grimmond, C.S.B., Ni, G.-H. (2016). How do green roofs mitigate urban thermal stress under heat waves? Journal of Geophysical Research: Atmospheres, 121, 5320-5335. Sutton, R. K. (2015). Introduction to Green Roof Ecosystems. In Green Roof Ecosystems (pp. 1-25). Springer International Publishing. Teemusk, A., and Mander, Ü. (2010). Temperature regime of planted roofs compared with conventional roofing systems. Ecological Engineering, 36, 91-95. VanWoert, N. D., Rowe, D. B., Andresen, J. A., Rugh, C. L., Fernandez, R. T., and Xiao, L. (2005). Green roof stormwater retention: effects of roof surface, slope, and media depth. Journal of Environmental Quality, 34(3), 1036–1044. von Bismarck-Osten, C., Birmili, W., Ketzel, M., Massling, A., Petäjä, T., and Weber, S. (2013). Characterization of parameters influencing the spatio-temporal variability of urban aerosol particle number size distributions in four European cities. Atmospheric Environment, 77, 415-429.

17 Weber, S., Kordowski, K., and Kuttler, W. (2013). Variability of particle number concentration and particle size dynamics in an urban street canyon under different meteorological conditions. Science of the Total Environment, 449, 102-114. Williams, N. S., Lundholm, J., and MacIvor, J. S. (2014). Do green roofs help urban biodiversity conservation? Journal of Applied Ecology, 51(6), 1643-1649. Yang, W.-Y., Li, D., Sun, T. and Ni, G.-H. (2015). Saturation-excess and infiltration-excess runoff on green roofs. Ecological Engineering, 74, 327-336.

18 Figure captions Fig. 1. Map of Braunschweig with city districts (black lines) (map source: Stadt Braunschweig - Open GeoData, 2016, Lizenz: dl-de/by-2-0, modified).

19 Fig. 2. Overview of the used geodata: a) Land use map of Braunschweig, b) part of the building ground plan (grey polygons), c) traffic count map with AADT data of the major roads of Braunschweig and d) climate function map with climatopes of Braunschweig (according to Steinicke et al., 2012), highlighted in color are the used climatopes in the method.

20 Fig. 3. Spatial distribution of GRPA classified as ‘appropriate’ and ‘limited appropriate’ (14,138 buildings with a total rooftop area of 5.7 km², green polygons) in Braunschweig.

21 Fig. 4. Example of GRPA (‘appropriate’ and ‘limited appropriate’) in the inner city area of Braunschweig.

22 Fig. 5. Frequency distribution of (a) GRPA binned into different size classes and (b) percentage of actual roof area that is suitable for greening due to limitations by roof constructions. Data are binned into 10 % classes. The Figure depicts buildings classified as ‘appropriate” and ‘limited appropriate”.

23 Fig. 6. Spatial distribution of GRPA in the city of Braunschweig (based on the percentage share of buildings in relation to the total amount of buildings). The figure depicts the (a) the spatial distribution of all ‘appropriate” buildings for green roofing and the roofs classified with ‘high benefit” in the UESS categories (b) thermal urban climate, (c) air quality, (d) retention potential, (e) biodiversity and (f) the overall assessment. The grid pattern has a horizontal resolution of 500 x 500 m. Grid cells without appropriate green roofs are not shown.

24 Table captions Table 1. Green roof UESS based on four categories: thermal climate, air quality, water retention and biodiversity. Table 1 Green roof UESS

Thermal climate

Air quality

Urban retention

Urban biodiversity

High benefit

Strong thermal stress in the builtup area

AADT > 5000 vehicles/ day

Degree of surface sealing > 50 %

Potential area > 100 m² and distance ≤ 50 m

Moderate benefit

Moderate thermal stress in the builtup area

0 < AADT ≤ 5000 vehicles/ day

0 % < Degree of surface sealing ≤ 50 %

Potential area ≤ 100 m² and distance ≤ 50 m OR Potential area > 100 m² and distance > 50 m

Low benefit

Low thermal stress in the built-up area

AADT = 0 vehicles/ day

Degree of surface sealing = 0 %

Potential area ≤ 100 m² and distance > 50 m

25 Table 2. Categories for different land use types in Braunschweig based on the degree of surface sealing (modified after Steinicke et al., 2012). Table 2 Green roof UESS

High benefit

Moderate benefit

Land use type

Degree of sealing

Centre development, district centre

95%

Asphalt surfaces, parking lots

95%

Industrial and commercial area

87%

Block development

78%

Linear and high-rise development

55%

Detached and terraced houses

41%

Rail traffic areas

25%

Allotments, sport fields and the like

10%

Meadows, agricultural area

0%

Wood, garden, fruit growing

0%

Forest

0%

Water

0%

Low benefit

26 Table 3. Final PRIOGR grouped by the green roof UESS categories. The table also gives the overall assessment. The percentage value estimates number of buildings classified to each category in relation to the total number of buildings in Braunschweig (absolute number of building in brackets).

Table 3 Green UESS

roof

Thermal climate

Air quality

Retention potential

Biodiversity

Overall assessment PRIOGR

High benefit

1.3 % (1258)

1.8 % (1779)

2.7 % (2742)

1.3 % (1313)

0.9 % (867)

Moderate benefit

3.1 % (3073)

1.7 % (1660)

4.7 % (4716)

4.3 % (4340)

3.5 % (3550)

Low benefit

2.7 % (2748)

5.1 % (5157)

1.1 % (1114)

2.9 % (2943)

4.2 % (4179)