Urban Fabric Types and Microclimate Response Assessment...

Urban Fabric Types and Microclimate Response ‐

Assessment and Design Improvement.

Final Report R. Stiles, B. Gasienica‐Wawrytko, K. Hagen, H. Trimmel, W. Loibl, M. Köstl, T. Tötzer,

S. Pauleit, A. Schirmann, W. Feilmayr

SUMMARY REPORT

Vienna, April 2014

A project within the program

ACRP 3rd Call

Climate and Energy Fund of the Federal State – managed by Kommunalkredit Public Consulting GmbH

CONTENTS

1 SUMMARY REPORT ............................................................................................................................................. 1

1.1 Project data .......................................................................................................................................................... 1

1.2 Abstract ................................................................................................................................................................ 2

1.3 Motivation and objectives .................................................................................................................................... 3

1.4 Work Packages ..................................................................................................................................................... 4 1.4.1 WP 2 – Generation of urban fabric typologies .............................................................................................. 4 1.4.2 WP 3 -‐ Characterisation of urban fabric types and identification of open space typologies ........................ 8 1.4.3 WP 4 -‐ Investigation of interactions between urban open space design and microclimate ....................... 13 1.4.4 WP 5 -‐ Application of findings to the urban fabric and open space typologies .......................................... 14 1.4.5 WP 6 -‐ Recommendations of planning and design measures ..................................................................... 20 1.4.6 WP 7 -‐ Dissemination and public discussion ............................................................................................... 22

1.5 Conclusions ........................................................................................................................................................ 22

1.6 Discussion and outlook ...................................................................................................................................... 24

Urban fabric types and microclimate response – assessment and design improvement

Summary Report 1

1 SUMMARY REPORT

The short summary is based on the official Final Report for the ACRP submitted in April 2014.

The full-‐length report (in german) is available for download on the project website www.urbanfabric.ac.at.

1.1 Project data

Title:

Urban Fabric and Microclimate Response – Assessment and Design Improvement (Urban Fabric + Microclimate)

Programme:

ACRP 3rd Call, Thematic Area 4: Understanding the climate system and consequences of climate change Klima-‐ und Energiefonds, Bundesministerium für Innovation, Verkehr und Technik (bmvit)

Project start and duration:

1.5.2011 – 20.1.2014 (33 months)

Coordination of project:

Vienna Technical University, Department of Landscape Architecture (Team: Prof. DI Richard Stiles, Dr. Katrin Hagen, DI Heidelinde Trimmel, DI Beatrix Gasienica-‐Wawrytko)

Project partners:

Austrian Institute of Technology GmbH; Energy Department (Team: Dr. Wolfgang Loibl, Dr. Tanja Tötzer, Mag. Mario Köstl)

TU München; Strategic Landscape Planning and Management (Team: Prof. Dr. Stephan Pauleit, DI Annike Schirmann)

Scientific advisory board:

The project has been accompanied by a scientific advisory board. Members: J-‐Prof. Dr. Fazia Ali-‐Toudert (TU Dortmund); Dr. Maria Balas (Umweltbundesamt); Prof. Dr. Christiane Brandenburg (BOKU Wien); Prof. Dr. Jürgen Breuste (Universität Salzburg); Prof. Dr. Michael Bruse (Johannes Gutenberg Universität Mainz); DI Jürgen Preiss (Wiener Umweltschutzabteilung MA22); Prof. Dr. Erich Mursch-‐Radlgruber (BOKU Wien); DI Thomas Proksch (Planungsbüro Land in Sicht, Wien); DI Eva Prochazka (Stadtentwicklung und Stadtplanung MA18); Dr. Maja Zuvela-‐Aloise (Zentralanstalt für Meteorologie und Geodynamik).


Summary Report 2

1.2 Abstract

Urban areas are particularly vulnerable to the impact of climate change; they are also the living environment of choice for a significant majority of Europe‘s population. Global warming has an increasing influence on the urban climate and will consequently affect the future health and well-‐being of a large proportion of the people of Europe. The aim of this project was to understand better the way in which the small scale structure of the urban fabric contributes different to the urban heat island effect and other urban climate phenomena. The results were used to develop specific strategies for counter-‐acting and mitigating these effects at a local level. A major focus of the project was on characterising the morphology of the urban landscape and on understanding the interaction between different types of urban fabric and the urban microclimate. The aim was to identify the different climate sensitivity of the urban fabric – using the example of Vienna -‐ and to investigate concrete design measures aimed at modifying open spaces in order to counteract the effects of overheating during hot summer days.

Using 500x500m square quadrants based on grid applied by Statistik Austria an urban fabric typology for the city of Vienna has been generated. Data sets were compiled describing key aspects of the urban climate and urban morphology in terms of terrain, open space and built structure, all of which influence the microclimatic conditions and parameters. A „two-‐step“ cluster analysis was then carried out, which resulted in the identification of 9 urban fabric types, that closely reflected the varying morphology of Vienna’s entire urban landscape. Subsequently, further investigations concentrated on the three urban landscape classes which were considered most climate-‐sensitive, namely those within the dense inner city and in the main urban development areas northeast of the Danube. These were then subjected to a further cluster analysis in order to increase the resolution of the classification as a result of which each was resolved into three sub-‐classes (WP2).

The quadrants representing these ‘critical’ urban fabric types were then analysed and characterised, after which a statistically representative sample of quadrants was drawn from each class. These were then analysed further using other geospatial data sets, in particular the Grünraummonitoring (‘green space monitoring’), the Flächenmehrzweckkarte (‘multi-‐purpose digital map’) and aerial photographs, as a result of which one sample quadrant for each urban fabric type was identified. Five sample quadrants, which are subject to distinct heat exposure regimes were selected and have been further investigated and characterised. The characterisation included a description of the typical open space patterns for each quadrant and was followed by the simulation of the status-‐quo and future climate conditions using the programme ENVI-‐met version 4.0. By comparing the respective maps, the climate sensitivity of different open space patterns could be identified with respect to the specific conditions of the urban fabric types in question (WP3).

In parallel the interaction of open space and microclimate has been investigated on the one hand by means of a literature review and on the other hand by studying the climate conditions within different open space situations by means of measurements and simulations with ENVI-‐met 4.0. Comparing the results it was possible to validate the simulation model for further investigation steps (WP4).

The characterisation of the sample quadrants led to specific design recommendations aimed at maximising thermal comfort. Design measures included tree planting, the replacement of sealed hard surfaces with permeable materials, and -‐ where appropriate -‐ roof planting (on the basis of the City of Vienna’s map of Gründachpotential -‐ ‘green roof potential’). The effects of these measures were simulated under the same climatic conditions as the status-‐quo. In response to the specific conditions prevailing in each sample quadrant, the design variants focused on different aspects, such as street orientation and width, influence of density of development and the influence of adjacent areas, etc. The evaluation of the data generated focused on those climatic factors most relevant for thermal comfort:


Summary Report 3

wind speed, mean radiant temperature, air temperature and humidity and on the thermal comfort value itself using the index PMV (predicted mean vote). The results are presented as maps depicting mean values and diurnal variations (WP5).

Based on the simulation results, and taking into account findings of a previous project, a general catalogue of open space design measures was compiled. As a final step a combination of measures was defined for each sample quadrant, as a representative of an urban fabric type. These reflected their specific characteristics in term of their typical open space patterns and their climate sensitivity. The aim in each case was to have an optimal influence on thermal comfort (WP6).

1.3 Motivation and objectives

Urban areas are the living space or place of residence for a significant proportion of Europe‘s population. But at the same time, because of their tendency to overheat during hot summers, cities are particularly vulnerable to the impacts of climate change. There are strong interactions between the design treatment of open spaces and their microclimate.

The objective of this project was to better understand the way in which the small scale structure of the urban fabric contributes different to the heat island effect and other urban climate phenomena. It also aimed to use this information to develop specific strategies for counter-‐acting and mitigating these effects at the local level in order to help secure the future health and well-‐being of the urban population as global warming increasingly influences the urban climate.

Furthermore, the study aimed to investigate how the detailed urban morphology is likely to influence the changing urban microclimate, and to consider how urban design solutions incorporating appropriate mitigating measures can be tailored to suit different settlement structures.

The project followed a systematic approach which involved the analysis of the climatic conditions and the effects of possible design measures based on quadrants defined by a 500m x 500m grid laid over the entire area of Vienna. The quadrants were representative of different ‘urban fabric types’, which are also likely to be similar to those that might be found in other European cities, expecting the results of the project could be transferable.

An important aspect was the close exchange with other specialists from related disciplines. An expert panel was assembled, including researchers on urban climate issues, city authorities and urban planners. The panel came together in the crucial phases of the project to discuss the (interim) results and advise on following project steps.


Summary Report 4

1.4 Work Packages

Work packages

Title Methodology

WP 1 Project management Project management; Expert meetings

WP 2 Generation of urban fabric typologies and sample quadrant selection

Spatial and climate data acquisition / extraction;

Spatial analysis, multivariate statistics

WP 3 Characterisation of urban fabric typologies and identification of open space typologies

Numerical microclimate simulations considering heat exchange, ventilation and cooling through physical properties and open space interior; Microclimate impact analysis; Literature review

WP 4 Investigation of interactions between urban open space design and microclimate

Literature study;

Atmospheric condition monitoring at reference sites and parallel numerical microclimate simulations for model validation;

Microclimate impact analysis

WP 5 Application of findings into the urban fabric and open space typologies

Design recommendations;

Simulations

WP 6 Recommendations of planning and design measures

WP 7 Dissemination and public discussion Dissemination of results and public discussion of recommendations with urban stakeholders

WP 8 Report: Formulation of policy recommendations

Final Report

1.4.1 WP 2 – Generation of urban fabric typologies

Urban typologies were defined using data describing the terrain, urban morphology and open space characteristics, as well as selected microclimatic indicators for each of the 500x500 m quadrants. based on the sample grid used by Statistik Austria. Several steps were necessary to generate the final urban fabric types:

First, a large set of variables for Vienna was acquired, extracted, and generated from other datasets from which a series of indicators were developed. In total around 250 indicators were extracted, 44 of which were selected to be used in the multivariate statistical analysis. The indicators selected were classified into four groups: climate, topography, open space and buildings.

The next step was to analyse these four groups using factor analysis in order to eliminate redundant information and to identify the most influential parameters. Using this approach, a limited number of factors for each group was selected to serve as “super variables”, containing the information of the entire indicator set. The data subset “Climate” was represented by one factor, the data subset “Topography” by


Summary Report 5

two factors and the last two subsets “Open Space” and “Buildings” were represented by three factors respectively.

These factors were then used in a cluster analysis to generate different forced numbers of clusters. After reviewing alternative results (applying different cluster numbers, different similarity measures and metrics), a cluster analysis resulting in 9 clusters was finally selected, providing the best differentiation of built-‐up districts, but without distinguishing too many clusters in less densely populated areas. Finally, the cluster types relating to statistical cases representing the 500x500m grid squares were linked with the respective geospatial entities.

In this way the distribution of the nine different urban fabric types was determined; each representing a combination of physical morphology and climate sensitivity (Fig. 1 and Table 1).

Fig. 1: Final Cluster analysis of urban fabric types.

Urban Fabric Type Climate Building / Land Use / Vegetation

Type 1 Industrial and commercial zones

Strong influence of urban heat island phenomena

Heterogeneous building structure ; High percentage rate of sealed surfaces;

Perimeter block development dominant block characteristics, besides this linear development since the 1960ies and 1970ies; large industrial areas with generous green spaces – therefore high percentage of grassland

Type 2 Densely built-‐up inner urban areas

Highest number of hot nights, warmest winter

Cluster with highest percentage of sealed surfaces; highest buildings, historic building structure predominates (especially perimeter block development) with low rate of vegetation in inner courtyards; less vegetation area overall

Type 3 Urban expansion areas on level terrain

Highest number of hot summer days; but cooling at night

Heterogeneous building structure; approx. half the cluster surface is sealed; level terrain; high proportion of grassland and agricultural use; moderate amount of shrub


Summary Report 6

vegetation and small trees

Type 4 Low density development on sloping terrain (West Vienna)

Cooler and more humid because of the proximity to the higher altitude of the Vienna Woods

Terrain with very varied elevation; areas of detached housing and villa-‐lined thoroughfares with big private garden estates; mainly perimeter block development with a higher amount of vegetation in inner courtyards; higher amount of shrub and tree vegetation (height < 0,5m and > 3,5m

Type 5 Urban fringe areas on level terrain (Vienna Basin)

Hot summer days, cool nights

low-‐density residential areas, with low building heights; recreations areas with large areas of low vegetation; few trees

Type 6 River corridor (Danube)

Moderately influenced by urban heat island effects – similar to Cluster 5; faster cooling process at night

Very low percentage of sealed surfaces; very low numbers of building – high proportion of water areas and trees

Type 7 Un-‐built agricultural land

Lowest amount of precipitation;

little vegetation shade

Detached housing areas with low building density; lower amount of building shade, low percentage of sealed surfaces – large proportion of agricultural land

Type 8 Urban fringe on wooded slopes

Cool, few sealed surfaces subject to heating up

Detached housing areas with a high proportion of vegetation;

Highest amount of shrub and vegetation with heights between 3,5m and 15m

Type 9 Wooded hills (Vienna Woods)

Cool, humid, high amount of shade vegetation, highest amount of precipitation

Low building density; high amount of forest, low proportion of low vegetation; high variations in altitude

Table 1: Characteristics of urban fabric types

The urban fabric types which appeared having the most potential to be explored further were those represented by clusters 1-‐5 and 8. These fabric types turn out to be more vulnerable to certain impacts of climate change (due to the high proportion of sealed surfaces, the low proportion of green open space, the high proportion of hot summer nights, etc.), and were expected as the most promising targets for future mitigation measures. The other fabric types are characterised by, for example, a higher amount of vegetation and fewer sealed surfaces, fewer hot summer nights and colder winters, etc. (e.g. Fabric Type 9 “Wooded hills”), and as a result the impact of climate change here is likely to be much less critical.

Although the selected clusters characterising the city’s urban fabric types generally presented a convincing representation of Vienna’s urban morphology, a further attempt was made to see, if it were possible to derive a more detailed resolution of the urban structure by sub-‐dividing some of the more heterogeneous classes – in particular urban fabric types 2 and 3, and the Danube river corridor Cluster 6. For these clusters further analysis was therefore conducted, leading to the generation of another three ‘sub-‐types’ each (Fig. 2 and Table 2).


Summary Report 7

Fig. 2: Final Cluster analysis of urban fabric subtypes

Urban Fabric Sub-‐type Climate Building / Land Use / Vegetation

Sub-‐type 2a

Late 19th century perimeter block development on sloping terrain (West Vienna)

Hot; highest amount of precipitation of the three others sub-‐clusters; “coolest” sub-‐cluster

Perimeter block structure, about 86% sealed surfaces, elevations in the terrain structure, low amount of vegetation structure, but trees with a height until 3,5m and grassland dominates

Sub-‐type 2b

Late 19th century inner urban perimeter block development (Inner urban Vienna; Floridsdorf)

Hot; high number of hot summer nights; high proportion of sealed surfaces subject to heating effects

Perimeter block structure; high proportion amount of sealed surfaces; small differences in elevation; grassland and small trees structure dominates the cluster vegetation

Sub-‐type 2c

Historic city centre Highest number of hot days and hot nights, influenced by urban heat island – high proportion of sealed surfaces

Perimeter block structure and historic buildings; high percentage of sealed surfaces; grassland and small trees structure dominates the vegetation; highest proportion of water surfaces of the three sub-‐clusters

Sub-‐type 3a

Post WW II urban expansion areas – (South/South East Vienna)

Hot days and warm nights; average amount of precipitation; 50% sealed surfaces which influences the heat island effect

Detached housing areas – land in agricultural use; average building height about 6m; low building density; high amount of grassland (agricultural areas); fewer trees

Sub-‐type 3b

Compact development in urban expansion areas and old village centres (North and East of Vienna)

Warm days and warm nights – highest rate of warm nights among the sub-‐clusters

Different types of built structures (residential and agricultural buildings); highest building density of these three sub-‐clusters; average of building height of about 11m

Sub-‐type 3c

Single family houses (West Vienna)

Coolest sub-‐cluster of the three, highest amount of precipitation among the three sub-‐clusters;

Detached housing areas with more perimeter block structure than in the other two sub-‐clusters; building height


Summary Report 8

about 50% sealed surfaces of about 8m; low building density; high level of shrub and small trees structure; terrain with most variation in elevation of the three sub-‐clusters

Sub-‐type 6a

Riverine woodland (National Park ‚Lobau‘)

Cool, low proportion of hot nights

the lowest percentage of buildings and very low building heights; Highest amount of shrub and tree vegetation with heights between 3,5m to 15m

Sub-‐type 6b

Waterside green spaces (east of the Danube)

Cool; highest rate of precipitation

Higher percentage of buildings – but quite low from a overall point of view; an average building height; detached housing areas; about 30% sealed surfaces; high amount of grassland and shrubs; high percentage of water areas

Sub-‐type 6c

Waterside landscape parkland (Danube banks and Prater)

Lowest rate of winter severity; cool in comparison to sub-‐clusters 2+3

Sub-‐cluster with the highest amount of water; very low building and percentage of sealed surfaces; low variations in terrain; high proportion of shrubs and grassland

Table 2: Characteristics of urban fabric sub-‐types

1.4.2 WP 3 -‐ Characterisation of urban fabric types and identification of open space typologies

The aim of this work package was to identify the characteristics of different urban fabric types in terms of their open space structures and microclimatic conditions.

(1) A selection of sample quadrants was made and those representing each of the selected urban fabric types were identified in order to investigate their microclimatic characteristics as a function of their morphology. The quadrants were selected according to the climate sensitivity open space patterns and as being characteristic of average indicator values for their respective urban fabric types. The final selection of sample quadrants within the urban fabric types 1-‐5 took place in several steps.

(a) Statistical analysis was carried out to obtain a significant number of samples which were representative of a particular urban fabric type. The number of samples required depended on the number of and spread of quadrants within an urban fabric type (Fig. 3).

(b) Representative sample quadrants were selected for each urban fabric type for further analysis and simulations. The main focus was on the identification of typical urban open space patterns, making use of information from the Grünraummonitoring (‘green space monitoring’) and the Flächenmehrzweckkarte (‘multi-‐purpose digital map’) databases, which were integrated using ArcGIS. The most representative quadrants were identified by making use of the spatial data, focussing on categories covering at least 5% of the total quadrant area.

(c) Comparison of aerial photos of the most representative quadrants for the final selection to assure the applicability of the simulation method (e.g. localisation of important open space patterns within sample quadrant and avoiding falsification of simulation results due to boundary effects).


Summary Report 9

Fig. 3: Random sample and final selection of quadrants for further investigation

The quadrants selected were: number 555 -‐ representing urban fabric type 1 as an example for a typical commercial and industrial area; number 723 -‐ representing sub-‐type 2a, characterised by late 19th century perimeter block development; number 919 -‐ representing sub-‐type 2b, characterised by late 19th century perimeter block development with integrated parks near the southern Gürtel area; number 983 -‐ representing sub-‐type 3b, characterised by residential development with apartment blocks separated by green open spaces; and number 1264 -‐ representing sub-‐type 3a, characterised by detached houses and agricultural land (Fig. 4).

555 – Cluster 1 723 – Cluster 2b 919 – Cluster 2c

1264 – Cluster 3a 983 – Cluster 3b

Fig. 4: Orthophotos of the selected sample quadrants


Summary Report 10

(2) Typical urban open space patterns were identified for each urban fabric type as represented by the quadrant selected, with the help of an analysis of rectified aerial photos, which were compared with the databases referred to above. This led to a classification with respect to potential open space design measures (Table 3).

Road system linear street area

widenings Cross-‐section of road

specific junctions

potential to abandonate road sections

Courtyard fragmented within block isolated

(partially) connected

entire block enclosed

opened

Green area fragmented ("mosaic")

connected (“flow“)

extensive ("solitaire")

Square fragmented ("mosaic")

connected (“flow“)

extensive ("solitaire")

Other open areas parking areas

paved areas within industrial sites

rail tracks

waste land

agricultural land

Roof area potential for roof planting

Table 3: Classification of identified open space structures within quadrants

The open space structures identified were digitized for all the sample quadrants using ArcGIS, in order to facilitate an analysis of the percentage distribution typical of each urban fabric type (Fig. 5). The keywords relating to the respective characterisation are presented in Table 4 and these led to different focuses for further investigation.


Summary Report 11

555 – Urban fabric type 1

723 – Urban fabric type 2b

919 – Urban fabric type 2c

983 – Urban fabric type 3a


Summary Report 12

1264 – Urban fabric type 3b

Fig. 5: Open space structures and their percentage distribution within sample quadrants

(3) The microclimatic conditions of the status quo of each selected quadrant were simulated using the ENVI-‐met model, and the results were compared to the maps of the open space structures obtained in step 2 and discussed with regard to the typical and critical climate situations (Fig. 6). The atmospheric input data used for the simulations referred to current climatic extremes using a day exhibiting heat wave conditions, measured in the inner city of Vienna, and reaching the 99-‐percentile of the daily maximum air temperature. Further simulations were conducted under future climate conditions projected for the year 2050 based on the results of the regional climate model COSMO-‐CLM (AIT; reclip:century-‐Simulation). The results showed that the current climate sensitivity will be intensified significantly, emphasising the urgent need for immediate counter-‐measures.

Fig. 6: Comparison of open space structures (left) and status-‐quo thermal comfort conditions (PMV) showing the example of sample quadrant 919 (Urban fabric type 2b)

Comparable open space patterns within the sample quadrants exhibited different climate sensitivity due, for example, to their particular proportions. Open spaces with sealed surfaces heat up strongly, while green open spaces with ground cover vegetation and trees result in the lowest temperatures and the highest comfort values. However, even within green courtyards, sealed areas form hotspots, which result in clearly identifiable thermal discomfort zones especially during the afternoon. Critical areas within the dense city structure could also be identified, for example, in areas where the streets widen and where there are junctions. Figure 6 shows an example of the comparison of the two maps for sample quadrant 919.


Summary Report 13

1.4.3 WP 4 -‐ Investigation of interactions between urban open space design and microclimate

The fourth work package comprised a review of the microclimatic conditions in the context of open space adaption strategies aimed at influencing the local climate under current and future climate conditions.

In addition, a field study investigating certain reference locations at AIT’s Seibersdorf Campus helped to explore the microclimatic conditions of different open space situations such as open grassland, wooded areas, courtyards (observing sites at the south and north side) a non-‐shaded asphalt area and a shaded gravel area in the undercroft of a building. Six micro-‐climate monitoring stations were installed and a set of variables was recorded and stored as mean value every 15 minutes. These were wind speed, wind direction, temperature (at the ground surface and in 2m height), relative humidity, radiation, precipitation (only at the grassland site) and soil humidity at the grassland and the forest sites. The measurements were carried out from the end of July to the beginning of December 2011.

This long monitoring period covered different seasons and weather events and helped to improve the quality of the modelling exercise for different seasonal conditions. Additionally, it allowed the identification of a reference date which was representative of a longer stable weather period. The analysis of the micro-‐climatic data showed an extended period of hot sunny weather around the 21st of August 2011, which is where the research was focussed, providing the basis for various micro-‐climatic simulations. A comparison of the empirical results with the simulations was undertaken to calibrate the ENVI-‐met model. This helped to improve the understanding how the model works and which parameters are sensitive to changes. It also provided insights into the conditions in different urban settings and their influence on microclimate.

As a first step, the input data for the simulation were prepared. The raster layer of the Seibersdorf Campus test site (190x190 pixels, 2m cell size) was generated; ground surface, buildings (heights, roofs, materials) and plants defined and local soils and plants were added to the ENVI-‐met parameter-‐databases. Initially ENVI-‐met version 3.1 was used. This could only allow for the definition of temperature, humidity, cloud cover, wind speed and direction as initial framework conditions. During each simulation the temperature and humidity gradient was calculated within the model, depending on the irradiance conditions with respect to the diurnal variation of the season and on the day-‐time specific solar angle according to the geographic location. This resulted in a significant deviation between simulation and measurements. Subsequently, a newer version, ENVI-‐met (version 4.0) was used, which allows a “simple forcing” of the hourly temperature and humidity gradation. As a first simple forcing attempt, the monitored hourly temperature values (and the related relative humidity) were included. However, the simulation results still remained 4-‐5°C below the monitored values especially between the hours noon and the early afternoon. The results matched the average temperature progression during August quite well, but still showed a deviation from the values of the specific day. So the external temperature forcing was adapted, achieving a peak of 2-‐4.5 °C above the monitored temperatures. The new simulation results now showed a curve similar to the measurements (Fig. 7).


Summary Report 14

Fig. 7: Temperature gradation of the ENVI-‐met simulation and the measured temperature on 21st August 2011 at the AIT Seibersdorf campus.

1.4.4 WP 5 -‐ Application of findings to the urban fabric and open space typologies

The aim of this work package was the definition of design variants aimed at the amelioration of the local climate conditions. It included the respective simulations using ENVI-‐met and the evaluation of the results in the form of simulation maps, extracted mean values and diurnal variations.

(1) Based on the input files and results of the status quo simulations (work package 3) different design variants were defined and modelled using ENVI-‐met to investigate the respective microclimatic effects. The modelled design variants were:

(a) different forms of tree planting with respect to appropriate measures for the identified critical climatic aspects within each open space type. The implementation of tree planting with regard to the type (deciduous), size (crown 12m) and planting distance (canopy closure) was based on the results of a previous project (FREIRAUM UND MIKROKLIMA 2011).

(b) the de-‐sealing of the ground surface within the affected open space types. This involves a simple change of surface material as well as potential changes of use e.g. the transformation of certain traffic lanes to public space.

(c) implementation of extensive roof planting based on the data of the Gründachpotentialkataster (‘green roof potential map’).

The design measures were simulated individually and in appropriate combinations, allowing conclusions to be drawn regarding specific open space situations including an “optimal variant” for the entire quadrant area. Depending on the characteristics of each sample quadrant the focus of the design measures was placed on the corresponding aspects (Table 4).


Summary Report 15

Quadrant Open space structure Focus of design variants Variants

555

Urban Fabric Type 1

characterised by extensive paved surfaces and parking areas, high potential for roof planting

De-‐sealing of ground surfaces in extensive paved areas as far as reasonable, unsealing of ground surfaces together with tree planting along building sites and on parking areas, area-‐wide extensive roof planting

4

723

Urban Fabric Type 2a

characterised by orthogonal road system (N, W) and fragmented, partially connected courtyards

tree planting along the streets with focus on street orientation as well as on the respective street sides (facades)

5

919

Urban Fabric Type 2b

Characterised by orthogonal road system with widenings (NW, NE), rail tracks and adjacent paved operational areas as well as all types of courtyards and extensive green areas.

tree planting along the streets with focus on the south façade along the Gürtel-‐road, de-‐sealing of ground surfaces within areas of street widening and possible abandoning of street section along the Gürtel-‐road, de-‐sealing of parking and paved operational areas as far as reasonable, de-‐sealing and tree planting within larger courtyards

10

983

Urban Fabric Type 3a

characterised by linear street areas with widenings at road junctions, open courtyards, linked green areas and parking areas.

tree planting along the streets with focus on junction widenings, de-‐sealing and tree planting in possible closed street sections, de-‐sealing and tree planting within parking areas

9

1264

Urban Fabric Type 3b

characterised by linear street areas, fragmented green areas and agricultural land

tree planting along the streets with focus on street orientation as well as on the respective street sides (facades), de-‐sealing and tree planting within parking areas

6

Table 4: Open space structure characteristics and focus of the investigated design variants

(2) The simulations were conducted using ENVI-‐met version 4.0, simulating 36 hour run time durations for selected seasonal characteristics focussing on hot, cloud-‐free summer days with high temperature amplitude.

The resulting data was analysed through simulation maps of mean values and diurnal variations. The focus was placed on the PMV (predicted mean vote) value, which integrates the effects of wind speed, mean radiant temperature, potential air temperature and humidity. Figure 8 gives an overview of the simulation maps generated for the different variants of sample quadrant 919 as an example of the approach taken for all sample quadrants.


Summary Report 16

Fig. 8: Overview of simulated variants for the sample quadrant 919 and respective maps of mean radiant temperature, potential air temperature, wind speed, specific humidity and PMV (predicted mean vote)

Generally, tree planting shows the greatest effect on all microclimatic factors analysed, especially within the immediate area shaded by the tree crown. The PMV value is reduced from “extreme hot conditions” to “slight warm conditions”. A clear correlation can be observed with specific pattern of tree distribution in relation to the adjacent open space structures and to the width and orientation of the respective open spaces. Figure 9 shows the differences between the status quo and the design variants of tree planting along the east-‐west (VB2) versus north-‐south orientated streets within sample quadrant 723. The maps of the respective PMV values and the differential maps show the greater effectiveness of taking measures within the east-‐west orientated streets, which lowers the PMV by 3,5 degrees across a wider street area. Figure 10 shows the diurnal variations for different microclimatic factors of the status quo (purple) versus the design variant of street planting in the east-‐west orientated streets (blue). The diagrams illustrate a significant decrease of the mean radiant temperature and a similarly significant increase of the specific humidity due to tree planting. Although the wind speed is only reduced slightly, the air temperature can be lowered by up to 1.5°C during the hottest period of the day and by 3°C at night.


Summary Report 17

Fig. 9: Simulation maps for the PMV-‐values showing the status quo and the design variants for tree planting in the east-‐west versus the north-‐south orientated streets for quadrant 723


Summary Report 18

Fig. 10: Diurnal variations for the climatic factors: mean radiant temperature (top.left), windspeed (top.right), specific humidity (bottow.left) and air temperature (bottom.right) for the status quo and for the design variant of tree planting along the east-‐west orientated streets fo sample quadrant 723

Just de-‐sealing the ground surface generally results in a slight reduction of the PMV value. Figure 11 shows the maps for the PMV values for the status quo and the design variant of permeable ground surface (especially parking areas) for quadrant 555, illustrating the effect referred to above.

Fig. 11: Simulation maps for the PMV-‐values showing the status quo and the design variant for de-‐sealing of ground surfaces for quadrant 555

The following simulation maps for sample quadrant 919 highlight further microclimatic effects. Figure 12 shows the PMV values for the status quo and the design variant of de-‐sealing of the industrial area to the south-‐west showing a slight reduction at 3 p.m. as mentioned above. Yet looking at the respective differential maps of diurnal variation in Figure 13 a cooling effect of air temperature within the residential area to the north-‐west can be observed due to the south-‐easterly wind conditions.


Summary Report 19

919

Status quo vs.

unsealing of areas

PMV value Differential map

Fig. 12: Simulation maps for the PMV-‐values showing the status quo and the design variant for de-‐sealing of ground surfaces for quadrant 919

Fig. 13: Differential maps between the status quo and the design variant for de-‐sealing of ground surfaces for quadrant 919 showing the PMV-‐values differences at 9 a.m., 3 p.m., 6 p.m., 9 p.m. and 0 a.m.


Summary Report 20

Roof planting generally has less direct effect at ground level. Looking at the vertical dispersion of microclimatic data values, however, the amelioration effect is highlighted. Figure 14 shows a vertical section of the potential air temperature for the design variant involving roof planting for quadrant 555. The map illustrates a significant cooling effect at higher air levels and on the leeward side of the buildings.

Fig. 14: Vertical section of simulation map showing the vertical distribution of air temperature for the design variant: roof planting for quadrant 555

(3) Finally the status quo situation for selected quadrants was compared to their respective “optimal variants” in order to demonstrate the effects of the simulated design measures within the different open space structures. To show the effects of individual measures on the entire quadrant area and the changes during the course of the day, average mean values were calculated for 9 times of day. The main focus has been put on dense urban areas, as represented by quadrants 723, 919 and 983, where large numbers of citizens are likely to be affected. This final analysis provided an important basis for the “packages of measures” proposed for the sample quadrants in their capacity as representatives of the individual urban fabric types in Workpackage 6.

The detailed simulation results and the final analysis can be found in the full-‐length report published on the project website (www.urbanfabric.tuwien.ac.at).

1.4.5 WP 6 -‐ Recommendations of planning and design measures

The results of the simulations have been evaluated with respect to the overall open space patterns, leading both to general recommendations and to specific packages of measures for each sample quadrant and corresponding urban fabric type.

(1) The catalogue of general recommendations (Maßnahmenkatalog) has been drawn up according to the type of design measures concerned: tree planting, de-‐sealing of ground surfaces and roof planting. Additional information and recommendations are based on a literature review and on the results of a previous project, dealing with aspects of tree size, type, distribution, planting distance and different forms of de-‐sealing ground surfaces (FREIRAUM UND MIKROKLIMA 2011


Summary Report 21

(2) Specific packages of measures (Maßnahmenpakete) have been defined according to the characteristic open space structures within the respective sample quadrants and ranked on the basis of their effectiveness, thereby defining a hierarchy of priorities.

For each urban fabric type a short description of the characteristic topography, the urban structure and climatic conditions is presented. The sample quadrants are discussed on the basis of the percentage area of existing open space patterns, which define their representative status for the urban fabric type. Based on the general recommendations in the catalogue, packages of measures have been defined and ranked in the form of priorities which take into account the climatic effect of the recommended measures (locally as well as on the nearby surroundings), the potential areas involved, the extent of implementation and the potential density of users. In a second ranking (see Priority*) the percentage occurence of the open space structures in question has additionally been taken into account. This differentiation allows for the consideration of different aspects with respect to local or area-‐wide implementation, for example, or to time and effort required. Figure 15 shows an example of the percentage distribution of open space structures within quadrant 555 representing urban fabric type 1 (Industrial and commercial zones) leading to the package of recommendation in Table 5.

Fig. 15: Percentage distribution of open space patterns within sample quadrant 555

Table 5: Package of measures for sample quadrant 555 representing urban fabric type 1 and ranked in order of priority.

The complete catalogue of recommendations and the specific packages of measures can be found in the full-‐length report published on the project website (www.urbanfabric.tuwien.ac.at).


Summary Report 22

1.4.6 WP 7 -‐ Dissemination and public discussion

An important aim of the project was the public dissemination of the results in order to raise the awareness of climate and open space issues in urban planning. As a result, poster exhibitions presenting information about urban microclimate in general as well as specific information about the project were organised in June and August 2013 as a basis for public discussion evenings. The 16th, 21st and 23rd districts of the City were chosen for dissemination events as these were the locations of three of the sample quadrants studied. The exhibitions were organized in cooperation with each of the local Gebietsbetreuung (‘district urban renewal offices’). A concluding presentation with a final discussion was organized at the Gebietsbetreuung in the 16th district, and this also offered the opportunity to present the first simulation results. The audience included the project partners, members of the expert team, and colleagues from related projects as well as interested citizens, who responded to the invitation of the Gebietsbetreuung.

1.5 Conclusions

General Conclusions

• Climate change projections consistently point towards an increasing occurrence and intensity of heat

waves, which are expected to become a considerable public health problem, especially in cities, due to

urban heat island effects.

• When dealing with urban heat island effects the influence of the built environment and the open

spaces between must be considered. The typology of the different urban fabric types represented

through the sample quadrants illustrates the close interaction between open space structure and local

climate conditions.

• Urban green is very important to decrease the local temperature regime, to avoid higher irradiance and

support nocturnal cooling by energy flux to the free atmosphere improving outdoor and indoor thermal

comfort. The microclimate simulations, virtually testing different adaptation measures, show a clear

correlation between cooling effects and the distribution of trees and further the width and orientation

of open spaces. Tree planting was observed to have the greatest effect on all microclimatic factors as it

helps to enlarge the shaded area and to reduce the extent of areas exposed to high radiant and surface

temperature, further increasing cooling through transpiration.

Typology of urban fabric types and open space patterns of Vienna

The typology considers a large number of indicators including aspects of climate as well as topography, built up area and open space.

• The urban space typology shows close agreement with similar urban typologies addressing other

criteria, as well as reflecting the relationships between urban structure and urban climate pattern. The

characterization of different urban space types, on the basis of the sample quadrants investigated also

illustrates the close interrelationship between open space structure and the local climate conditions.

• One important finding was that the microclimate conditions showed a high level of variation between

the different urban fabric types but also within individual quadrants across relatively small distances.

The explanation lays in the layout and volume of the built structures as well as the percentage of paved


Summary Report 23

surfaces with respect to heat storage, to the shading conditions and general air ventilation as well as

open space characteristics.

• The simulation showed that the right location of planting measures can be more important than the

overall extent of planting. In some cases a single line of trees between street and settlement area has

more cooling effects on the microclimate of the settlement behind the treeline than several rows of

trees in the narrow streets within the settlement area.

• Comment: Additional support is provided when the open space is to some extent irrigated providing

moisture which accelerates evaporation from the soil and transpiration from vegetation and thus

cooling.

Measures

The implementation of the proposed amelioration measures in open spaces helps to improve local thermal conditions. The analysis of the microclimate simulations show the effect of various measures (planting trees, unsealing of paved surfaces, roof greening).

• Tree planting

Planting trees helps -‐ in addition to the increase of evaporative cooling through transpiration -‐ by enlarging the shaded area and thus reducing the extent of areas exposed to irradiance and radiant temperature.

Taking into account the average effects over the course of the day, tree planting causes a significant reduction of mean radiant but also air temperature. The minimum, maximum and mean air temperature are all reduced.

The minimum air temperature occurs earlier than without tree plantations.

Comment: locally, e.g. in courtyards, a negative effect may occur especially in the morning hours so that the temperature is a little higher with additional trees than without. This can be explained by reduced wind and air exchange due to smaller sky view characteristics.

• De-‐sealing

De-‐sealing of paved surface generally reduces the surface temperature and increases the area able to provide evaporation as long as the soil contains a certain water concentration.

De-‐sealing of large surfaces can cool down the air temperature within neighbouring dense city quarters in the direction of the wind.

Comment: de-‐sealing has a further effect providing extended porous surfaces serving as drainage facility that can act to reduce flooding, which is expected to increase because of climate change. In addition it can contribute to recharging groundwater. .

• Roof planting

Roof planting generally has effect within the roof area itself and on the equivalent levels of adjacent buildings.

Roof planting has a cooling effect even on ground level in the direction of wind, particularly when implemented on lower buildings.


Summary Report 24

Comment: roof planting (as well as planting on facades) also contributes to enlargement of shading and building insulation enhancing indoor cooling during summer days

Recommendations

In quadrants with similar open space characteristics, differences could be identified due to local variations in the urban fabric (orientation, ratio height / width, adjacent land or buildings, etc), leading to the need for different priorities with regard to design measures.

• The results can serve as a planning guideline and to provide decision support for urban design and

should create greater awareness for the need for climate sensitive open space design given

appropriate presentation of measures and impacts.

1.6 Discussion and outlook

Project highlights

• The urban space typology shows close agreement with similar city typologies addressing other criteria,

as well as reflecting the relationships between urban structure and urban climate pattern.

• The simulation results support the theoretical knowledge in the urban climate debate facilitating a

better understanding and stronger arguments for the proposed measures.

• The expert meetings proved to be very constructive and the participants were very supportive of the

project. They resulted in a high level of interdisciplinary and transdisciplinary discussion. The

contributions of the experts were particularly valuable in refining the project approach, and both sides

were able to profit from the intense discussions.

• The findings of the research project are not only sound from a scientific point of view, but also of high

relevance for practitioners. Representatives from the City of Vienna also showed great interest in this

project. There has been a close exchange with the MA22 for Environmental Protection opening the

potential for future cooperation.

Target groups able to benefit from the results of the project

• City authorities responsible for open space maintenance may benefit from the results as help guiding

the practitioners in implementing the suggested measures real time.

• Researchers in the field of urban climate and open space design can profit from the results for their

own research.

• The project results constitute a basis for urban design processes, strengthening the awareness of urban

climate issues and providing practicable measurement proposals for urban planners.

• Last but not least the exhibition posters and public discussions serve the awareness raising with regard

to urban climate issues and to possible mitigation measures which can be implemented by local people.


Summary Report 25

Difficulties during the project period encountered in the achievement of the project targets

• The choice of the indicators required a very intensive review of the literature which led to a small delay

in the beginning of the study.

• Not all important indicators could be extracted from the available data sources (e.g. sky view factor),

because this would have required additional programming efforts, which were outside the scope of this

project.

• Initial problems with the climate simulations due to the usage of the latest undocumented Beta-‐version

of the simulation programme ENVI-‐met, which was more suited for this project and thus recommended

by the developer during the 1st expert meeting, which led to a delay in WP 3 of about 3 months.

• Despite the dissemination efforts in form of the exhibitions, there was not enough opportunity to

discuss the different measures proposed in the project with all stakeholders. It would have been

especially interesting to consider the final recommendations with a broader group of stakeholders

including the proprietors of the open spaces.

General outlook

• The results obtained suggest that the approach used based on generating an urban fabric typology as

the basis for a differentiated assessment of the local impacts of climate change is one which could

usefully be refined and extended to other cities. This could involve expanding the types of open space

characteristics considered and range of climate conditions at the city scale as a means of supporting the

urban population in adapting to climate change and higher temperatures.

• It is possible that the results may encourage local activities to implement selected measures in sample

areas. Monitoring of the results of this work and comparing these to those from the ex ante assessment

carried out in this study would provide important information.

• Potential adaptation measures countering heat stress ought not only to focus on the treatment of open

spaces, but also on their interaction with the building design considering, for example, building facades

and the implementation of evaporative cooling instead of standard air conditioning.

• The effects of the proposed measures have only been investigated for a small portion of the city. How

they would affect the city area as a whole, for example in terms of large scale ventilation or sultriness,

needs to be studied. This could be answered using the model MUKLIMO3

• The design measures have only been simulated for one set of specific extreme conditions during

summer (hot/dry/windstill). How the recommended measures could affect the microclimate during

other weather conditions and seasons needs to be investigated in further detail.

• Within the framework of the project, the focus has been on current climate conditions against the

background of their intensification as a result of climate change. How the recommended measures

perform in the face of changing climate, for example with respect to increased drought and reduced

winter length is a question for which answers need to be found.


Summary Report 26

• The results are based on the assumption that the measures will be implemented, something which is

dependent on various factors, such as legal or political restrictions, technical problems, financial

difficulties, conflicts of use, lack of knowledge and acceptance of the resident population, etc. To

improve the chances of the proposed measures being carries out, the current governance structures

and adaptation policies need to be investigated and opportunities for interventions in policy identified.

(This could partly be answered during the current UHI-‐project)

Further steps that will be taken by the project team

• A further cooperation with the Vienna environment protection department MA 22 will be appreciated

to start implementation, allowing real world tests.

• Further work to test and refine the approach to the definition of the urban fabric typology and to make

it applicable to both a wider range of subject areas (in addition to climate change) as well as in a wider

geographic context (other cities) needs to be pursued.

• A follow up project will conduct city-‐wide urban climate simulations to roughly identify not typical

sample quadrants but hot spot zones where residents display a greater vulnerability to heat stress.

Further simulations will be conducted to obtain a detailed view of heat stress for different weather

conditions in those neighbourhoods and living labs will be established, inviting local residents,

stakeholders, experts and city representatives to share experience of local heat stress, compared with

simulation results.

Urban fabric types and microclimate response –

assessment and design improvement.

Final Report

Website: www.urbanfabric.tuwien.ac.at

TU Wien

Institut für Städtebau, Landschaftsarchitektur und Entwerfen

Fachbereich für Landschaftsplanung und Gartenkunst

Operngasse 11, A – 1040 Wien

www.landscape.tuwien.ac.at

Prof. DI Richard Stiles, Dr. Katrin Hagen, DI Heidelinde Trimmel, DI Beatrix Gasienica‐Wawrytko

Prof. Dr. Wolfgang Feilmayr (Fachbereich Stadt‐ und Regionalforschung)

AIT Austrian Institute of Technology GmbH

Energy Department

Giefinggasse 6, A – 1210 Wien

www.ait.ac.at

Dr. Wolfgang Loibl, Dr. Tanja Tötzer, Mag. Mario Köstl

TU München

Lehrstuhl für Strategie und Management der Landschaftsplanung

Hans‐Carl‐von‐Carlowitz‐Platz 2, D – 85354 Freising

www.landschaftsentwicklung.wzw.tum.de

Prof. Dr. Stephan Pauleit, DI Annike Schirmann

Urban Fabric Types and Microclimate Response Assessment...

Documents

Transcript of Urban Fabric Types and Microclimate Response Assessment...