Towards an Enhanced Green Roof System

World Green Roof Congress, 19-20 September 2012, Copenhagen Page 1

Towards an Enhanced Green Roof System

Christian Berretta, University of Sheffield, ([email protected]), United Kingdom

Tobias Emilsson, ZinCo GmbH, (tobias.emilsson@zinco‐greenroof.com), Germany

Nigel Dunnett, University of Sheffield, ([email protected]), United Kingdom

Virginia Stovin, University of Sheffield, ([email protected]), United Kingdom

Ralf Walker, ZinCo GmbH, (tobias.emilsson@zinco‐greenroof.com), Germany

Abstract

Green roof research has generally been developed as single stranded projects

investigating either plant, stormwater attenuation or aesthetic performance. There

are few examples of integrated research projects linking plant performance and

substrate design to stormwater management either on a local roof or drainage

basin scale.

The University of Sheffield Green Roof Centre, together with ZinCo GmbH, is involved

in the project “Green Roof Systems” within the EU FP7 Marie Curie (IAPP). The main

aim of the project is to enhance traditional intensive and extensive green roof

systems by revisiting the fundamental basis of green roof system design. The aim is

to optimize both the stormwater management and the plant performance with a

renewed focus on the aesthetics.

The project is divided into three Work Packages (WP). In WP1 a standardized plant

screening protocol has been developed and used to investigate plant performance

for a range of species in relation to growing media depth and moisture availability.

The protocol has been tested in two climatic contexts: continental (Stuttgart, DE)

and maritime climate (Sheffield, UK). WP2 is focused on the detention effect in the

substrate and drainage layer, water transfer between components and physical

characterization of substrates optimised for retention and plant survival during

drought. Evapotranspiration rates have been studied as well as vertical fluxes from

drainage layer to substrate. WP3 is focused on studying the complete green roof

system by using the knowledge developed in the previous parts of the project. In

WP3 we are implementing a physically-based hydrological model specific for green

roofs, validated on experimental data acquired through test beds characterized by

traditional and innovative green roof systems and evapotranspiration tests from

climate chamber.

The paper will explain how key findings from WP1 and WP2 have informed the

development of the enhanced systems that will be trialled and modelled as part of

WP3.


Authors’ Biographies

Dr Christian Berretta’s research focuses on hydrological and environmental

processes monitoring and modelling in urban environment, Sustainable

Drainage Systems (SuDS) to restore pre‐development hydrological condition

while controlling targeted pollutants, and the assessment of source area runoff

impact on aquatic ecosystems and human health. He developed his research

activity at the University of Genoa, Italy (2001‐2007) and at the University of

Florida, US (2007‐2011). He is currently working as Marie Curie senior research

fellow at the University of Sheffield, UK.

Dr Tobias Emilsson is working as Marie Curie senior research fellow at Zinco

GmbH. His main work is currently focused on substrate design and water

relations of extensive green roof substrates. Tobias Emilsson has a background

in Plant ecology and a PhD in technology focused on Extensive green roofs. His

previous work has involved vegetation development and nutrient runoff from

extensive green roofs.

Dr Nigel Dunnett is Director of The Green Roof Centre, Sheffield, UK, and

Professor of Planting Design and Vegetation Technology in the Department of

Landscape, University of Sheffield. He has a background in botany, horticulture

and ecology. His work revolves around innovative approaches to planting

design, and the integration of ecology and horticulture to achieve low‐input,

dynamic, diverse, ecologically‐tuned designed landscapes, at small and large

scale.

Dr Virginia Stovin is a Senior Lecturer in the Department of Civil and Structural

Engineering at The University of Sheffield. Her work focuses on urban drainage

structures and processes, most recently on the hydrological performance of

SuDS. She is an enthusiastic proponent of SuDS retrofitting, and is a co‐author

of the recently‐published CIRIA retrofitting guidance. As part of the Sheffield

based Green Roof Centre, she has constructed 11 green roof test beds. The

long term records from these beds are being combined with laboratory studies

to underpin the development of green roof hydrological performance

modelling tools suitable for urban stormwater management planning.

Ralf Walker is the Head of R+D at ZinCo GmbH in Germany and has undertaken

lots of developments in the field of Green Roofs. His background is in

horticultural engineering, soil science and plant nutrition as well as in plants.


Background

Industrial Context

Modern green roof technologies were developed in Germany and guidelines for roof greening

have been published by the German organisation for landscaping research (FLL) since 1984.

Green roofs have become a common feature of built environment because of their multiple

benefits, including stormwater attenuation, biodiversity support, cooling effect of buildings and

aesthetic value (Oberndorfen et al., 2007).

In the last five years the context and reasons for green roof installation have changed

somewhat. The demand for extensive green roofs that are less resource and maintenance‐

intensive systems and which have high aesthetic value is increasing. This ecological view

demands that extensive green roofs become biologically more diverse whilst also offering

improvements in delivery of ‘ecosystem services’ such as stormwater retention, carbon

sequestration, energy conservation and nutrient cycling. The model light‐weight sedum

extensive roof does not appear to deliver, in the long‐term, these desired objectives (Compton

and Whirlow, 2006; Brenneisen, 2006).

The key drivers for the evolution and development of commercial green roof products for

extensive green roofs include developing a longer water supply out of the systems in order for

the vegetation to survive drought periods, whilst ensuring that under wetter conditions there

shall not be an over‐supply of water, which may promote changes in vegetation, such as

establishment of grasses dominating over drought resistant species. A new focus on

evapotranspiration and inter‐event moisture balance and transfer within components of the

system is required. This relates to a better understanding of the plant physiology and growing

medium physical properties. There is an increasing pressure to include native species and to

widen the plant diversity of green roofs in the contest of an increasing interest in biodiversity

potential. Furthermore a new focus on stormwater attenuation represents another

development driver.

For intensive green roof there is also the need to minimize or eliminate irrigation to increase

sustainability. This will require the investigation of suitable plant species alternative to current

planting regimes as well as the development of green roof system components, substrate and

drainage layers, that optimize moisture fluxes. Particularly, there is a specific need to research

alternatives to lawn or turf grass that can give a uniform, evergreen surface, without the need

for intensive irrigation.

Problem

Plant selection for extensive green roofs at an international level is still largely dependent on

lists produced from research in Germany in the last 20 years and, for intensive green roofs, on

the use of standard lawn mixtures and landscape plants. This assemblage of plants was

developed under one climatic regime. There is a need to further extend the range of plants

that are used and that are suitable under different climatic regimes through the development

of rigorous and standardized methodologies that enable the characterization of plant species


according to their optimal growth requirements and their tolerance limits to environmental

stress. Often plant testing and screening are performed on a small scale, and not undertaken on

a rigorous basis that gives a full and detailed characterization of the plant’s requirements.

Much of the existing knowledge about the hydrological performance of green roofs is derived

from field or laboratory experiments in which observations of rainfall and runoff have been

used to derive empirical ‘black‐box’ performance functions. The predictive value of these

relationships is, however, restricted to each study’s specific system configuration and climatic

influences. This means they cannot be utilised with confidence in other contexts. At the same

time, as the individual influence of each system component (plant, substrate, drainage layer) is

lost in the overall system performance, it is difficult to optimise the design of either individual

components or complete systems to meet specific performance objectives. There is the need

for single component based understanding of performance as well as, linked to fundamental

physical properties of the system, to enable modelling and system design.

By having a better understanding of how each component influences the processes occurring in

green roof systems, it is possible to adapt the configurations or combinations of components to

meet specific criteria, such as aesthetic, stormwater management, lower maintenance. If

structural limitations provide the boundaries for the design, climatic characteristics dictate the

most effective combination of substrate and vegetation. Vapour pressure gradient, radiation,

wind, temperature, hydrological regime in terms of rainfall intensity, depth, antecedent dry

weather periods, internal intermittency causing frequent wetting‐drying cycles are all factors to

take into consideration in the design. In green roofs, especially extensive ones, design criteria

can be conflicting. One example is the tension between optimizing the system for stormwater

management or for inter‐event plant survival. Substrates with higher maximum water holding

capacity are usually characterized by higher organic matter content and a larger number of

small pores. If these properties favour plant survival during drought, they are also likely to

induce greater matric pressure and increased resistance to water balance changes, thus

decreasing the stormwater management benefit due to a slower regeneration of the available

water capacity within the system. Other examples are choosing native plant species or species

with higher aesthetic value over drought resistance species or favouring retention over

detention through the design.

In this framework, the University of Sheffield Green Roof Centre, together with ZinCo GmbH, is

involved in the project “Collaborative Research and Development of Green Roof Systems

Technologies” that aims at enhancing traditional intensive and extensive green roof systems by

revisiting the fundamental basis of green roof system design. The objective is to provide a

profound understanding of green roof system performance and of the potential for

optimization to meet the new challenges described previously. The project is funded under the

European Union People programme as a Marie Curie Industry Academia Partnerships and

Pathways (IAPP) project. It is the largest international green roof project to date, involving 11

researchers from an academic institution and a commercial partner and has a long time span,

running over 4 years. This paper presents the research methodology developed during this

project.


Learning Objectives:

• A deeper understanding of the processes occurring in green roof system through a

multidisciplinary integrated approach

• The influence of single components in the performance of the complete systems and potential

for their optimization

Approach

The project is divided into three Work Packages (WP). In WP1 a standardized plant screening

protocol has been developed and used to investigate plant performance for a range of species

in relation to growing media depth and moisture availability. The protocol has been tested in

two climatic contexts: continental (Stuttgart, DE) and maritime climate (Sheffield, UK). The

experimental work is conducted in genuine roof environments in urban areas (Figure 1). A cross

factorial experimental design is used, which involves 3 different levels of water availability

obtained by different irrigation regimes (low, moderate and abundant) and three different

depth of the growing medium (5, 10 and 15 cm). A growing medium composed of 55% crushed

brick, 30% pumice, 10% coir fibre and 5% composted bark was specifically adopted for the

project as a reference substrate not containing peat. 46 plant species (29 forbs, 10 grasses, 7

succulents) have been tested. Spaces of sowing was 10 cm to permit plant interaction at an

early stage. The arrangement of plant species in each module was determined randomly. At this

stage of the project data have been collected for two growing seasons (2010 and 2011) in the

University of Sheffield site and 1 growing season (2011) at ZinCo in Germany. The following

data were collected: percentage germination, shoot extension, maximum height of flowering

stem, mean diameter, number of inflorescence or flowering stem, species survival and

percentage dieback of vegetative growth.

Figure 1 Experimental sites showing the plant trials at the University of Sheffield, UK and at

ZinCo, DE.

WP2 focused on investigating the hydrological processes occurring in a green roof system. Prior

to a rainfall event the substrate is characterized by an initial moisture content MC0. The

maximum moisture content that a substrate can hold is referred to as its field capacity, WCmax.

During a rainfall event the substrate will retain moisture up until field capacity is reached. The

amount of rainfall retained therefore equates to the difference between MC0 and WCmax. As a

Univ. of Sheffield, UK ZinCo , DE


result of plant evapotranspiration, the moisture content will tend to decrease during dry

periods, but it is approximately close to WCmax soon after a storm event that generated runoff.

Any excess rainwater is temporarily stored within large air pores, but will typically drain from

the roof under gravity within two hours. This temporary storage effect is referred to as

detention, and it provides an important stormwater management function through delaying

and reducing the impacts of storm peaks on sewer systems or watercourses. The rate at which

temporarily stored moisture exits from the detention storage depends on substrate physical

characteristics (‘vertical’ detention) and drainage layer characteristics (‘horizontal’ detention).

After the storm event has ceased, the roof will continue to drain until the transient detention

storage is empty and the substrate is at field capacity (Kasmin et al, 2010). Over subsequent

days, the substrate will then lose moisture gradually as a result of plant evapotranspiration.

The rate at which moisture content decreases depends on the plant physiology, the substrate

characteristics and the climatic conditions. During dry periods the moisture level in the

substrate can decrease to the level that that plants experience drought‐stress. The permanent

wilting point represents the condition in which moisture is no longer available to plants. The

soil structure and the pore size distribution characterize its moisture release behaviour or pF

curve.

WP2 aims to understand and quantify each of these processes and has therefore focused on

the establishment of experimental methodologies to assess the following:

‐ Evapotranspiration rates;

‐ pF curve quantification for substrates;

‐ Detention in the drainage layer;

‐ Detention in the substrate;

‐ moisture vertical flux (drainage layer to substrate);

‐ Substrate amendments to enhance runoff detention and plant soil moisture availability.

All aspects of the investigation have required either new methodologies to be developed or for

existing approaches to be significantly modified to account for the specific features of green

roof substrates. At the same time, physical characteristics of substrate were investigated

through FLL tests (FLL 2008) and pore space distribution. Table 1 provides a list of the tests

performed for each green roof element.

Although WP2 focuses primarily on the performance of individual elements within the green

roof system, the work is placed into context through the continuous monitoring of 9 external

test beds installed at the University of Sheffield. The test beds have been established to assess

the extent to which substrate type and vegetation treatment affect long‐term runoff retention

and detention performance. In particular, three vegetation options (Sedum, meadow flower

mixture, no vegetation) and three substrates were selected for investigation. Two commercial

substrates manufactured by Alumasc, namely Heather with Lavender (HwL) Substrate and

Sedum Carpet (SC) Substrate, were considered alongside a Lightweight Expanded Clay

Aggregate (LECA) based substrate. The Heather with Lavender and Sedum Carpet substrates

contain crushed brick and selected mineral aggregates, enriched with a small amount of mature

compost (Alumasc, 2011). The LECA‐based substrate contains 80% of LECA, 10% of loam (John

Innes No. 1) and 10% of compost (Poë et al, 2010). The field installations include weather


stations, and selected beds incorporate water content reflectometers for moisture content

vertical gradient measurement.

Table 1 Tests performed during the project for each component of the green roof system

Component Test Methodology

Ve

ge

tati

on

Plant Screening

Programme

46 plant species (29 forbs, 10 grasses, 7 succulents) tested

in 2 climatic conditions (maritime, Sheffield, UK and

continental, Stuttgart, DE); 3 irrigation regimes; 3 depths

of growing medium

Ve

ge

tati

on

‐

Su

bst

rate

Evapotranspiration 3 substrates and 3 vegetation options tested in a climate

chamber to simulate spring (5.01‐9.76 °C, 12 hours

sunlight) and summer conditions (13.76÷19.84 °C, 17

hours sunlight)

Phytometer (substrate test

using plants as bio‐

indicators)

8 synthetic/inorganic and 6 organic amendments tested vs

3 plant species at greenhouse condition set at 22 °C and

relative humidity 70% (OECD 2006)

Su

bst

rate

Pressure plate extraction

(pF curves)

Substrates and amended substrate tested from ‐0.35 to ‐

15 bar pressure (specific method)

FLL tests (FLL 2008) Granulometric distribution, apparent density (dry

condition and at max water capacity), total pore volume,

max. water holding capacity, permeability, organic

content, pH, nutrients.

Pore space distribution Image analysis of sections of substrate cores solidified in

resin ‐ 4 substrate options

Detention Small‐scale laboratory rainfall simulator for detention

process vs substrate depth, organic content, rainfall

intensity, presence of moisture mat

Su

bst

rate

‐

Dra

ina

ge

lay

er

Moisture vertical flux from

drainage layer to substrate

Moisture balance observation in controlled climatic

condition (35 °C and 20% relative humidity) through

specifically designed trays

Dra

ina

ge

lay

er

Detention 5×1 m rainfall simulator to test 4 drainage layer

components vs rainfall intensities, roof length, roof slope.

Retention/detention of

novel component

Hydraulic tests of different detention enhancing devices

Co

mp

lete

syst

em

Field test beds Rainfall – runoff – climatic conditions and moisture

content vertical gradient monitored in 9 test bed

configurations (3 substrates and 3 vegetation options)


WP3 is focused on studying the complete green roof system by using the knowledge developed

in the previous phases of the project. To achieve this objective the combination of green roof

elements that provided the most promising performances as well as innovative elements will

undergo a second field monitoring program at the University of Sheffield. Furthermore, in WP3

we are implementing a physically‐based hydrological model specific for green roofs, validated

on experimental data acquired through the field tests. The model will also be used for

simulating the impact of green roofs on a catchment scale.

Analysis

The integrated multidisciplinary method

The adopted research approach is represented in Figure 2.

Time

Mon 07 Mon 14 Mon 21 Mon 28

Flo

wra

te [

l/m

in]

0.0

0.1

0.2

0.3

0.4

0.57/5/2012 14/5/2012 21/5/2012 28/5/2012

Rai

nfal

l Int

ensi

ty [

mm

/h]

0

20

40

60

80

100

Vol

umet

ric

Wat

er C

onte

nt [

m3 /m

3 ]

0.0

0.1

0.2

0.3

0.4

0.5

Top WCR Mid WCR Low WCR

Pore space

distribution

ET test

Pressure plate

extraction

June 2010

Hadfield roof

Measuring

Interpreting

Enhancing

Test beds

Hadfield roof

Components

Physical

characteristics

Processes Complete

system

Moisture Content in test beds

Figure 2 Scheme of the integrated research approach adopted in the project.


A deeper knowledge of each element’s performance, as well as of the combination of elements

(complete system), allows interpretation and understanding that can result in enhancing the

system by new product development or more effective combination of traditional elements.

The described process restarts from the testing of the new or more effective solutions. The

adopted approach requires a combination of expertise including horticulture, plant ecology,

plant physiology, hydrology, civil engineering and the collaborative partnership between

academia and the industry. Also the duration of the project allows the possibility to collect

representative data and when needed to repeat series of tests after more promising solutions

have been determined.

Selected outcomes from the project are briefly highlighted below.

Plant screening programme. The main result of WP1 is the rigorous characterization of plant

species and their performance. Preliminary analysis of the data collected during the first

growing season in the UK showed that leaf extension growth response is influenced by depth of

the rooting medium (Figure 3). However there is clear differentiation at this growth stage

between horticultural plant groups, with grasses showing a greater relative increase in growth

with increasing substrate depth if compared to succulents behaviour. For many species under

high water availability a significant drop in shoot extension was observed due to a restriction of

resources by competition from neighbour plants.

Figure 3 Leaf length classified by plant categories (forbs, grasses and succulents) and growing

medium depths measured after the first growing season at the University of Sheffield site

Substrate detention tests. Preliminary tests (Yio et al., 2012) have identified the effects that

substrate depth and composition (type and percentage of organic matter) have on runoff

detention. In Figure 4 the laboratory small scale rainfall simulator is shown and the measured

cumulative runoff of 6 substrate options. Laboratory data has shown that the detention in

green roof substrates increases as a function of depth and organic matter content. The latter is

associated with a reduction in permeability. A modified reservoir routing model was used to

simulate the detention process. The model parameters are largely independent of rainfall

intensity, and it appears feasible to predict them from known physical characteristics of the

substrate, specifically its depth and permeability.


Figure 4 Small scale rainfall simulator for vertical detention tests and measured runoff of

different substrates characterized by different organic material (coir and composted bark)

and organic content at 0.10 mm/min inflow rate (Yio et al., 2012)

Drainage layer detention tests. Different drainage layers have been tested in a 5 × 1 m rainfall

simulator (Figure 5). Preliminary results showed a moderate detention effect with no

significance difference between different drainage layers, while an increased detention effect

was observed with the use of a moisture mat combined with the drainage layer. A runoff model

based on storage routing and a power‐law relationship between storage and runoff and

incorporating a delay parameter were created. A sensitivity analysis showed the influence of

roof slope and drainage material. More details are reported elsewhere (Vesuviano and Stovin,

2012).

Figure 5 Rainfall simulator for horizontal detention tests and hydrographs of five drainage

component configurations at a roof slope of 1.15o, drainage length 5 m and inflow rate

0.6 mm/min (Vesuviano and Stovin, 2012).

Phytometer tests. Preliminary results of the Phytometer tests support the choice of substrates

containing 15% of organic matter (bark or peat) and inert material (bricks and pumice) to

increase the maximum water holding capacity while maintaining the permeability above the FLL

target. Pressure plate extraction results (pF curves) showed a large remaining water reservoir at

‐15 bar extraction pressure (wilting point). Results showed that no correlation was found

between the FLL water holding capacity values and the pressure plate extraction


measurements. Also a higher maximum water holding capacity was not reflected in prolonged

plant survival or increased aesthetics (Emilsson et al., 2012)

Evapotranspiration test. Poë and Stovin (2012) quantified the net contribution to ET losses

during typical UK spring and summer conditions for the two plant species Sedum and Meadow

Flower. The latter showed a higher contribution in restoring the available water content during

spring conditions and in the first days of summer. In summer after circa 14 days a change in

behaviour was observed due to drought stress condition. Results are explained in relation to

climatic condition and substrate characteristics, thus providing information on the impact of

drought‐resistant species over more aesthetic ones on the hydrological process in an extensive

green roof.

Field test of green roof systems. Clear differences in runoff retention have been observed

dependent upon substrate characteristics and plant species choices as shown in Figure 2 where

cumulative retention for the month of June 2010 is reported for the 9 green roof system

configurations tested at the University of Sheffield (Poë et al, 2010).

The enhanced system

The described results obtained in WP1 and WP2 are used to define the green roof

configurations to undergo a second field test and to develop a hydrological model specific for

green roofs (WP3). As for the field test, different vegetation mixes with higher aesthetic value

will be compared to traditional ones. A moisture mat will be included in the tested systems

combined with drainage layers. A novel drainage layer that favours moisture flux to the

substrate, thus functioning as reservoir for plants during drought, has been developed and will

be tested. A novel ‘slow draining’ system specifically designed to enhance the detention effect

will be tested as alternative to traditional drainage layer. The reference substrate will be

compared with the two substrates that provided the best performances in the retention and

phytometer test. Also, the new configurations will provide novel data for rainfall‐runoff model

validation.

Results and Business Impacts

Key Findings

This paper presents a research approach that aims to investigate each element of traditional

green roof system to quantify their impact in the hydrological processes occurring on a green

roof. From a deeper knowledge of the performance of traditional element this project explore

potentials of enhancing these systems by developing novel elements but also by testing

combination of elements in the field as well as in the laboratory. A rigorous characterization of

plant species was conducted that will provide essential information on the design of a green

roof to meet specific criteria. Different tests highlighted the need for more detailed ways of

measuring substrate characteristics such as permeability and maximum water holding capacity.

Other studies (Fassman and Simcock, 2012) confirm the growing interest in using alternative

test methods to FLL ones. One of the objectives of this project is to develop a hydrological

model specific for green roof. Model approaches were proposed within the project for the

simulation of the detention effect of the drainage layer and the substrate. In developing a

model for long term simulation of a green roof system a key challenge will be to describe the


changes in the system over time. Green roof systems are living elements and changes in

vegetation as well as in the substrate due to organic matter decomposition and vegetation

establishment (especially for plant species supporting the increase biodiversity demand) are

expected. The first phase of the monitoring program investigating complete systems at the

University of Sheffield (Poë et al, 2010) will end in October 2012 after more than 2 years of

continuous data collection. There will be then the opportunity to conduct a series of test that

aim at investigating these changes, especially in the substrate physical characteristics.

Business Impacts

This research intends to provide a better understanding of the processes occurring in green

roof systems and on the influence of each specific element. A rigorous characterization of plant

species performance as well as substrates and drainage layers can lead to the design of more

efficient systems or to the development of new products. Novel design of drainage layers and

combination of elements will be presented at the end of the project as well as a hydrological

model specific for green roof systems.

Conclusions

The University of Sheffield Green Roof Centre, together with ZinCo GmbH, is involved in the

project “Collaborative Research and Development of Green Roof Systems Technologies” that

aims at enhancing traditional intensive and extensive green roof systems by revisiting the

fundamental basis of green roof system design. This paper presents the integrated

multidisciplinary approach adopted in this research and highlights preliminary findings of this

project.

Key Lessons Learned:

• A multidisciplinary research approach is the key to address the new challenges of green roof

system design

• Integrated physical property measurement of each element and laboratory and field tests of

single elements and complete systems can lead to hydrological modelling specific for green

roofs

References

Brenneisen, S. (2006). Space for Urban Wildlife: Designing Green Roofs as Habitats in

Switzerland, Urban Habitats, 4, pp. 27‐36.

Compton, J.S., and Whitlow, T. (2006) A Zero Discharge Green Roof System and Species

Selection to Optimize Evapotranspiration and Water Retention, Proceedings of Greening

Rooftops for Sustainable Communities, Boston, MA, May, 2006.


Emillsson, T., Berretta, C., Walker, R., Stovin, V., and Dunnett, N. (2012) Water in Green Roof

Substrates – Linking Physical Measurements to Plant Performance. World Green Roof

Conference 2012, Copenhagen, 19 ‐20 September, 2012.

Fassman, E.A., and Simcock, R. (2012) Moisture Measurements as Performance Criteria for

Extensive Living Roof Substrates. Journal of Environmental Engineering, 138, pp. 841‐851.

FLL (Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau) (2008) Guidelines for the

Planning, Execution and Upkeep of Green‐roof Sites.

Kasmin, H., Stovin, V. and Hathway, E. (2010) Towards a generic rainfall‐runoff model for green

roofs. Water Science & Technology, 62, 4, pp. 898‐905.

Oberndorfer, E., Lundholm, J., Bass, B., Coffman, R., Doshi, H., Dunnett, N., Gaffin,S.,Köhler, M.,

Liu, K., and Rowe, B. (2007) Green Roofs as Urban Ecosystems: Ecological Structures,

Functions, and Services, BioScience , Vol. 57,10, pp. 823‐833.

Poë, S., Stovin, V., and Dunsinger, Z. (2011) The Impact of Green Roof Configuration on

Hydrological Performance, Proceedings of the 12th International Conference on Urban

Drainage, Porto Alegre/Brazil, 11‐16 September, 2011

Poë, S., and Stovin, V. (2012) Advocating a Physically‐based Hydrological Model for Green

Roofs: Evapotranspiration during the Drying Cycle. World Green Roof Conference 2012,

Copenhagen, 19 ‐20 September, 2012.

Vesuviano G and Stovin V. (2012) A Generic Hydrological Model for a Green Roof Drainage

Layer, Proceedings of the 9th International Conference on Urban Drainage Modelling

(9UDM), Belgrade, Serbia, 3‐7 September, 2012.

Yio, M.H.N., Stovin, V., and Werdin, J. (2012) Experimental Analysis of Green Roof Detention

Characteristics, Proceedings of the 9th International Conference on Urban Drainage

Modelling (9UDM), Belgrade, Serbia, 3‐7 September, 2012.

Documents

Towards an Enhanced Green Roof System