ChapterPDF Available

Abstract and Figures

Green roof monitoring is critical to understand and improve the design, implementation, and management of green roof ecosystems. Creating resilient, less resource intensive living roofs fitting their larger eco-regional context, specific local setting, and unique project objectives means understanding inputs and outputs. This chapter addresses monitoring abiotic inputs and outputs related to green roof hydrology (precipitation and irrigation, storage, outflow, and evapotranspiration), water quality, energy fluxes, temperatures, meteorological conditions (wind), and gas/carbon exchange. This chapter presents monitoring approaches and equipment needs from literature and researcher interviews detailing several relevant examples. Important design, educational, and management opportunities relating to effective monitoring programs are discussed.
Content may be subject to copyright.
Lee R. Skabelund Kimberly DiGiovanni Olyssa Starry
302 Seaton Hall Villanova University Portland State University
Kansas State University USA Portland, Oregon, 56478 USA
Manhattan, KS 66506-2909 USA
lskab@k-state.edu
2 Monitoring Abiotic Inputs and Outputs
Lee R. Skabelund, Kimberly DiGiovanni & Olyssa Starry
Abstract Green roof monitoring is critical to understand and im-
prove the design, implementation, and management of green roof
ecosystems. Creating resilient, less resource intensive living roofs
fitting their larger eco-regional context, specific local setting, and
unique project objectives means understanding inputs and outputs.
This chapter addresses monitoring abiotic inputs and outputs related
to green roof hydrology (precipitation and irrigation, outflow,
evapotranspiration, and storage), water quality, energy fluxes, tem-
peratures, meteorological conditions (wind), substrate characteristics
and gas/carbon exchange. It presents monitoring approaches and
equipment needs from literature and personal communications with
researchers; several related case studies are also detailed. Discussion
also focuses on how important design, educational, and management
opportunities relate to effective monitoring programs.
2.1 Introduction
Intensively managed ecosystems (including regularly tended residential yards and
urban gardens and parks) generally follow a high input, high output model requir-
ing relatively large subsidies of time, substances, energy, and materials such as la-
bor, water, and nutrients while frequently shedding stormwater and contributing
various pollutants directly and indirectly to the environment (Arvidson 2012).
Because any ecosystem leaks or exports nutrients, materials, and energy (Vitousek
& Reiners 1975), inputs often become waste outputs and potentially degrade or
harm their surroundings, particularly surface water features, air quality, and ambi-
ent temperatures (Odum 1969, Oke 1978, Spirn 1984). We must and can create
“sustainable urban social-ecological systems” (Byrne & Grewal 2008 p. 1) that in-
clude green roofs and other living infrastructure to minimize harm.
What benefits and concerns are related to green roof ecosystem inputs and out-
puts? What is known about such inputs and outputs?
2
Stormwater outputs often have financial impacts so the quantity of water stored,
and its timed release, must be reconciled with precipitation intensity, as well as
substrate and vegetation characteristics. Many urban areas now closely monitor
stormwater runoff (for one example, see Kurtz et al. 2010). How does one monitor
and account for the constituents that drain or flow from a green roof? How do the-
se flows compare among rooftops?
For heat attenuation, assessing the insolation and heat flows on and within a living
roof system are important. How do substrate and vegetation type influence hy-
drology and microclimate? And, how do plant growth above and within a sub-
strate, the production of new vegetation from seeds, and human management
strategies influence energy flows, hydrologic processes, climatic conditions, and
the creation of a living, supportive substrate? Teasing apart these complex interac-
tions relies first on solid qualitative and quantitative data about each. This chapter
touches on some of these concerns.
Researchers, designers, and managers must account for constituents are retained or
leave a green roof. They need to ask: What current methods and equipment are be-
ing used to measure and analyze inputs and outputs? How are the data analyzed
and used to improve green roof design, monitoring, and management?
This chapter begins by defining green roof ecosystem inputs and outputs in the
following section. Next, a general overview of green roof monitoring is provided.
The content of this chapter focuses primarily on monitoring the inputs and outputs
associated with water (hydrologic factors) and energy fluxes from green roof sys-
tems; additional information is also provided on substrate characterization and
gas/carbon exchange. Throughout the chapter, brief case studies are presented
demonstrating green roof monitoring applications related to research, design and
management goals. Challenges and lessons learned from green roof monitoring
across the United States and other countries are presented. Opportunities for the
future of green roof monitoring are also discussed.
2.2 Defining Inputs and Outputs
Since many interconnected variables exist in every ecosystem (both natural and
created) and researchers cannot feasibly monitor everything, so they need to clear-
ly define what is to be monitored and why. Monitoring specific abiotic inputs and
outputs is vital to our understanding of the interactions and functions associated
with both biotic and abiotic conditions on a green roof.
For this chapter, we define inputs as substances and energy added to a green roof
(for example, water in the form of precipitation and irrigation, added nutrients,
and energetic inputs that include sunlight and wind energy). Outputs are defined
3
as substances and fluxes modified on or leaving a green roof (for example, the
outflow of water nutrients in substrates or runoff, evapotranspiration, and heat en-
ergy). Inputs and outputs from green roof systems are generally conceptualized in
Figure 2.1.
Fig. 2.1 Green Roof Inputs and Outputs Model (Olyssa Starry and Rich Pouyat, unpublished).
Closely examining frequently stated “green roof benefits” related to urban heat
loads, energy use, stormwater management, and carbon sequestration can help re-
searchers effectively monitor the dynamic conditions and changes that influence
important green roof attributes and functions. Thus, we monitor those conditions
and factors related to optimizing energy and water inputs, and reduce or eliminate
negative outputsnamely, excess heat, carbon dioxide, nutrients, and heavy met-
als. In doing so we can better understand how to create low input (water and ener-
gy conserving), and low output (less runoff and non- or minimally-polluting)
green roof ecosystems. Our ability to sequester carbon and achieve other relevant
project goals can also be enhanced.
Of course, this approach to evaluation is only one perspective on green roof inputs
and outputs. Some, but not all, inputs are related to management and presumably
should be minimized. Similarly, some, but not all, outputs are pollutants that
should be minimized. For example, carbon input, would ideally be maximized as
sequestration.
Further, ways to monitor green roof inputs and outputs can vary. “Some monitor-
ing is directed at the inputs and outputs themselves (i.e. the energy and material
4
fluxes in the system). Some monitoring is directed at describing processes (such as
evapotranspiration or microbial activity) that drive those fluxes.” Other monitor-
ing is directed at physical conditions (such as temperature or wind speed) which
may directly or indirectly influence fluxes and other aspects (John Lambrinos,
May 2014, pers. comm.)
2.3 Planning for Green Roof Monitoring
It is impossible to learn from implemented green roofs without closely observing
and understanding monitoring goals and objectives. This section discusses: 1)
monitoring approaches and goals in light of the needs and demands associated
with observational studies versus experimental monitoring designs; 2) expecta-
tions regarding equipment and maintenance; 3) data collection and analysis and
the technical expertise needed to successfully undertake effective green roof sys-
tem monitoring; and 4) monitoring precautions.
2.3.1 Monitoring Approaches and Goals
The most important part of any research project and accompanying monitoring
process is articulating specific needs, goals, and objectives of the study. To do this
requires a reasonable understanding of the available literature and the project con-
text (including the type of site, regional and local setting, and funding, expertise,
personnel, equipment, and other necessary support systems).
Relating monitoring to the specific type of site(s) and study under consideration
(before deciding what type of monitoring to undertake) is essential. Will monitor-
ing be done on an existing green roof, on a new or proposed green roof, or on
models, mock-ups, modules, or platforms? Will monitoring activities examine in-
tegrated or modular systems, or both?
It is also vital to understand conditions associated with the particular green roof
study system of interest under consideration. Rainfall, temperature, relative hu-
midity, wind speed, wind direction, and solar radiation all change seasonally, and
can be affected by surrounding buildings, structures, and vegetation. For building
retrofits, “data collected before renovation can be a valuable measure of the new
green roof’s performance” (Onset 2012, p. 3).
An example of the types of monitoring questions appropriate at this stage of the
design process would be: How hot and cold does this location get? How do adja-
5
cent building masses influence sun/shade patterns and wind movementsand thus
precipitation, relative humidity, temperatures, and other conditions on the roof?
An approach to monitoring needs tempering by our purposes and goals and it will
be different if our primary interest is in improving design and management (as op-
posed to simply trying to better understand ecosystem functions on a green roof).
For example, our methods and analysis will likely be different if we are looking
for a trend (over time and space) or specifically trying to address a narrower ques-
tion by comparing the effects of a treatment against a control (Karban and
Hunzinger 2006).
Based on the goal and objectives of a project, the level and intensity of monitoring
activities can vary. Welker et al (2013) describes a three-tiered (low, medium,
high) approach to monitoring. It provides a framework for monitoring allowing for
a balanced approach between project objectives and monitoring costs. For exam-
ple, a low level approach for monitoring the hydrology and ecology of a green
roof might include visual inspections while a high level approach would include
the use of a sensor system collecting continuous data (Welker et al. 2013).
Generally, two overarching approaches, observational monitoring and experi-
mental research/monitoring can describe green roof monitoring. Each of these ap-
proaches may include qualitative and quantitative data collection and analysis, and
these approaches (with their accompanying monitoring activities) can be carried
out simultaneously to address specific green roof research questions, hypotheses,
and/or practical design and management issues.
Observational monitoring studies rely on systematic collection, recording, and
analysis of relevant data over some period of time. Observation of green roof con-
ditions in space and time are frequently and systematically recorded to document
changes, dynamics, and particular conditions for selected variables that are likely
to address research questions of interest. Hand-written notes regarding observa-
tions accompanied by photographs taken from the same locations through time
supplement data collected using a number of different monitoring devices and
equipment. In addition, one or more control rooftops may also be monitored so
that comparisons can be made between the green roof and nearby black, dark-
gray, and/or white or light-colored roofs, as time and circumstances afford.
Experimental research/monitoring systematically focuses on one or more of the
following: vegetation types, substrate depths, substrate types, roof slopes, micro-
climatic conditions, supplemental irrigation, nutrients added, shading devices, etc.
Experimental monitoring also relies on systematic collection, recording, and anal-
ysis of relevant data over a period of time, but includes a manipulation or compar-
ison targeting a specific research question and hypothesis. Standard statistical de-
signs and protocols can lead to inference of significance about the data. Also,
other researchers must be able to replicate methods. Researchers often need to
6
balance the need for replicated treatments in their design against the feasibility of
including multiple roof-scale measurements in their experiment. Depending on the
research question, working with modules or experimental containers may be an
option. All such designs require consultation with a statistician beforehand. As
with observational studies, one or more control rooftops may also be monitored
for comparison.
Experimental research requires more statistical rigor (i.e., replication) than obser-
vational work. Observational work requires less replication but is also less gener-
alizable. Neither approach is better than the other, but the distinction is important.
Often the two approaches are integrated. Someincluding Tilman 1989, and Ha-
vens and Aumen 2000argue that these two approaches need to be integrated.
Again, it is important to note that green roof monitoring and data collection can
done to support green roof management, to evaluate performance relative to par-
ticular green roof project goals or models, or to collect data as part of the process
of testing specific hypotheses about how green roofs function.
2.3.2 Monitoring Equipment and Maintenance
Monitoring equipment can be as simple as a hand-held thermometer, manual rain-
gauge, and container to collect stormwater runoff, or as complex as a series of tip-
ping buckets, dozens or more temperature sensors, and multiple flow sensors all
connected to one or more data logger(s) and a satellite-operated wireless data dis-
tribution network.
Great value comes in using basic probes or sensors to take repeated measurements
of green roof systems. Although single measurements of variables such as sub-
strate temperature or moisture content are only snapshots in time of systems con-
stantly in flux, if samples are taken at a regular intervals over a long time scale,
overall rates, trends, and patterns can be discerned.
Spot measurements are especially useful for constructing experimental designs so
that comparisons between two different green roof treatments can be made. These
types of measurements can be conducted in the context of field or classroom vis-
its. They can also be incorporated into regular monitoring system maintenance and
green roof management.
Commercial green roof monitoring systems can be coupled with a weather station
and include temperature sensors above and below the roof deck. Data from these
systems can be accessed in real time. In some instances all weather stations, sen-
sors, and loggers can be ordered from a single established company.
7
Complex monitoring networks may combine educational and commercial systems
to collect continuous data with increased replication (Lea-Cox 2014, pers. comm.).
A larger roof area can also be sampled to account for variation. Being able to ac-
cess data remotely is also important for roofs on which safety or security concerns
limit frequent visits. Remote access enables roof data to be visualized during
storms from the safety (and comfort) inside. Sensor information can be transferred
to a network, computer, or handheld device synchronously, making data more ac-
cessible in the classroom or laboratory in real time (see Figure 2.2).
In addition to educational and research opportunities, green roof monitoring sys-
tems document green roof performance to improve green roof design and man-
agement and promote adoption. In a survey of architects and building managers in
Chicago and Indiana, Hendricks and Calkins (2006) identified ways that designers
can increase public understanding of the benefits associated with new or innova-
tive green roof practices. They noted that recognition of the environmental ser-
vices provided (as revealed by monitoring) is a key incentive for early adopters.
As such, monitoring is seen as vital to the “green building” certification process,
particularly since LEED™ (Leadership in Energy and Environmental Design) and
other rating systems require pre- and post-implementation monitoring, and have
the potential to improve design, implementation, and management while also
deepening our collective knowledge about green roof dynamics and functioning.
2.3.3 Data Collection
Depending on the nature of the study, data collected for abiotic green roof condi-
tions and processes may be documented directly by a researcher in a hand-written
notebook or portable device or computer—and/or the data may be wirelessly sig-
nal-fed to a data-logger/computer or physically transmitted via wires/cables linked
to a data-logger/computer. For micro-meteorological data, weather stations can be
installed in association with each module or platform, or positioned in a reasona-
ble location related to an integrated green roof and/or any number of green roof
modules/platforms. Figure 2.2 depicts a type of networked wireless sensor system
that could be developed and deployed to assist with data collection for green roof
research.
2.2.4 Data Analysis and Technical Expertise
Technical expertise needed for green roof monitoring typically includes personnel
familiar with scientific research methods (especially quantitative approaches and
tools) and statistical analysis (especially when more sophisticated modeling and
8
analyses are required or desirable). Familiarity with rigorous, systematic data col-
lection and analysis procedures and the ability to trouble-shoot equipment or de-
vice failures is vital. For example, in order to provide useful measurements, soil
moisture sensors need to be placed properly and calibrated appropriately in order
to give accurate readings for green roof substrates (Starry 2013). Some commer-
cial organizations provide guidance specific to green roofs regarding weather sta-
tion selection, logging capacity, configurations, setup, data download, and de-
ployment options; sensor placement and positioning for different sensor types and
purposes; and sensor cable protection, weather station grounding, battery mainte-
nance, and sensor calibration (Onset 2012).
Fig. 2.2 A Wireless Sensor Network System supports green roof monitoring and facilitates real
time data collection and analysis (Adapted with permission from Lea-Cox 2012).
2.2.5 Monitoring Precautions
In any monitoring scenario the unique nature of green roofs needs to be consid-
ered. Safety precautions include fall awareness and fall protection training for any
person maintaining or otherwise walking on a green roof in situ (Omar et al. 2013)
and these pertain to monitoring as well.
Additional care should also be taken not to disturb the green roof system (when
studying an actual green roof system), roofing membrane, or monitoring devices
and equipment while on the roof or working in the vicinity of modules or plat-
9
forms. To this end all green roof monitoring activities should be dated and record-
ed as to their time and purpose in a logbook or other recording device.
2.3 Topics of Green Roof Monitoring
Different types of data collection approaches are used to quantify inputs, trans-
formations, outputs, as well as the factors that control them. Data can be collected
and evaluated on an established or newly implemented green roof (in situ) and/or
on modules or microcosms located on top of a roof (or constructed on platforms
established as experimental prototypes on the ground or roof).
In this section, we discuss monitoring:
Amount and frequency of precipitation and irrigation
Substrate moisture and interception
Water outflows (stormwater runoff)
Evapotranspiration (ET)
Water quality
Surface energy balance (including latent heat fluxes)
Temperature dynamics (surfaces and sub-surfaces)
Wind speed, direction, and dynamics
Substrate attributes and changes
Gas exchange and carbon sequestration
2.4 Monitoring Hydrology
This section addresses: 1) monitoring of water inputsamount, rate, and duration
of precipitation and supplemental irrigation; and 2) monitoring of water outputs
(quantity and quality of green roof runoff, and evapotranspiration rates). The wa-
ter balance is important to consider in evaluating green roof hydrology and system
performance. The water balance of a green roof is represented schematically in
Figure 2.3 and given as 𝑃𝐸𝑇 !𝛥𝑆 =𝐷 where P is precipitation (or irrigation),
ET is evapotranspiration, ΔS the change in storage, and D is drainage (outflow
assuming no surface runoff) all defined as volume flow rates.
10
2.4.1 Water Inputs: Precipitation and Irrigation
Water inputs to green roof systems include precipitation and irrigation. Water in-
puts are important considerations for green roof performance. The amount, rate,
and duration of water inputs to green roofs are dictated by precipitation patterns,
which vary geographically and temporally, and the presence of supplemental irri-
gation schemes. Water inputs can be monitored through various means: tipping
bucket rain gauges (heated or non-heated), non-recording rain gauges, plate
Fig. 2.3 Green roof water balance from DiGiovanni (2013) with permission from ASCE. Note:
Water drainage is often referred to as runoff or outflow since it leaves the green roof ecosystem.
gauges, weighing type rain gauges, radar rain data, and flow meters. For example,
the amount and frequency of precipitation can be monitored using a manual rain
gauge or by using a tipping bucket connected to a data-logger to record rainfall
and snowmelt at some pre-selected interval ( e.g., every five to 15 minutes). While
a nearby weather station can be used to estimate onsite precipitation, spatial varia-
bility in weather patterns and structural differences (e.g. if the green roof falls in
the lee of other buildings) warrant onsite measurement of precipitation.
Irrigation inputs are often measured with flow meters, but this is often challenging
if more than one water delivery pipe is active. Irrigation needs and demands can
be measured over time and an observational and/or experimental design employed
to examine e.g. water conservation practicesfrom rooftop water harvesting and
re-use, to drip and sub-surface irrigation, to the complete elimination of irrigation
(see Chap. 4).
Green roofs maybe observed every few days, and when vegetation shows signs of
wilting or browning, irrigation can be applied for a specified amount of time
and/or quantity of water. In other cases, soil moisture sensors, linked to programed
irrigation systems, can provide supplemental water.
Irrigation not only impacts roof performance (biomass production, vegetative
health, and summertime cooling potential), but it can strongly influence substrate
11
moisture and stormwater drainage monitoring results (Fassman-Beck et al. 2013).
For example, by taking up retention capacity and generating runoff, irrigation can
become a disservice. Thus, precipitation and irrigation must be accounted for
when preparing green roof monitoring plans and analyzing collected data.
The Upper Seaton Hall Green Roof (see case study) focuses on monitoring vegeta-
tion, substrate temperatures and soil moisture, micro-meteorological conditions,
and other factorswhile seeking to understand the influences of irrigation (and
non-irrigation) on this integrated living roof system in the central Great Plains.
2.4.2 Water Storage: Substrate Moisture and Interception
Green roof water storage can be achieved in the substrate and drainage layers as
well as on the green roof vegetation through canopy coverage and to a lesser de-
gree through internal plant storage. Monitoring water storage on a green roof pro-
vides important information regarding both stormwater management benefits as
well as information relevant to water conservation/irrigation and plant survival.
Per event, the stormwater retention of a green roof is dictated by the storage ca-
pacity and also by intra-storm ET (which restores water storage capacity.) Intra-
storm evapotranspiration (though often assumed negligible) can be an important
mechanism particularly for low intensity long duration storm events (DiGiovanni
et al. 2010). This section focuses on monitoring substrate storage and also discuss-
es interception and intra-storm ET.
Volumetric moisture content sensors of various types quantify substrate storage.
Substrate moisture levels can be measured via sensors in the substrate or modeling
based on local and/or regional climatic conditions and trends. Modeling and anal-
ysis of substrate moisture levels through time is possible.
According to John Buck, who has worked on a number of green roof monitoring
projects with Civil & Engineering Consultants, Inc., many monitoring networks
include Decagon soil moisture volumetric water content sensors. The sensors have
a built-in calibration, but can be specifically calibrated to individual substrate
compositions. These sensors measure volumetric water content based on the die-
lectric permittivity, or the tendency of water to polarize in an electric field. A
number of sensors with different capabilities are available; particular attention
should be given to the volume of soil to be interrogated by each type of sensor.
Some sensors may be used for deeper roofs and others more appropriately in shal-
lower green roof substrate (John Buck, July 2014, pers. comm.).
Soil moisture data can help with water balance calculations. The sensors can also
help detect field capacity, or excess water held after initial drainage. Researchers
attest that it is best to take these measurements in the field; substrate disturbance
12
Case%Study:%Supplemental%Irrigation%Interrelationships%with%Vege8
tation,%Substrate,%and%Micro8Meteorological%Conditions%(Kansas%
State%University%Research%8%Manhattan,%Kansas)%
The Upper Seaton Hall Green Roof in Manhattan, Kansas relates plant suc-
cess to climatic factors, substrate, and irrigation in the wind-swept Great
Plains.
Along with plant growth and survival, precipitation, irrigation, stormwater runoff
and meteorological variables are monitored on this south-facing (full-sun) green
roof in the Flint Hills Eco-region. The primary objective of the project is to deter-
mine how selected native grasses and forbs fare with varying substrate depths (7.5
to 17.5 cm) with and without supplemental irrigation. The east side of the green
roof received supplemental irrigation for four years, the west side for three years
(with no irrigation provided in 2011). In mid-August 2012, supplemental irrigation
was entirely discontinued. Plant survival and growth are monitored annually. A
Campbell Scientific data-logger records air temperature, relative humidity, green
roof surface and sub-surface temperatures, rainfall, and wind speed and direction
every five minutes. Runoff is also measured. Instruments installed in June 2009 on
the green roof (Fig. 2.4) included a Campbell Scientific (CR23X micro-logger)
and associated BP solar panel (10W 16.8V); RM Young weather station; three
surface temperature probes; six sub-surface temperature probes; and one Texas
Electronics (TR-525I) tipping-bucket. Hard-wired cables connect instruments and
sensors to the data-logger. The Kansas State Climatologist loaned and installed
most monitoring equipment. The manual, all-weather rain gauge (Productive Al-
ternatives) was placed at the south end of the green roof near the tipping bucket. In
2013, one additional 5TE Temperature/Soil Moisture Sensor (purchased from
Decagon Devices) and one U23-002 Hobo Pro v2 Ext Temp/Relative Humidity
Data Logger (Onset) were added to better understand stresses on vegetation fol-
lowing the elimination of irrigation.
Fig. 2.4 Kansas State University’s Upper Seaton Hall Green Roof (a) has been monitored since
May 2009. Campbell Scientific CR23X micro-logger (b) (Lee R. Skabelund)
13
can affect lab results. An initial watering-in period of a few days after placement
makes sensor readings more consistent by removing air pockets around a truly
embedded sensor (Griffin 2013). Sometimes, erroneous data from volumetric wa-
ter content sensors can arise. Erroneous data can be attributable in some cases to
faulty sensor placement (see Figure 2.5 for the typical placement of a Decagon
soil moisture sensor). Figure 2.6 shows a Decagon data logger.
John Buck (July 2014, pers. comm.) found that sensors near the drainage layer are
nearly always very moist (especially in rainy and/or irrigated locations). The more
dramatic changes in soil moisture (quite moist to very dry) happen closer to the
surface. Using multiple sensors at different levels within the profile is one option,
but sensors should be about 20cm apart so they do not interfere with one another.
Per John Lea-Cox (2014, pers. comm.) the University of Maryland’s green roof
research has specifically targeted contributory factors related to stormwater runoff.
Substrate composition and depths, choice of plant species, vegetative metabolism
(e.g. C3 vs. CAM), and regional environmental conditions (especially rainfall fre-
quency and intensity) are critical green roof design factors related to stormwater.
Lea-Cox noted that their monitoring and modeling work has primarily focused on:
1) quantifying substrate moisture content by sensing water flux between rainfall
events; 2) modeling environmental fluxes using modified Penman-Monteith equa-
tions; and 3) assessing plant species effects, including differences in water use
over time.
Determining variabilityespecially in the root zoneis important as it allows re-
searchers to simplify sensor installations (i.e. right sensor, right place). Variability
is naturally linked to understanding the influence of vegetation and substrate types
on soil water retention (or matric potential) and hydraulic conductivity. Decagon
Devices data loggers (see Figure 2.6) are seen by Lea-Cox and others as having
intuitive, easy-to-use software for green roof monitoring (i.e. no programming
skills requiredessentially “plug and play”).
Fig. 2.5
Decagon soil
moisture sensor
set to sense mois-
ture top-to-bottom
of 7-12 cm sub-
strate profile
(Olyssa Starry and
Liz Ensz, un-
published)
14
Interception and intra-storm ET are gaining increased research focus; some re-
searchers seek to quantify the interception capacity of green roof vegetation
through laboratory techniques under simulated rainfall conditions (Fassman and
Simcock 2009, Rostad et al. 2011) and evaluation of methods to determine intra-
storm ET.
2.4.3 Water Outputs: Quantity
Green roof outflow (i.e., drainage and/or runoff) constitutes the liquid water leav-
ing a green roof system during or immediately following a storm event. It can be
monitored directly by several techniques and used to understand green roof water
retention in relation to precipitation and evapotranspiration. Some of the instru-
ments utilized for the measurement of green roof outflow include in-line flow
measurement devices and large capacity tipping bucket gauges, as well as flumes,
weirs and cisterns with accompanying water level or mass measurement devices.
In some instances custom designed and constructed devices are utilized for moni-
toring green roof outflow, for example, the orifice restricted device (ORD) devel-
oped by Voyde et al. (2010).
Stormwater outflow or runoff can be measured by channel/gutter systems to route
water to a tipping bucket, flume, cistern (which requires regular emptying), or
other measuring device in association with water level sensors or a scale (in the
case of cisterns). Measurements can be recorded manually or using a datalogger to
find the amount of green roof outflow for a storm event.
For full-scale systems, the total amount of green roof water retention can be simp-
ly quantified by subtracting the amount of outflow from the total estimated precip-
itation falling on the green roof. Inline flow meters typically provide rates of out-
flow utilized for runoff hydrographs displaying flow rates versus time and
revealing outflow volumes. Also outflow (and thus stormwater retention) from a
Fig. 2.6 Decagon
Em50 Data Logger
with five sensor ports
and a communication
port to link directly to
a computer or other
suitable device (Lee
Skabelund)
15
green roof can be simplistically estimated by assuming that rain in excess of the
field capacity of the green roof substrate flows out or off of the system.
Utilizing various techniques, the outflow and stormwater retention performance of
green roofs is being evaluated by researchers worldwide. Of note, 66 percent of
publications on green roofs are from USA and EU representing primarily the per-
formance of green roofs in temperate environments (Blank et al. 2013).
Extensive documentation of the stormwater retention performance of green roof
systems supports that green roofs effectively reduce the volume of stormwater, at-
tenuate peak flows and increase the time to peak in comparison to conventional
roof surfaces (Charlesworth et al. 2013, Fioretti et al. 2010). Pioneering studies of
green roof stormwater retention and runoff reduction in the US were completed by
Berghage et al. (2009). Reviews of other early works are available in the literature
e.g. Berndtsson (2010). Continued research in the field has yielded published
works from various researchers including Carson et al. (2013), Fassman-Beck et
al. (2013), Rosatto et al. (2013), Song et al. (2013), Palla et al. (2012), Stovin et
al. (2012) and Voyde et al. (2010) (Chap. 4, Chap. 5) Roof-scale runoff or out-
flow is one of the most challenging aspects of green roof monitoring, so these sys-
tems need to be carefully designed and evaluated. The Portland, Oregon Hamilton
Eco-Roof (see case study) focuses on monitoring runoff and stormwater retention
at the roof-scale.
2.4.4 Water Outputs: Evapotranspiration (ET)
Evapotranspiration (ET) describes the combined effect of evaporation and transpi-
ration through loss of water from substrate as well as from vegetation. Monitoring
ET is important to inform irrigation practices and also because ET is linked to a
variety of green roof benefits. For example, ET restores the retention (water stor-
age) capacity of a green roof and the ability to capture stormwater and also pro-
vides other benefits including micro-meteorological regulation and carbon seques-
tration. ET achieved during dry days between storm events has the greatest
influence on green roof stormwater retention (Voyde, 2010).
Evapotranspiration can be monitored using sensors including weighing lysimeters
(see Fig. 2.8), and/or modeled capturing site specific wind, temperature, and other
meteorological conditions based on regional and/or local meteorological condi-
tionsand then related to substrate and vegetation type. Simple or sophisticated
analysis of ET patterns, dynamics, and changes through time are possible
(Chap.3).
Measurement of ET can be achieved using a variety of methods (Jensen et al.
1990). Methods for the measurement of ET include the use of lysimeters, soil wa-
16
ter depletion techniques, and energy balance. Weighing lysimeters are widely con-
sidered the only directly quantitative means to measure ET (Tanner 1967, Jensen
et al 1990; Rana and Katerji 2000; Xu and Chen 2005). Other ET measurement
techniques include eddy covariance and scintolometers. Nouri et al. (2012) re-
views ET measurement techniques for urban landscape vegetation, including
green roofs.
Measurement of ET can be costly and estimates of ET can be used. Estimates of
ET can be achieved by various techniques including temperature-based, radiation-
based, and combination-based equations. Penman-Monteith based combination
equations are widely recognized to yield the most accurate results in a wide varie-
ty of climates. Inputs required for the estimation of ET, depending on method, in-
clude solar radiation, temperature, relative humidity and wind speed monitored us-
ing pyranometer, temperature-relative humidity probes and wind sensors
respectively. Furthermore, the development of appropriate coefficients to adapt
ET estimates from reference conditions for green roof vegetation can be devel-
oped (Starry 2013, DiGiovanni 2013, Schneider 2011, Sherrard et al. 2012) as
well as evaluation of coefficients to account for substrate moisture conditions
(DiGiovanni et al. 2011, DiGiovanni et al. 2013).
Despite the importance of evaporative processes in managing stormwater and
providing other valuable ecosystem services, comprehensive monitoring and
measurement of ET from green roofs is rare. Reported studies measuring evapo-
transpiration from green roofs are limited to a small body of literature including
Berghage et al. (2009), Voyde et al. (2010), Sherrard et al. (2012), DiGiovanni et
al. (2013), DiGiovanni (2013), Starry (2013), and Wadzuk et al. (2013).
17
Case%Study:%Hydrology:%Water%Quantity%8%Precipitation%&%Storm8
water%Runoff%(City%of%Portland,%Oregon%Monitoring)%%
The Hamilton Eco-roof is an example of a long-term, roof-scale hydrological
monitoring study that is one type of designed experiment.
Two (7.62 cm and 12.7 cm depth) integrated eco-roofs (west 243 m2 and east 234
m2), each with a different substrate, were installed on the Hamilton apartment
building (autumn 1999) and monitored from 2001 to 2012. The eco-roof is divided
into two drainage areas (west and east), each with a separate roof drain. Both in-
clude areas of impervious patio and vegetation. A distinct substrate was used for
each area to compare the performance of thicker to thinner substrates. Fiberglass
flumes (Fig. 2.7) measure outflow from each section while a rain gauge measures
precipitation input (Kurtz et al. 2010, p. 18). A small, V-trapezoidal Plasti-Fab
flume is installed adjacent to, and immediately upstream of, each primary roof
drain. The primary roof drain is sealed and isolated to direct all flow through the
flume prior to entering the drain. An American Sigma Model 950 bubbler-type
flow meter is used to measure water level in each flume. Because of spillover and
other physical challenges “it is not unusual to measure more runoff coming from
the east side than the total rainfall that fell on the roof. This makes the use of the
east side data problematic…” (ibid, p. 18-19). Nevertheless, this technique, when
properly executed, yields valuable data. The flume setup has been verified to
measure a range of flows accurately from inter-event outflow to medium and large
events. According to city employees, these data are used to communicate useful
answers to questions about green roof performance, and improve policy and de-
sign guidance for other projects, including new green roof construction and green
roof retrofits (Timothy Kurtz April 2014, pers. comm.). Level data are converted
to flow values by using a formula created by manually establishing the level to
flow relationship specific to these flumes. Initial monitoring indicated that the
formula provided for the flumes by the manufacturer was not accurate enough for
this project so BES calculated a more accurate formula (Timothy Kurtz July 2014,
pers. comm.).
Fig. 2.7
Drain (a) and
flume (b) setup on the
Hamilton Ecoroof for
stormwater outflow
monitoring (City of
Portland, BES)
18
Fig. 2.8 Green roof lysimeter setup at the Utah Natural History Museum in Salt Lake City, Utah
(graphic by University of Utah & Natural History Museum of Utah; photo by L.R. Skabelund)
The University of Utah case study is an example of a project employing innova-
tive monitoring tools and techniques for ET monitoring on green roofs.
2.4.5 Water Output: Quality
The quality of water outflow from a green roof system is important an important
consideration of green roof performance. Green roofs can improve water quality
through filtration and adsorption in the substrate (Wang et al. 2013), plant uptake
of nutrients, and microbial action, though these processes are not well studied
(Dietz, 2007). Furthermore, the export of nutrients and other constituents based on
rainfall intensity (Teemusk and Mander 2007) and fertilization (DeCuyper et al.
2005) can cause concentrations in green roof outflow to exceed standards and/or
objectives set for receiving water bodies (Van Seters et al. 2009).
Water quality can be measured in relation to precipitation and rooftop runoff by
taking one or more samples from a single integrated green roof, control roof, or
from a series of experimental green roof modules set on platforms. Understanding
the base nutrient conditions of the substrate is critical (Chap. 5).
19
Case%Study:%Monitoring%Evapotranspiration%%
(University%of%Utah%Research%8%Salt%Lake%City,%Utah)
This green roof research program uses several innovative tools to examine
the functions and performance in an arid, high-elevation climate.
At the University of Utah researchers have been monitoring two green roofs since
fall 2013 (Fig. 2.9). The Natural History Museum green roof covers 1,115 m2, and
the Marriott Library roof 632 m2. One of the major purposes of this study is to
monitor ET, so researchers set up four weighing lysimeter systems. ET data col-
lected by the lysimeters will be compared to ET measurements made by a Camp-
bell eddy covariance system. Eddy covariance is one of the most widely accepted
micro-meteorological methods to directly measure fluxes like water vapor and
based on the covariance between wind speed and humidity (measured separately
but simultaneously at high frequency). Eric Pardyjak, Department of Mechanical
Engineering, uses the eddy covariance tower for this project. Supported by the Ur-
ban Transitional Arid-region Hydro-sustainability program, Dr. Steven Burian and
PhD student, Youcan Feng, use a weighing lysimeter and expect accurate green
roof ET time series from a point scale, not easily achieved by most other means.
Yet, “…accounting for Utah winds that generate errors for the scale read-
ingsrequires a lot of time to calibrate sensors and validate results (Feng, 2014,
pers. Comm.).” Burian and Feng recognize instrumentation limitations e.g. the
lack of good tools to measure the outflow time series from the lysimeter unit; a
tipping bucket would miss measurements and it is hard to apply depth sensors on a
green roof to provide reliable outflow data. The same is true for irrigation moni-
toring. While Salt Lake City’s arid climate requires irrigation, researchershave
not found good tools to measure the flows from irrigation tubes or sprinklers.”
(Youcan Feng, 2014, pers. comm.).
Fig. 2.9 Green roof monitoring setup on two roofs in Salt Lake City, Utah: a) Natural History
Museum of Utah; b) and c) University of Utah J. Willard Marriott Library (Lee R. Skabelund)
20
Water quality samples from green roofs can be collected by grab sampling or with
automated samplers. Care needs to be taken to sample different regions of the run-
off hydrograph intentionally. Subsequent laboratory analyses can be performed
for determining the concentration or presence of various constituents including nu-
trients and metals. Basic water chemistry parameters can also be monitored either
in the laboratory or through data collection in the field. To quantify parameters
such as conductivity, temperature, dissolved oxygen, pH and turbidity, a variety of
probes, sondes, or other measuring devices can be utilized to collect discreet or
time-series data sets. Some studies have evaluated the promising use of turbidity
as a proxy for TSS concentration in green roof outflow (Al-Yaseri, 2013).
Reviews of water quality studies from full-scale and laboratory green roofs includ-
ing factors that impact green roof performance are presented in the literature, e.g.
Berndtsson (2010) and various other studies have been reported (see for example
Berghage et al. 2007; Teemusk & Mander 2007; Dunnett et al 2008; Simmons et
al. 2008; Berghage et al. 2009; Bliss et al. 2009; Van Seters et al. 2009; Gregoire
and Clausen 2011; Schroll et al. 2011; Morgan et al. 2012; Toland et al. 2012;
Alsup 2013; Clark 2013; Gnecco 2013; Seidl et al. 2013; Wang et al 2013 and Za-
pater-Pereyra 2013) (Chap. 5).
2.5 Monitoring Energy Flows and Temperatures
In this section, we address green roof surface energy balance, temperature dynam-
ics (surface and subsurface), and thermal fluxes to and from buildings with green
roof systems. This section discusses approaches to monitoring e.g. surface tem-
peratures on a green roof system in comparison to conventional roof surfaces and
also addresses ways to monitor or otherwise understand the energy flows associat-
ed with a specific green roof ecosystem, as well as monitoring the relationships
between green roof energy flows and specific types of green roof plant systems.
2.5.1 Surface Energy Balance
The surface energy balance of a green roof generally differs from that of conven-
tional roofs. In comparison, green roofs can provide energy benefits to individual
buildings (as further discussed in Section 2.5.3) and impact ambient conditions.
Energy benefits can be associated with green roofs by passive cooling through
evaporative processes and latent heat fluxes as well as reflection of solar radiation
(characterized by albedo) and reduction of sensible heat fluxes. With widespread
adoption of green roofs, mitigation of the urban heat island (UHI) effect can be
21
achieved reducing, “peak energy demand, air conditioning costs, air pollution and
heat-related illness and mortality” (PlaNYC 2008).
Surface energy balance is represented as: 𝑅!=!𝜆𝐸 +𝐻+𝐺 where Rn is net radi-
ation, H is the heat flux to the air also known as sensible heat flux, λ is the latent
heat of vaporization, E is the rate of vaporization (evapotranspiration), and G is
the heat flux to the soil also known as soil heat flux (Hanks, 1992). Monitoring
surface energy balance components (see Fig. 2.10) can be achieved through the
use of various sensor technologies and estimation techniques.
Fig. 2.10 A green roof energy budget consists of various measurable components, DiGio-
vanni (unpublished) adapted from Gaffin et al (2011)
Net radiation includes incoming and outgoing long-wave and short-wave radia-
tion. Incoming solar radiation (often referred to as short-wave radiation) and out-
going (reflected) solar radiation are measured using pyranometers. Incoming and
outgoing long-wave and/or infrared radiation are monitored using pyrgeometers.
Monitoring surface and air temperatures as well as wind speed can estimate sensi-
ble heat flux. Surface temperature can be monitored using direct contact thermo-
couple sensors and infrared radiometers. Air temperature can be measured using
probes coupled with an appropriate solar radiation shield. Latent heat fluxes asso-
ciated with evapotranspiration (Sect. 2.4.4, Chap. 3) can be monitored by various
techniques or by backing out the term in the energy balance if all other parameters
are known. Studies evaluating energy balance and latent heat fluxes related to
green roofs include: Jim & He (2010), Susca et al. (2011), Coutts et al. (2013),
Kim & Park (2013), Nagengast (2013), Peng (2013) and Song et al. (2013).
22
Case%Study:%Monitoring%Green%Roof%Energy%Flows/Fluxes%&%Energy%
Balance%(Columbia%University%Research%8%New%York%City)%%
Columbia University and City College of New York researchers have deployed a
robust, high quality sensor network on a diverse array of green roofs in NYC. The
sensors enable quantification of various components of the energy balance. The five
monitored green roofs, presented in Gaffin et al. (2009), vary in their layout, materi-
als and structure. Monitoring stations used identical sensor selections paired with a
monitored “control” roof. Back-to-back pyranometers (Kipp and Zonen CMP3)
measure incoming and outgoing short-wave radiation and determine surface albedo.
Incoming and outgoing long-wave radiation are quantified using surface temperature
and relative humidity inputs, with surface infrared radiometers (Apogee Instruments
SI-111, previously IRR-P) and temperature/relative humidity probes (Campbell Sci-
entific [CS] CS215). The surface infrared radiometers were noted as, “particularly
useful for monitoring green roof leaf temperatures which have a complex geometry”
(Gaffin et al 2009 p 2655). Gaffin et al (2009, 2654) also noted that, “sensors are in-
creasingly becoming available to measure all four SW and LW fluxes.” Net radiom-
eters like Kipp and Zonen CNR2 and Hukesflux NR01 include back-to-back pyra-
nometers and pyrgeometers. The sensible heat calculation requires air temperature
and wind speed. These are monitored with temperature/relative humidity probes
(CS215) and wind sensors (RM Young 05013). Latent heat fluxes due to evapotran-
spiration (ET) were quantified using inputs of wind speed, temperature and relative
humidity. Drexel University also developed a custom-designed and built weighing
lysimeter system providing ongoing direct measurement of ET. The substrate heat
flux is quantified using monitoring data from heat flux plates (Hukesflux HFP01,
though a self-calibrating model may be preferred e.g. HFP01SC), substrate tempera-
ture sensors (CS107) and volumetric water content sensors (CS616). Data from sen-
sors is recorded on data-loggers (CS-CR1000). Fig 2.11 shows one of the NYC
green roof monitoring setups.
Fig. 2.11 Meteorological
and hydrological moni-
toring at the Ethical Cul-
ture Fieldston School
green roof (Kimberly
DiGiovanni).
23
2.5.2 Temperature Dynamics
Temperatures of green roofs are important to for understanding benefits of green
roofs related to cooling, reduction of the urban heat island (UHI) effect and build-
ing energy savings. Temperature dynamics and changes can be measured using
surface and sub-surface temperature sensors.
It is helpful to monitor different kinds of conventional or control roofs as each
“conventional” roof type will likely perform differently in some respectswith
the same being true for different green roof substrate types and depths including in
different climatic conditions. The following is an example.
On the 6,875 square meter (m2) Walmart Green Roof in Chicago, Illinois the ener-
gy impact of an extensive (approximately 7.5 cm) green roof was analyzed and
compared with an adjacent white roof based on 2006-2009 monitoring. The fol-
lowing parameters were measured at points distributed across the two rooftop
types: 1) surface temperatures; 2) temperatures under the roof membrane; 3) tem-
peratures below the roof deck; and 4) temperatures in the substrate profile for the
green roof only. Heat flux (Q, in watts per square meter)a measure of energy
flowing in or out of the store through the roofwas also monitored. HVAC air in-
take temperatures were measured from July 2009 to July 2010 (Walmart et al.
2103). “To analyze energy impact of the green roof, the heat flux data collected
from the roof was integrated into a simplified building model [then] into the full
store energy model”helping researcher “interpret field data [and] allot heat flux
differences properly.” The model translated temperature difference into “energy
use difference” by the rooftop and air handling units by “modeling air temperature
difference on the green side as precooling or preheating…” (Walmart et al. 2103,
p. 21). Models were run for the Chicago store, then data were extrapolated to a
model in Houston, Texas to gauge likely green roof energy impacts. Average an-
nual conditions were studied as well as peak heating and cooling periods to deter-
mine the green roof’s effect on store energy use (Walmart et al. 2103).
2.5.2.1 Surface temperatures: Green roofs and control roofs
Surface temperature can be monitored using direct contact thermocouple sensors
and infrared radiometers. Researchers in diverse geographic regions have used
these techniques for measuring green roof surface temperature and in comparison
to conventional roof surfaces. Sidwell et al. (2008) evaluated a southern Illinois
green roof and a black roofing membrane (ethylene propylene diene terpolymer or
EPDM) control using temperature sensors; the green roof fluctuated between
23.6°C and 29.8°C; the EPDM control roof 19.1°C and 46.3°C. Monitoring by
Dvorak and Volder (2013) in central Texas found that an un-irrigated modular Se-
dum green roof was 18.0°C cooler at the surface and 27.5°C cooler below the
24
modules in comparison to a control roof during summer months. DeNardo et al.
(2005) in Pennsylvania found an 8.9 cm green roof substrate surface to be 6°C
warmer in winter months and 19°C cooler in summer months (over a control).
Monitoring by Wong et al. (2003) in Singapore revealed intensive rooftop garden
temperatures 30°C cooler than a control roof. In Japan, Onmura et al. (2001)
found temperatures to be 30 to 60°C cooler on a green roof compared to a nearby
control roof.
2.5.2.2 Sub-surface temperatures: Substrate
It is important to understand how warm or cold it gets beneath the substrate sur-
face since green roofs can insulate buildings. In combination with substrate mois-
ture levels substrate temperatures also strongly influence evapotranspiration rates
and vegetative health.
Temperature profiles can be achieved for different roof surface types using ther-
mocouples. Pearlmutter and Rosenfeld (2008) applied thermocouples to different
locations on small building “cell” replicates to compare various methods of roof-
cooling including from mesh shading, soil, and gravel. A heat flux plate was also
placed under the soil and simultaneous measurements made of global radiation
(using a Kipp and Zonnen pyranometer), wind speed (via a LSI constant tempera-
ture hot-wire anemometer) and ambient air temperature (using a Campbell
21xdatalogger). They found that though roof shading material provided more
overall daily cooling, gravel covered roofs optimize daily cooling potential which
is important for the desert climate in which their work was conducted.
Dvorak and Volder (2013) placed thermistors at multiple depths in green roof
modules to compare the effects of irrigation on cooling. Silicone was used to seal
the periphery of temperature probe wires. A Campbell CR1000 data logger tem-
peratures more frequently, but data was condensed to mean hourly intervals. Per
their abstract: “Ambient air temperatures were collected on the rooftop with non-
forced ventilation shielded air temperature instrumentation (Humidity and Tem-
perature Probe HMP155 and 41005-5 radiation shield Vaisala, Helsinki, Fin-
land).” Substantial temperature reduction in unirrigated modules was noted (com-
pared to standard roof surfaces).
The role of plants and water availability in green roof cooling has been the focus
of several studies. Most of these use similar thermistor technology to measure
temperature as previously mentioned, but from each study, different monitoring
advice can be gleaned. Jim (2012) noted for their study on plant effects in the trop-
ics, for example, their concern about the effect of advection on adjacent plots and
recommend larger study plots to address this as well as placing sensors in the
middle of these plots. This study found that grass plots cooled more effectively
25
than groundcover or shrub. Butler and Orians (2011) used a Maxim ibutton high
capacity temperature logger DS1922L which revealed that temperature regulation
might be one mechanism via which Sedum could act as nurse plants.
2.5.3 Building Thermal Fluxes (Insulating Properties)
Green roofs impact the heat gain and loss to and from buildings and influence the
heating and cooling loads. Green roof substrate provides insulation and vegetation
reflects solar radiation more effectively than most conventional roof surfaces pre-
venting solar heat gain and increasing the thermal resistance of the building (Eu-
morfopoulou & Aravantinos 1998).
Monitoring of thermal fluxes in and out of buildings can be achieved using a vari-
ety of sensors including thermocouples, thermistors, temperature probes (inside
and ambient air temperature), and heat flux plates. The heating and cooling im-
pacts of green roofs can be evaluated through energy usage data, typically metered
through power-supply companies. Studies evaluating the thermal gain and thermal
resistance of buildings with green roofs include: Pierre (2010), Fioretti (2010),
Becker & Wang (2011), Zhao & Srebric (2012), Chan (2013), Darkwa (2013),
Moody & Sailor (2013) and Olivieri (2013).
2.6 Monitoring Wind Speed and Direction on Green Roofs
Wind speed and dynamics can influence the stability of the entire green roof sys-
tem. Wind also has a major influence on the movement and drying of substrates
and the viability of green roof vegetation. Thus, green roof monitoring should
document wind speed and directionin tandem with monitoring other relevant
micro-meteorological, hydrologic, and substrate variables.
Monitoring the impact of wind dynamics on vegetation and/or green roof system
movement, dislodgement, breakage, and overall green roof stability can be ac-
complished by using a wind tunnel for modules, or cameras and observations for
an entire green roof system. Wind scour or loss of substrate can be measured us-
ing devices placed in the substrate and observed over time and/or using one or
more high-resolution video cameras to record movement of particles during windy
periods. Simple or highly sophisticated wind scour modeling and analysis are pos-
sible.
26
2.6.1 Stability of Substrate and the Entire Green Roof System
Zhang et al (2007) indicate that green roof “soil erosion” induced by winds de-
creases with higher levels of plant cover. Roots bind plant masses to the substrate,
thus providing a windbreak from erosive forces. To observe such phenomena,
University of Central Florida researchers implemented two, full-scale green roofs
to continuously monitor wind effects, using “a grid of very low differential pres-
sure transducers and a high speed anemometer for wind speed and direction.” A
geosynthetic erosion control blanket was used on one roof, significantly reducing
substrate loss (Wanielista et al. 2011 p v). The researchers note that field data
from several monitoring stations with high wind velocities may better define de-
sign parameters for all green roof-building options.
Cao et al. (2013) explored wind load characteristics for green roof modules. A se-
ries of wind tunnel experiments were carried out on a scaled-down module in-
stalled in different positions on two types of building models. They investigated
peak force and moment coefficients of the model rooftop and the effects of para-
pets and other design parameters Retzlaff et al. (2010) employed a subsonic, recir-
culating wind tunnel to evaluate wind uplift and wind scour of partially and non-
vegetated modular green roof systems.
2.6.2 Built Context Influences Wind Patterns and Dynamics
Wind and wind variability are strongly influenced by building mass, height and
parapets thus influencing green roof systems. For example, visual observations of
wind and snowfall on two Kansas State University green roofs indicate patterns of
wind and precipitation respond to building mass, location, and height (Skabelund
2014, research notes). Use of an anemometer to monitor wind speed and direction
provides useful information on the dynamics associated with wind strength, direc-
tion, and patterns related to the urban context.
2.7 Monitoring Substrate Attributes
Without understanding the specific attributes of green roof substrate characteris-
tics it is unlikely that we can create green roof ecosystems that are resilient and al-
so minimize resource demands, especially supplemental irrigation and nutrients.
In this section, we address ways to monitor substrate pH, nutrients, organic matter,
metals, and other constituents that are seen as vital to plant systems (but potential-
ly detrimental to downstream aquatic systems). We also examine ways to effec-
27
tively assess changes in soil attributes over time, noting that once a green roof is
installed, substrate properties can be sampled to inform maintenance decisions like
fertilization frequency. Monitoring substrate attributes can inform balanced
maintenance and management decisions (e.g. fertilization to secure system surviv-
al and health targeted to reduce nutrient and metal loads in green roof outflow).
2.7.1 Substrate pH, Nutrients, Minerals, and Organic Matter
Substrate chemical parameters and organic matter content are expected to be high-
ly variable across time and space. Nutrient and pH studies are time and resource
intensive, so it is important that they incorporate additional roof metadata so that
findings can be generalized. The green roof community also needs to agree on ap-
propriate reporting units. For example, organic matter content is sometime report-
ed as volume per unit substrate and other times reported as mass per unit substrate.
Zheng and Clark (2013) evaluated five different Sedum species under variable
substrate pH conditions. By identifying species-specific characteristics and opti-
mizing substrate pH, Zheng and Clark suggested that Sedum growth can be opti-
mized.
Panayiotis et al. (2003) studied four substrates for their capacity to sustain Lanta-
na camara L. Physical and chemical evaluation included “weight determination at
saturation and at field capacity, bulk density determination, water retention, air
filled porosity at 40 cm, pH and electrical conductivity.” Plant growth evaluated
“shoot length, shoot number, main shoot diameter and the number of buds and
flowers” (Panayiotis et al. 2003, abstract).
2.7.2 Changes in Substrates through Time
Substrate changes related to pH, organic matter and specified minerals and met-
als, and substrate dynamics such as properties, characteristics, quality, and/or nu-
trient levels can be measured over time (Chap. 5).
Green roof substrates can be influenced by climate and vegetation changes over
time. For example, preliminary evidence (Griffin 2013) suggests substrate particle
size distribution can be influenced by freeze-thaw cycles such that the smaller size
fraction is increased proportionally compared to the others. In order to determine
this distribution, substrate is poured through sieves of various sizes. Two related
measurements that could be useful to monitor green roofs over time include sub-
strate depth and bulk density, which is the weight of the substrate divided by the
volume. Studies of German green roofs from the 1980s also confirm that green
28
roof substrates tend to get thinner (shallower) as they age, but also increase in or-
ganic matter content (Thuring and Dunnett 2014).
2.8 Monitoring Gas Exchange and Carbon Sequestration
Carbon sequestration on a green roof can be achieved as a function of vegetation
transpiration and related photosynthetic processes. By sequestering carbon, green
roofs help to mitigate climate change.
Carbon fluxes via gas exchange, and carbon sequestration by gains and losses of
CO2 through photosynthesis and respiration can be measured by direct measure-
ments of CO2 atmospheric exchange, or by measuring changes in C stocks over
time (i.e., collecting, drying and weighing substrate and root samples). Modeling
and analysis of atmospheric carbon fluxes, net ecosystem productivity, and carbon
sequestration through time is possible and sophisticated equipment exists to do so
(.Chap. 5).
Several efforts are being made to construct carbon budgets for green roofs. Foun-
dational work conducted in Michigan documented changes in plant biomass and
associated carbon content over time in order to assess green roof carbon sequestra-
tion (see Michigan State University case study for details regarding specific meth-
odology). It is important to note that this approach ignored carbon losses from the
system via respiration and leaching.
Researchers in Vancouver are updating calculations to incorporate respiration us-
ing experimental chambers “Li-Cor LI-8100” (Gaumont-Gauy and Halsall 2013).
In their 2012 study, five chambers, 314 cm2 in area, measured CO2 fluxes at the
roof-atmosphere interface for five different Sedum species. One additional cham-
ber measured CO2 flux from an unplanted surface to assess respiration. The net
ecosystem productivity (NEP) was determined as the balance between gross CO2
assimilation through plant photosynthesis and CO2 release through plant and de-
composer respiration. Gaumont-Gauy and Halsall (2013) found that net C assimi-
lation integrated across plant types was 440g m-2 yr-1. A range of uptakes were ob-
served for different species whereby endemic species native to the region
exhibited higher net carbon fixation compared to others; these findings are sup-
ported by Starry et al 2014 (in review) who also noted a range in uptake for differ-
ent Sedum species. Gaumont-Gauy and Halsall (2013) further note that their study
does not include carbon lost from the system via leaching; future work may in-
volve a more integrated study of all the different green roof carbon pathways.
29
2.9 Synthesis of Green Roof Monitoring: Approaches, Costs,
Challenges and Lessons Learned
2.9.1 Green Roof Monitoring Approaches
Green roof monitoring can range from simplistic to complex, and data extensive to
data intensive. Common approaches and tools include observation, experimenta-
tion, computer modeling, and in situ sensors. For measurements of water quality
and substrate attributes, samples require additional lab support. The importance of
integrated green roof monitoring is highlighted in the following section.
2.9.2 Integrated Green Roof Monitoring
Integrated green roof monitoring brings together observation, experimentation and
data collection in a manner that enables researchers to understand complex interre-
lationships over an extended period of time. Early work by Carter and Rasmussen
(2006) in Athens, Georgia (USA), Glass and Johnson (2009) in Washington, DC,
Berghage, et al. (2009) in Pennsylvania, and other researchers set the stage for in-
depth and integrated green roof monitoring now occurring in many locations in
North America. Following are three brief examples of integrated green roof moni-
toring in the U.S. and Canada:
University of Pittsburgh: One example of an integrated monitoring approach is
the studies by University of Pittsburgh. Researchers there evaluated green roofs in
comparison to conventional roof tops focusing on various interrelated areas in-
cluding stormwater management, water quality and thermal benefits. In evaluating
these factors, sensor systems including flumes, weir boxes, ultrasonic sensors, soil
moisture sensors, rain gauges, thermocouples, temperature probes, net radiome-
ters, laboratory water quality analyses, and data loggers connected by modem and
electronic networks were utilized (Neufeld et al. 2009).
EPA Region 8 Green Roof: The first large-scale extensive green roof in Colora-
do was created atop the ninth floor of the EPA building in Denver. Covered with
Sedum species, cacti, and grasses, this 1,858 m2 roof is near a gravel ballast con-
trol roof. Both roofs have: 1) weather stations to measure temperature differences;
2) instruments to monitor stormwater runoff rates and quantities; and 3) water col-
lectors for water quality analysis. In 2008-2009, Klett et al. (2012) evaluated green
roof vegetation (biomass) in relation to different substrates, zeolite amendments,
30
Case%Study:%Monitoring%Green%Roof%Carbon%Sequestration%–%(Mich8
igan%State%University%–Lansing,%Michigan)%
Michigan State University researchers were among the first to consider the green
roof carbon balance. Getter et al. (2009) monitored 13 green roofs (nine in Michi-
gan and four in Maryland). For twelve roofs, plant material and substrate were
harvested seven times across two growing seasons. Roofs ranged from one to six
years in age and from 2.5 to 12.7 cm in substrate depth. Replicate samples of
aboveground biomass were collected, dried, and ground. Carbon accumulation
was determined by multiplying dry matter weight by total C concentration. As part
of this study an additional roof was planted in different Sedum species, and root
and substrate samples were included in the Carbon analyses over two growing pe-
riods every two months. This carbon data was then used to support ecological ob-
servations about the different Sedum and compare carbon sequestration with car-
bon flows to and from the green roof.
Whittinghill et al. (2014) took a similar approach in their study of different land-
scape areas (including green roofs). Carbon content analysis was performed on
above-ground biomass, below-ground biomass (roots), and soil and substrate col-
lected at the end of the 2010 and 2011 growing seasons (ibid).
and irrigation regimes. They used digital image analysis (employing SigmaScan
Pro 5.0 image analysis software) and manually collected two-dimensional data. To
assess water-holding capacity of substrates, they collected volumetric moisture
content data using a Delta-T ThetaProbe ML2X. Their analytical methods includ-
ed multivariate analysis.
Vancouver Island University: VIU researchers used a sophisticated green roof
monitoring design strategy for evaluating gas/vapor exchanges, vegetation, mete-
orological conditions, water and energy fluxes, and water quality from green
roofs. Their integrated monitoring design strategy, initiated in January 2012, in-
cludes automated and portable CO2 and H2O exchange chambers, digital cameras,
weather stations (precipitation, radiation, temperature, relative humidity), water
level sensors, soil heat flux sensors, and sensors to monitor dissolved organic car-
bon in green roof runoff (CDOM/FDOM sensor). Data logging/acquisition sys-
tems (Campbell Scientific) and computational software (Matlab) are integral to
their green roof monitoring (Gaumont-Guay 2014, pers. comm.).
Portland State University: Researchers at the Green Building Research Lab are
pursuing several monitoring objectives. Projects include “very simple monitoring
setups (a weather station, soil moisture, soil temperature)” and also more complex
systems “involving those same sensors as well as surface heat flux sensors, net ra-
diometers, arrays of air temperature rakes, and HVAC monitoring” (Sailor 2014,
pers. comm.). One of PSU’s projects addressed reciprocal effects of solar panels
31
and green roofs and for other integrated monitoring, data loggers interfaced with
indoor environmental quality sensors and outdoor weather sensors to monitor air
temperature, CO2, occupancy, relative humidity, and equipment run time. PSU re-
searchers currently monitor stormwater and heat loads associated with a 3,716 m2
green roof (Williams 2013).
University of Toronto: Toronto’s Green Roof Innovation Testing Laboratory
(GRIT Lab) uses “real time data monitoring and ongoing field observation to
study the metrics associated with [green roof] systems” (ASLA 2013). The 372 m2
GRIT Lab section of roof is dedicated to conducting experimental researchwith
33 (1.22 m x 2.44 m) modules. Each module is instrumented with eight sensors
one soil moisture sensor, a rain gauge to measure runoff and flow rates from each
module, and five thermal sensors along a vertical axis to generate a thermal pro-
file. One infrared radiometer records the average surface temperature of a 0.914 m
diameter circle. At least 12 researchers are involved in this green roof research
projectintended to be holistic and integrated by evaluating interrelated process-
es, including meteorological conditions, heat/energy flows, gas exchanges, water
quantity/quality, soils, vegetation, and fauna (ASLA 2013). MacIvor (2014 pers.
comm.), noted that the GRIT Lab “green roof has been online since late 2010 but
has only been fully instrumented (full array of stormwater and thermal sensors,
data loggers and dedicated computer with macros scripts to recall and subset data
from all sensors), calibrated, and fully automated since June 2013.” Early monitor-
ing efforts at this site focused on irrigation and plant success (MacIvor et al.
2013), but ongoing research is targeting a number of different questions, especial-
ly those related to pollination.
2.9.3 Green Roof Monitoring Costs and Funding Sources
The cost of monitoring green roofs ranges widely. Basic observational monitoring
can be conducted for little cost while complex monitoring operations require hun-
dreds of thousands in equipment and personnel. Specific equipment costs are
available from the manufacturers, but also range within the same monitoring pa-
rameters depending on the sensor specification, range and accuracy.
From Michigan State University (MSU), Bradley Rowe (2014, pers. comm.) ex-
pressed that Campbell Scientific data loggers, etc. were the most cost intensive of
the equipment purchased for their green roof monitoring work. The Plant and Soil
Sciences green roof at MSU was instrumented with heat flux sensors, moisture
sensors, thermocouples, and a weather station, costing approximately $10,000 (see
Getter et al. 2011). The monitoring system was pieced together using either exist-
ing or purchased equipment. Most MSU funding for green roof monitoring came
from green roof suppliers, the USEPA, and internal university grantswith the
32
largest grants from Ford Motor Co. (and numerous smaller grants from companies
such as LiveRoof and XeroFlor America). As is the case with other monitoring
projects, many donated green roof materials were contributed.
At the University of Maryland, Lea-Cox (2014, pers. comm.) notes that “research
instrumentation costs are considerably higher than what would ultimately be in-
stalled on commercial green roofs.” Monitoring costs inevitably depend on the
size of the green roof, the complexity of the research, and the equipment already
available.
Retzlaff (2014, pers. comm.), describes Southern Illinois Edwardsville University
(SIEU), five green roofs monitored on five different campus buildings the larg-
est 1,486 m2 and the smallest 28 m2. SIEU also has four (4) green roof projects for
stormwater runoff and a green wall test area (with 18 modular green walls). SIEU
utilizes Hobo data loggers and “simple soil thermal devices” to monitor the tem-
perature of the green roof systems. The loggers cost approximately $300 each and
the thermal probes about $65. For measuring stormwater runoff, SIEU uses inex-
pensive five-gallon red plastic gas cans that they weigh to track runoff from each
storm event. SIEU researchers obtain meteorological data from a local reporting
station. They used a wind tunnel for wind researchthe original price was more
than $275,000 plus, but it was purchased as part of the new Engineering Building
(and used for many other projects). Between 2004 and 2014 SIEU received ap-
proximately $100,000 in external funds for their green roof and green wall re-
search projects. The largest grants were an EPA P3 award and direct funding from
the National Roofing Contractors Association. Retzlaff noted: “It would be great if
we could use expensive instrumentation for our research projects. Unfortunately
most of the funding we receive only covers the green roof or green wall materials
to conduct the testing.”
In short, complex monitoring systems are costly. Most monitoring systems de-
scribed in this chapter cost between $5,000 and $20,000 (including in-kind loans
or donations). They may also require the expertise of specialists or consultants to
maintain them and troubleshoot any site-specific challenges. More research is
necessary to quantify the benefits of these monitoring systems relative to costs.
2.9.4 Green Roof Monitoring Challenges and Lessons Learned
Monitoring green roofs can present a variety of challenges including collection of
representative data sets, instrumentation and maintenance of monitoring systems,
as well as management and interpretation of collected data. Researchers conduct-
ing green roof monitoring have experienced many challenges and offer lessons
learned from these experiences in this section.
33
Researchers recognize that collecting data that is representative of the overall
green roof system is a goal of green roof monitoring that can be a challenge to
meet. For example, modules and smaller platforms may be quite constraining in
regards to the growth demands or requirements of some grasses and herbaceous
vegetation. This is fine if the goal is to test selected species growth and viability in
these constrained systems and compare them with integrated (monolithic) green
roof systems but does not necessarily reflect the functional characteristics of the
larger, integrated systems. Furthermore, adjacent walls or structures can have a
significant influence on wind patterns and thus rainfall eventsconcentrating
more precipitation (rainfall or snowfall) on one portion of a green roof and reduc-
ing or eliminating rainfall/snowfall on another portion of the same roof.
Beyond creating monitoring systems to allow for representative data collection,
there are limitations with monitoring instrumentation. For example monitoring of
hydrologic inputs and outputs from green roofs can present potential challenges
and limitations including the following: 1) Flumes and gauges may not capture
low flows and are susceptible to debris blockages 2) Flow meters can also be
blocked and disabled by particles and magnetic flow meters require full pipe flow
for operation 3) Tipping buckets cannot capture precise precipitation rates that are
very small or very large/rapid and are generally only good for small areas as they
can be overwhelmed by large amounts of flow. 4) Cisterns, rain-barrels, and buck-
ets collecting runoff may overtop in larger storm events making accurate runoff
measurements impossible. (Rowe Mar. 2014, pers. comm.).
Other researchers have also experienced equipment related limitations to green
roof monitoring. For example, when asked about the pros and cons related to mon-
itoring hydrology from 12 mock-up green roof panels and three mock-up control
roof panels (1.524 x 1.524 meter) in Fayetteville, Arkansas, Mark Boyer, Univer-
sity of Arkansas, stated: “For me it was the tipping bucket capacity. I really want-
ed to be able to compare the lag time of runoff off of a green roof compared to a
conventional membrane roof and I thought the tipping buckets could do that for
me. We had attempted using weirs on the first green roof that installed, but there
were problems associated with that, so I was hopeful that the tipping buckets
would solve the problem. We tried using tipping buckets to measure the quantity
and timing of runoff but their capacity was exceeded and so we had to resort to
capturing all of the runoff and omitting the timing effect” (Apr. 2014, pers.
comm.).
Data collection and data management are another challenge present in green roof
monitoring. Collecting soil moisture and other data with a data logger is helpful,
but setting everything up and getting all equipment working the way is supposed
to work can be time consuming and very challenging (Rowe Mar. 2014, pers.
comm.). Downloading recorded data (especially for data recorded every 5-15
minutes) can also be quite time consuming. Some devices automatically save data
34
with file names indicating the date and time data was downloaded. Correlation
with daylight savings times may be needed for some devices. Some data may need
to be collected using a USB or other direct cable connection. Ethernet or wireless
connections may be able to speed this process up and costs may be minimal if
wireless or Ethernet connections already exist. Otherwise, costs will increase.
Linking data collection devices to the Internet can be very helpful and save time if
done well (enabling ready access to multiple users and allowing for sharing of re-
sults from anywhere that a researcher can access the Internet). Quick and ready as-
sessment of data is possible via networked monitoring and analytical equipment,
but requires well-trained and funded personnel.
Interpretation and analysis of monitoring data from green roofs presents further
challenges. Careful analysis and interpretation of monitoring results is required
prevent conclusions that are incomplete, problematic and misleading. For exam-
ple, Berghage, et al. (2009) note the importance of relating concentrations of green
roof outflow to total volume. “Although the runoff concentrations (from the green
roofs) were typically higher, the loading was not always higher” (p. 4-16).
Furthermore, Berghage et al (2009) supported that interpretation from green roof
studies must recognize that, as with all ecosystems, green roofs are dynamic sys-
tems with living properties that impact the system outputs. For example, Berghage
et al. (2009) found that outflow unplanted substrate sections was considerably
higher in both concentration of tested water quality parameters and in total volume
of outflow than planted green roofs, suggesting “that newly planted roofs are like-
ly to have much higher runoff loading rates than established roofs” (p. 4-17). The
study also demonstrated seasonal variation in runoff for some (but not all) runoff
constituents monitored, and this may be attributed to seasonal fluctuation in plant
uptake.
Researchers also recognize that monitoring needs are tied to regional and site-
specific conditions related to e.g. location and the design and size of the installa-
tion. Furthermore, establishing and maintaining a green roof monitoring network
requires sustained funding and appropriately trained personnel for the mainte-
nance and upkeep of monitoring equipment and acquisition of quality data sets.
2.10 Future Directions for Green Roof Monitoring
The future of green roof monitoring holds many opportunities, particularly at the
intersections relating hydrologic processes, evapotranspiration, energy transfer,
vegetation, nutrient cycling and other services provided by green roofs. Integrated
studies considering holistic and multi-faceted approaches to evaluating green roofs
35
are increasingly needed. Such integrated studies reveal findings valuable for un-
derstanding various interrelated processes and concomitant benefits.
It is important to note that monitoring (including experimentation) is sometimes
an afterthought in regards to green roof research. To improve monitoring out-
comes, researchers need to collaborate with practitioners as part of the green roof
planning and design process to create “designed experiments” (Felson et al. 2014)
to address various research needs.
One step that would help to unite green roof researchers collecting monitoring da-
ta is a platform that would facilitate the comparison of national green roof da-
tasets. Some national databases already exist, but these lack an option to search for
monitoring data. Table 2.1 below illustrates how such a database might be set up
to include information relevant to researchers.
Various topics in need of research attention exist beyond those mentioned previ-
ously within the context of integrated green roof research. Given continued reli-
ance on succulents, comparisons between Sedum-dominated, systems exclusively
supporting native grasses, forbs and other indigenous species, and mixed vegeta-
tive systems are needed in relation to long-term stormwater runoff and water qual-
ity trends. Furthermore, studies are needed to address the impacts of different
types of green roofs on air quality, a topic of research that has received little focus
as of yet.
Table 2.1 Green Roof Database Categories
Roof
location
Size, slope,
aspect, sub-
strate depth,
etc.
Hydrology-
related data
Energy-
related data
Microclimatic
data
Biodiversity
Studies
Roof
sample
10,000 sq. ft.
929 m2
Yes
Yes
Yes
Logbook
Observations
The development of monitoring networks incorporating automated sensor/system
technologies real-time, remote sensing networks and data management systems
with low-cost sensor technology will aid in advancing green roof monitoring initi-
atives. Mooney-Bullock, et al. (2012) provide an example of a low-cost sensor
network using new technology to monitor green infrastructure including green
roofs, revealing how real-time monitoring can be implemented in an affordable
manner.
Further, the assessment of neighborhood scale impacts of green roof adoption,
which has received limited attention, would expand the scope of green roofs bene-
fits and dovetail into research related to city-wide planning and green infrastruc-
36
ture planning and networks (e.g. Green City, Clean Waters and NYC Green Infra-
structure Plan).
Overall, the future of green roof monitoring presents many intriguing and practical
research opportunities. What is of interest about many green roof monitoring pro-
jects is the time and expertise required to design, install, test, calibrate, and vali-
date data generated by the instruments and equipment which can lead to the ques-
tion: Is simpler better? That depends on the research questions being asked and the
particular green roof types and contexts.
Works Cited
Al-Yaseri I Morgan S Retzlaff W (2013) Using turbidity to determine total
suspended solids in stormwater runoff from green roofs. J of Env Engin-ASCE
139(6):822-828
Alsup S Ebbs S Battaglia L Retzlaff W (2013) Green roof systems as sources
or sinks influencing heavy metal concentrations in runoff. J of Env Engineer-
ing-ASCE 139(4):502-508
Arvidson AR (2012) Greening the Landscape. WW Norton, NY
American Society of Landscape Architects (ASLA) (2013) Award of Excel-
lence: Green Roof Innovation Testing (GRIT) Laboratory. ASLA Professional
Awards project narrative, http://www.asla.org/2013awards/394.html
Beattie D Berghage R (2004) Green roof media characteristics: the basics. In:
Proc. of 2nd North American Green Roof Conference: Greening Rooftops for
Sustainable Communities, Portland, Oregon, 2-4 June 2004
Becker D Wang D (2011) Green roof heat transfer and thermal performance
analysis. Unpublished Carnegie Mellon University Report, Pittsburgh, PA
Berghage R Jarrett A Beattie D et al (2007) Quantifying evaporation and tran-
spirational water losses from green roofs and green roof media capacity for
neutralizing acid rain. Pennsylvania State University, State College, PA
Berghage R Beattie D Jarrett A Thuring C Razaei F (2009) Green roofs for
stormwater runoff control. T. P. S. University. USEPA, National Risk Man-
agement Research Laboratory, Water Supply and Water Resources Division.
EPA/600/R-09/026
Berndtsson J (2010) Green roof performance towards management of runoff
water quantity and quality: A review. Ecol Engineering 36(4):351-360
Blank L Vasl A Levy S Grant G Kadas G Dafni A Blaustein L (2013) Direc-
tions in green roof research: A bibliometric study. Build. & Env. 66:23-28.
37
Blanusa T Vaz Monteiro M Fantozzi F Vysini E Li Y Cameron R (2013) Al-
ternatives to Sedum on green roofs: Can broad leaf perennial plants offer better
‘cooling service’? Building & Environment 59:99-106.
Bliss DJ Neufeld RD Ries RJ (2009) Stormwater runoff mitigation using a
green Roof. Env Engineering Science 26(2):407-417.
Butler C Orians CM (2011) Sedum cools soil and can improve neighboring
plant performance during water deficit on a green roof. Ecol. Engin.
37(11):1796-1803
Byrne LB Grewal P (2008) Introduction to ecological landscaping: A holistic
description and framework to guide the study and management of urban land-
scape parcels. Cities and the Environment 1(2):article3
Cao J Tamura Y Yoshida A (2013) Wind tunnel investigation of wind loads on
rooftop model modules for green roofing systems. J of Wind Engineering and
Industrial Aerodynamics 118:20-34
Carson TB Marasco DE Culligan PJ McGillis WR (2013) Hydrological per-
formance of extensive green roofs in New York City: Observations and multi-
year modeling of three full-scale systems. Env. Res. Letters 8(2):24-36
Carter TL Rasmussen TC (2006) Hydrologic behavior of vegetated roofs. J
Am. Water Resources Association 42(5):1261-1274
Chan ALS Chow TT (2013) Evaluation of overall thermal transfer value (ottv)
for commercial buildings constructed with green roof. App. Ener. 107:10-24
Charlesworth SM Perales-Momparler S Lashford C Warwick F (2013) The
sustainable management of surface water at the building scale: preliminary re-
sults of case studies in the UK and Spain. Aqua, London (0003-
7214) 62(8):534-544
Clark MJ and Y Zheng (2013) Plant nutrition requirements for an installed se-
dum-vegetated green roof module system: Effects of fertilizer rate and type on
plant growth and leachate nutrient content. HortScience 48(9):1173-1180
Coutts A E Daly J Beringer and N Tapper (2013) Assessing practical measures
to reduce urban heat: Green and cool roofs. Buildi. & Env. 70:266-276
DeNardo JC Jarrett AR Manbeck HB Beattie DJ Berghage RD (2005) Storm-
water mitigation and surface temperature reduction by green roofs. Transac-
tions of the ASAE 48:1491-1496
Darkwa J Kokogiannakis G Suba G (2013) Effectiveness of an intensive green
roof in a sub-tropical region. Building Services Engineering Research & Tech-
nology 34(4):417-432
38
DeCuyper K Dinne K Van de Vel L (2005) Rainwater discharge from green
roofs. Plumbing Syst. Design Nov/Dec:1015
Dietz M (2007) Low impact development practices: A review of current re-
search and recommendations for future directions. Water Air & Soil Poll.
186(1-4):351-363
DiGiovanni K (2013) Evapotranspiration from urban green spaces in a north-
east United States’ city. Dissertation, Drexel University
DiGiovanni K Gaffin S Montalto F (2010) Green roof hydrology: results from
a small-scale lysimeter setup (Bronx, NY). Low Impact Development
2010:1328-1341. doi: 10.1061/41099(367)114
DiGiovanni K Gaffin S Montalto F Rosenzweig C (2011) The applicability of
classical predictive equations for the estimation of evapotranspiration from ur-
ban green spaces: Green roof results. World Environmental and Water Re-
sources Congress, Palm Springs, California. 22-26 May 2011.
DiGiovanni K Montalto F Gaffin S Rosenzweig C (2013) Applicability of
classical predictive equations for the estimation of evapotranspiration from ur-
ban green spaces: Green roof results. J of Hyd Engin 18(1):99-107. Figure 4
with permission from ASCE
Dunnett N Nagase A Booth R Grime J (2008) Influence of vegetation compo-
sition on runoff in two simulated green roof experiments. Urb Ecos 11:385-398
Dvorak B Volder A (2013) Rooftop temperature reduction from unirrigated
modular green roofs in south-central Texas. Urb Fores. & Urb Green 12(1):28-
35
Eumorfopoulou E Aravantinos D (1998) The contribution of a planted roof to
the thermal protection of buildings in Greece, Energy and Buildings 27(1):29-
36
Fassman E Simcock R (2006) Quantifying interception for New Zealand green
roofs. Ecole Plytechnique
Fassman-Beck E Voyde E Simcock R Hong YS (2013) Four living roofs in
three locations: Does configuration affect runoff mitigation? J of Hydr 490:11-
20
Felson AJ Pavao-Zuckerman M Carter T Montalto F Shuster B Springer N
Stander E Starry O (2014) Mapping the design process for urban ecology re-
searchers. BioScience 63(11):854-865
Fioretti R Palla A Lanza LG Principi P (2010) Green roof energy and water re-
lated performance in the Mediterranean climate. Building and Environment
45(8):1890-1904
39
Friedrich CR (2005) Principles for selecting the proper components for a green
roof growing media. p. 262-273. In: Proc. of 3rd North American Green Roof
Conference: Greening Rooftops for Sustainable Communities, Washington,
DC. 4-6 May 2005.
Gaffin SR Khanbilvardi R Rosenzweig C (2009) Development of a green roof
environmental monitoring and meteorological network in New York City. Sen-
sors 9(4):2647-2660
Gaffin SR Rosenzweig C Khanbilvardi R et al (2011) Stormwater retention for
a modular green roof using energy balance data. Columbia University.
Gaumont-Guay D Halsall R (2013) The carbon balance of a Pacific west coast
green roof. In: Proc. of Cities Alive: 11th Annual Green Roof and Wall Con-
ference. San Francisco. 23-26 Oct 2013
Getter KL Rowe DB (2009) Substrate depth influences sedum plant communi-
ty on a green roof. HortScience, 44(2):401407
Getter KL Rowe DB Andresen JA Wichman IS (2011) Seasonal heat flux
properties of an extensive green roof in a Midwestern U.S. climate. Energy and
Buildings 43:35483557
Getter KL Rowe DB Robertson GP Cregg BM Andresen JA (2009) Carbon
sequestration potential of extensive green roofs. Environment Science & Tech-
nology 43(19):75647570
Glass CC Johnson PA (2009) Monitoring of a new green roof for water quality
and quantity. In: Proc. of Greening Rooftops for Sustainable Communities
Conference, Awards & Trade Show. Baltimore, MD. 30 April to 2 May, 2008
Gnecco I Palla A Lanza LG LaBarbera P (2013) A green roof experimental site
in the Mediterranean climate: The storm water quality issue. Water Science &
Technology 68(6):1419-1424
Gregoire B Clausen J (2011) Effect of a modular extensive green roof on
stormwater runoff and water quality. Ecological Engineering 37(6):963-969
Griffin W (2013) The effects of green roof substrate composition on plant
growth and storm water retention of Mid-Atlantic green roofs. Diss. U of
Maryland
Hanks R (1992) Applied soil physics: Soil water and temperature applications,
2nd ed. Springer-Verlag
Havens KE Aumen NG (2000) Hypothesis-driven experimental research is
necessary for natural resource management. Env Management 25(1):1-7
Hendricks J Calkins M (2006) The adoption of an innovation: barriers to use of
green roofs experienced by Midwest architects and building owners. J of Green
40
Building 1:148-168
Jensen ME Burman RD Allen RG (1990) Evapotranspiration and irrigation
water requirements. ASCE Manuals and Reports on Engineering Practice No.
70. Am. Society of Civil Engineers, NY
Jim CY (2012) Effect of vegetation biomass structure on thermal performance
of tropical green roof. Ecological Engineering 8(2):173-187
Jim CY He HM (2010) Coupling heat flux dynamics with meteorological con-
ditions in the green roof ecosystem. Ecological Engineering 36(8):1052-1063
Karban R Hunzinger M (2006) How to do ecology: A concise handbook.
Princeton University Press, NJ
Kim S-C Park B-J (2013) Assessment of temperature reduction and heat budg-
et of extensive modular green roof system. Korean J of Horticultural Science &
Technology 31(4):503-511
Klett JE Bousselot JM Koski RD O’Connor TP (2012) Evaluation of green
roof plants and materials for semi-arid climates. USEPA National Risk Man-
agement Research Laboratory, Office of Research & Development,
USEPA/600/R-12-592 Sep. 2012: 78pp
Kurtz T, et al (2010) Stormwater management facilities monitoring report: De-
cember 2010. Bureau of Environmental Services, Sustainable Stormwater
Mgmt Program, City of Portland, OR: 172pp
Lea-Cox JD (2012) Using wireless sensor networks for precision irrigation
scheduling. In: Kumar M (ed) Problems, perspectives and challenges of agri-
cultural water management. InTech Press, Rijeka, Croatia, p 233-258
MacIvor JS Margolis L Puncher CL Matthews BJC (2013) Decoupling factors
affecting plant diversity and cover on extensive green roofs. J. of Environmen-
tal Management 130:297-305
Moody S, Sailor DJ (2013) Development and application of a building energy
performance metric for green roof systems. Energy and Buildings 60:262-269
Mooney-Bullock R Buffam I Bolan M (2012) Urban learning laboratory fea-
tures four educational, monitored green roofs. In: Proc. of Cities Alive: 10th
Annual Green Roof & Wall Conference. Chicago. 17-20 October 2012
Morgan S Celik S Retzlaff W (2012) Green roof stormwater runoff quantity
and quality. J of Env Engineering 139(4):471-478
Nagase A Dunnett N (2013) Establishment of an annual meadow on extensive
green roofs in the UK. Landscape & Urban Planning 112:50-62
41
Nagengast A (2013) Energy performance impacts from competing low-slope
roofing choices and photovoltaic technologies. Diss. Carnegie Mellon Univer-
sity
Neufeld RD Monnell J Ries RJ (2009) Monitoring Protocol and Data Collec-
tion: Comparison of the Runoff Water Quantity, Quality and Thermal Perfor-
mances of Two Green Roof Technologies: Thin vs. Thick. Sch. of Engin., U
Pittsburgh
Nouri H Beecham S Kazemi F Hassanli AM (2012) A review of ET measure-
ment techniques for estimating the water requirements of urban landscape veg-
etation. Urban Water 10(4):247-259
Oke TR (1978) Boundary Layer Climates. Methuen Company, NY
Olivieri F Di Perna , D'Orazio M Olivieri L, Neila J (2013) Experimental
measurements and numerical model for the summer performance assessment
of extensive green roofs in a Mediterranean coastal climate. Energy & Build-
ings 63:1-14
Omar MS, Quinn MM, Buchholz B, Geiser K (2013) Are green building fea-
tures safe for preventive maintenance workers? Examining the evidence. Am. J
of Industrial Medicine 56:410423.
Onmura S Matsumoto M Hokoi S (2001) Study on evaporative cooling effect
of roof lawn gardens. Energy & Bldgs. 33:653-666
Onset Computer Corporation (2012) Monitoring green roof performance with
weather stations. MKT1017-0912, Bourne, MA
Palla A Gnecco I Lanza LG (2012) Compared performance of a conceptual
and a mechanistic hydrologic models of a green roof. Hyd. Processes 26(1):73-
84
Panayiotis N Tsiotsiopoulou P Chronopoulos I (2003) Soil amendments reduce
roof garden weight and influence the growth rate of Lantana. Hortscience
38(4):618-622
Pearlmutter D Rosenfeld S (2008) Performance analysis of a simple roof cool-
ing system with irrigated soil and two shading alternatives. Energy & Build-
ings 40(5):855-864
Peng LLH Jim CY (2013) Green-roof effects on neighborhood microclimate
and human thermal sensation. Energies 6(2):598-618
Pierre J Bisby L Anderson B MacDougall C (2010) Thermal performance of
green roof panels in sub-zero temperatures. J of Green Building 5(2):91-104
PlaNYC (2008) Sustainable stormwater management plan. City of New York
42
Rana G Katerji N (2000) Measurement and estimation of actual evapotranspi-
ration in the field under Mediterranean climate: a review. European J of
Agronomy 13(2-3):125-153
Retzlaff W Celik S Morgan S Graham M Luckett K (2010) Into the wind
wind tunnel testing of green roof systems. In: Proc. of Cities Alive: 8th Annual
Green Roof & Wall Conference. Vancouver, BC. 30 Nov to 3 Dec 2010
Rosatto H Meyer M Laureda D et al (2013) Water retention efficiency of green
roof systems in extensive and intensive type covers. Revista De La Facultad
De Ciencias Agrarias 45(1):169-183
Rostad N White S DiGiovanni K Montalto F (2011) Canopy interception in
vegetated stormwater management features. In: Proc. of American Geophysi-
cal Union 2011 Meeting, San Francisco, California. 5-9 December 2011.
Rowe DB (2011) Green roofs as a means of pollution abatement. Environmen-
tal Pollution 159(8-9):2100-2110
Rowe DB Kolp MR Greer SE Getter KL (2014) Comparison of irrigation effi-
ciency and plant health of overhead, drip, and sub-irrigation for extensive
green roofs. Ecol Engineering 64:306-313
Schneider K (2011) Quantifying evapotranspiration from a green roof analyti-
cally. Master’s Thesis, Villanova University
Seidl M Gromaire M-C Saad M DeGouvello B (2013) Effect of substrate depth
and rain-event history on the pollutant abatement of green roofs. Environmen-
tal Pollution 183:195-203
Sherrard JA and JM Jacobs (2012) Vegetated roof water-balance model: Ex-
perimental and model results. J of Hydrologic Engineering 17(8): 858-868
Schroll E Lambrinos J Righetti T Sandrock D (2011) The role of vegetation in
regulating stormwater runoff from green roofs in a winter rainfall climate. Ecol
Engineering 37(4):595-600
Sidwell A Gibbs-Alley J Forrester K Jost V Luckett K Morgan S Yan T Noble
B W Retzlaff (2008) Evaluation of the thermal benefits of green roof systems.
In: Proc. of 6th Annual Greening Rooftops for Sustainable Communities Con-
ference, Baltimore, MD. 30 April to 2 May, 2008
Simmons M Gardiner B Windhager S Tinsley J (2008) Green roofs are not
created equal: The hydrologic and thermal performance of six different exten-
sive green roofs and reflective and non-reflective roofs in a sub-tropical cli-
mate. Urban Ecosystems 1(4):339-348
Song U Kim E Bang JH Son DJ Waldman B Lee EJ (2013) Wetlands are an
effective green roof system. Building & Environment 66:141-147
43
Spirn AW (1984) The granite garden: Urban nature and human design. Basic
Books, NY
Starry O (2013) The comparative effects of three sedum species on green roof
stormwater retention. Dissertation, U of Maryland
Stovin V Vesuviano G Kasmin H (2012) The hydrological performance of a
green roof test bed under UK climatic conditions. J of Hydrology 414:148-161
Susca T Gaffin SR Dell'Osso GR (2011) Positive effects of vegetation: Urban
heat island and green roofs. Environmental Pollution 159(8-9):2119-2126.
Teemusk A, Mander U (2007) Rainwater runoff quantity and quality perfor-
mance from a green roof: The effects of short-term events. Ecological Engi-
neering 30(3):271-277
Tilman D (1989) Ecological experimentation: Strengths and conceptual
problems. In: GE Likens (ed) Long-term studies in ecology: Approaches and
alternatives. Springer Verlag, p 136-157
Thuring CE, Dunnett N (2014) Vegetation composition of old extensive green
roofs (from 1980s Germany). Ecological Processes 3:4
Toland DC, Haggard BE, Boyer ME (2012) Evaluation of nutrient concentra-
tions in runoff water from green roofs, conventional roofs and urban streams.
Transactions of A S A & B E 55(1):99-106
Tsiotsiopoulou P, Nektarios A, Chronopoulos I (2003) Substrate temperature
fluctuation and dry-weight partitioning of Lantana grown in four green roof
growing media. J of Horticultural Science & Biotechnology 78(6):904-910
VanSeters T, Rocha L, Smith D, MacMillan G (2009) Evaluation of green
roofs for runoff retention, runoff quality, and leachability. Water Quality Re-
search Journal of Canada 44(1):33-47
Voyde E, Fassman E, Simcock R (2010) Hydrology of an extensive living roof
under sub-tropical climate conditions in Auckland, New Zealand. J of Hydrol-
ogy 394(3-4):384-395
Wadzuk B, Schneider D, Feller M, Traver R (2013) Evapotranspiration from a
green-roof storm-water control measure. J of Irrigation & Drainage Engineer-
ing 139(12):995-1003
Wang X, Zhao X, Peng C, Zhang X, Wang J (2013) A field study to evaluate
the impact of different factors on the nutrient pollutant concentrations in green
roof runoff. Water Science & Technology 68(12):2691-2697
Walmart et al. (2013) Green roof performance: A cost-benefit analysis based
on Walmart’s Chicago store. Jan. 31, 2013 report
Wanielista M, Minareci M, Catbas N, Hardin M (2011) Green roofs and wind
44
loading. Florida Dept. of Environmental Protection and U of Florida Storm-
water Management Academy, FDEP:934 stormwa-
wa-
ter.ucf.edu/sealofapproval/GreenRoofsandWindLoadingsDraftFinalRepor
t.pdf
Welker AL, Mandarano L et al. (2013) Application of a Monitoring Plan for
Storm-Water Control Measures in the Philadelphia Region. Journal of Envi-
ronmental Engineering. 139(8):1108-1118.
Williams C (2013) Portland State will build green roof research site on Walmart’s
new North Portland store. 23 Oct 2013 News: Portland State University.
Whittinghill LJ, Rowe BD, Schutzki R, Cregg BM (2014) Quantifying carbon se-
questration of various green roof and ornamental landscape systems. Land-
scape & Urban Planning 123:41-48
Wong NH, Chen Y, Ong CL, Sia A (2003) Investigation of thermal benefits of
rooftop garden in the tropical environment. Build. & Environ.38:261-270
Xu CY, Chen D (2005) Comparison of seven models for estimation of evapotran-
spiration and groundwater recharge using lysimeter measurement data in Ger-
many. Hydrological Processes 19(18):3717-3734
Zapater-Pereyra M, van Dien F, van Bruggen JJA, Lens PNL (2013) Material se-
lection for a constructed wetroof receiving pre-treated high strength domestic
wastewater. Water Science & Technology: Journal of International Association
on Water Pollution Research 68(10):2264-2270
Zhang CL, Zou X-Y, Yang P et al (2007) Wind tunnel test and cs tracing study on
wind erosion of several soils in Tibet. Soil & Tillage Research 94:269-82
Zhao MJ, Srebric J (2012) Assessment of green roof performance for sustainable
buildings under winter weather conditions. J of Central South University of
Technology 19(3):639-644
Zheng, Y and MJ Clark (2013) Optimal growing substrate pH for five sedum spe-
cies. Hortscience 48(4):448-452
... One hesitation of adopting green roofs in Utah and other parts of the arid west is uncertainty about their performance due to a lack of performance data. Among these few green roof project examples in Northern Utah ③ (Table 1), the project at University of Utah's main library is perhaps the only one that collects performance data, but its main focus is to understand the evapotranspiration rates of green roof plants [22] [23] . This to Provo, with Salt Lake City (state capital) in between. ...
... Over a period of 2 years the green roof sequestered 375 g C m − 2 . A study from a Sedum covered Canadian green roof using surface chambers to quantify carbon exchange reported a net carbon uptake of 440 g C m −2 year −1 (Skabelund et al., 2015). ...
Article
The CO2 surface-atmosphere exchange of an unirrigated, extensive green roof in Berlin, Germany was measured by means of the eddy covariance method over a full annual cycle. The present analysis focusses on the cumulative green roof net ecosystem exchange of CO2 (NEE), on its seasonal variation and on green roof physiological characteristics by applying a canopy (A-gs) model. The green roof was a carbon sink with an annual cumulative NEE of -313gCO2m(-2)year(-)(1), equivalent to -85gCm(-2)year(-)(1). Three established CO2 flux gap-filling methods were applied to estimate NEE and to study the performance during different meteorological situations. A best estimate NEE time series was established, which chooses the gap filling method with the highest performance. During dry periods daytime carbon uptake was shown to decline linearly with substrate moisture below a threshold of 0.05m(3)m(-3), whereas night-time respiration was unaffected by substrate moisture variation. The roof turned into a temporary C source during dry conditions in summer 2015. We conclude that the carbon uptake of the present green roof can be optimized when substrate moisture is kept above 0.05m(3)m(-3).
Article
Full-text available
Green roof systems have been increasingly implemented to enhance vegetation cover and associated ecosystem services in urban spaces, with primary goals being the reduction of peak surface runoff, enhanced water quality, and mitigation of urban heat island effects. Recently, green roofs have also received attention as a means to enhance carbon sequestration, but direct measurements of greenhouse gas fluxes from established green roof systems are largely lacking. Here we present observations of CO2 and CH4 fluxes from substrates of experimental extensive green roof units that varied in vegetation type (Sedum spp., and a native meadow species mix), substrate depth, substrate type (high vs. low organic matter content), and irrigation. We predicted that substrate CO2 effluxes would be higher in high-organic-matter substrates and that systems with high organic matter would potentially act as CH4 sources. Substrate fluxes were low compared to natural soils, with seasonal means ranging from an efflux of 0.1–0.4 µmol CO2 m-2.s-1 and uptake of ~0.00–0.04 nmol CH4 m-2.s-1, with higher fluxes late in the growing season. CO2 fluxes showed large increases in response to irrigation and were higher from the high-organic-matter substrate and with increased substrate depth. The strength of the CH4 sink increased in response to prior irrigation treatments, and CH4 emissions were detected only on low-organic-matter substrates early in the growing season. No effects of vegetation type were detected for either CO2 or CH4 flux. Our results indicate that high levels of organic matter in green roof substrates may enhance aerobic soil respiration but are not associated with CH4 emissions, which instead were only detected in low-organic-matter substrates.
Chapter
Full-text available
This chapter presents case studies of three tallgrass prairie conservation sites and nine ecoregional green roofs located in the Coastal Plains and Interior Lowlands of North America. Prairies were once one of the top ten ecologically rich ecosystems in the world with over 300 species of forbs and 70 species of grasses native to the tallgrass prairie. Today, less than one percent of tallgrass prairie remains across North America. On green roofs, however, 145 species of plants native to tallgrass prairies have been trialed on the ecoregional green roofs featured in this chapter located in Kansas, Nebraska, Texas, and Missouri. This chapter demonstrates how the diverse tallgrass prairie can inspire a variety of habitat types for green roofs including vegetation from dry, mesic, and wet prairie habitats on sloped and flat roof decks.
Chapter
Full-text available
This chapter investigates the theoretical background of environmental and ecological factors that can be used to inform the design of ecoregional green roofs. Ecoregions are defined by major or minor delineations of plant communities and their interactions with other resident or transient organisms. By observing and learning how native vegetation adapts and thrives in its natural settings, green roof researchers, educators, and designers can learn how to make decisions about the resourceful use of native vegetation on green roofs. This chapter discusses how green roofs must respond to environmental factors such as heat stress, drought, and varied slope and soil conditions, and how these factors can inform the design of green roofs with native vegetation. The chapter ends with a discussion regarding how ecoregions are defined in this book and are employed in the case studies in Part II of this book.
Article
Green roofs can provide environmental benefits that include increased building insulation, mitigating urban heat islands, providing aesthetic value, reducing runoff and storm water flooding in urban environments, improving air quality by sequestering pollutants, cooling photovoltaic panels to improve their function, and providing habitat for fauna and flora. Until very recently, improvements of green-roof environmental services had been achieved largely by horticulturalists, engineers, and architects. In recent years, ecologists have increased their participation, implementing ecological theory for enhancing biodiversity, and selecting specific plant assemblages for other environmental services such as carbon sequestration and for providing cooler roofs. Moreover, ecologists can use green roofs as relatively novel habitats for testing and developing ecological theory. This special issue is devoted to fostering input from ecologists for advancing the environmental and ecosystem services of green roofs. A wide range of ecologists can explore the topic of the ecological aspects of green roof design and implementation including island biogeography theory, niche theory and null models, the role of environmental heterogeneity, invasion ecology, and plant selection. They can contribute ecological methodology and study design for strong inference.
Conference Paper
Full-text available
The Green Roof Innovation Testing (GRIT) Lab at the University of Toronto is a multi-year research project comparatively analyzing thirty-three extensive green roof modules with variables of composition and maintenance. Each module is continuously monitored through an array of nine thermal and hydrological sensors. The paper will focus on the troubleshooting and calibration processes of two sensors contributing to the hydrological modeling of the modules: 1. Decagon 5TE Moisture sensors embedded within the planting substrate, and 2. Hydro Services TB6 rain gauge, measuring water volume drained through each module. Each of these sensors was designed for slightly differing applications, and thus needed modification to accommodate the specific conditions of the research project. The calibration process involved pairing substrate moisture with the dielectric permittivity of the two planting substrate types in the modules, and designing, fabricating and testing new components of the rain gauges.
Article
Full-text available
Flash floods in urban areas caused by overload of drainage networks are a recurrent problem of raising importance. Greenroofs retain part of the stormwater, lowering surface flow and generating runoff hydrographs with lower and delayed peak flows. Therefore, this technology can contribute to mitigate the overload of drainage networks. The results of the study that was carried out in City of Buenos Aires along almost two years, showed that the retention capacity of the tested lots tasted varied, depending upon precipitation, coverage and depth of the substrate. With precipitation less than or equal to 20 mm, the retention fraction was high (73% to 100%), and when precipitation reached 35 to 40 mm, the maximum percentage of retention was around 60%. However, when the rainfall was approximately 100 mm, the retention fraction was reduced substantially, reaching values nearing 30%. The results of the test showed that green roofs system represent a good alternative in the integrated management of water runoff in urban watersheds.
Article
Full-text available
This study evaluated nutrient concentrations in runoff water from conventional roofs, green roofs, and urban streams, focusing on the impacts of compost addition at the industry standard of 15% by volume to green roofs at installation. Water samples were collected during selected rainfall events (n = 9) during calendar year (CY) 2008 and from the urban stream approximately monthly during baseflow conditions. Water samples were analyzed for ammonium-nitrogen (NH4-N), nitrite-N (NO 2-N), nitrate-N plus NO 2-N (referred to as NO 3-N), total N (TN), soluble reactive phosphorus (SRP), total P (TP), and total organic carbon (TOC). The concentrations of SRP, TP, TN, and TOC were significantly greater in runoff from the green roofs that received compost during installation than from the conventional roofs or the green roof without compost addition, while the NH 4-N, NO 2-N, and NO 3-N concentrations in stormwater runoff were generally not significantly different across the conventional or green roofs. Nutrient concentrations in the study streams, except for TOC, generally increased with the percentage of urban and pasture land use in the stream catchment, and the exponential relationship was generally strongest (higher R 2) for NO 3-N, TN, and P. Nutrient concentrations in stormwater runoff from the green roof without added compost were within the range observed across the study streams. However, nutrient concentrations in stormwater from the green roofs with compost were more variable when compared to the selected streams, and P concentrations were significantly greater in stormwater from the compost-amended green roofs compared to that measured in the study streams. The data collected in this study provide evidence that compost applied at the industry standard of 15% by volume to maintain plant growth is contributing to increases of nutrients in stormwater runoff 17 to 23 months after installation. Future studies should focus on compost additions to green roofs that maximize plant growth and survival while minimizing nutrient (particularly P) loss in stormwater. © 2012 American Society of Agricultural and Biological Engineers.
Conference Paper
Full-text available
The Civic Garden Center of Greater Cincinnati " s Green Learning Station is home to an accessible flat roof comparing extensive, bio-tray and intensive vegetated roof systems and the city " s first sloped green roof. The intensive roof section is a vegetable garden producing food year-round. Monitors embedded in the soil and downspouts measure soil moisture and temperature, runoff volume and rate. Additional data is being collected to compare runoff water quality between the green roofs and their traditional counterparts. The Green Learning Station roof is open to the public (accessible via a prominent staircase) with signage explaining the systems and their benefits. Tours are given regularly to secondary science classes, college courses, the general public and building and design professionals. A self-guided tour uses QR codes linked to videos to prompt visitors to interact with the site. The project is an example of a successful collaboration between a non-profit (Civic Garden Center), the private sector (in-kind donors Tremco Incorporated, Urbanalta, Green City Resources, Melink) and the public sector (University of Cincinnati professors and students, Metropolitan Sewer District of Greater Cincinnati, Cincinnati Park Board and US EPA). In 2007, the Civic Garden Center (CGC) began dreaming about how it might convert its storage building, a former gas station, into a demonstration site for sustainable landscape practices, including a wide variety of green infrastructure for stormwater management. Since 1942 the organization had been educating Cincinnatians about best practices for urban gardening, which had long included environmentally safe methods of pest and weed control and responsible
Article
To determine the optimal growing substrate pH values for Sedum plants, Sedum album , Sedum reflexum ‘Blue Spruce’, Sedum spurium ‘Dragon’s Blood’, Sedum hybridum ‘Immergrunchen’, and Sedum sexangulare were grown in containers using peatmoss and perlite-based substrates at five target pH levels (i.e., 4.5, 5.5, 6.5, 7.5, and 8.5). Optimal pH levels, calculated from dry weight regression models, were 6.32, 6.43, 5.71, 6.25, and 5.91 for S. album , S. reflexum , S. spurium , S. hybridum , and S. sexangulare , respectively, and 5.95 overall. Sedum spurium dry weight varied the most among pH treatments (i.e., 9.5 times greater at pH 6.3 vs. 8.3), whereas S. reflexum varied the least (i.e., 1.3 times greater at pH 6.3 vs. 4.4), indicating species-specific growth responses to growing substrate pH. These findings identified a narrow range of optimal growing substrate pH levels within a wider pH range tolerated by five Sedum spp. Therefore, by adjusting substrate pH to optimal levels, Sedum growth can be maximized.
Article
Four potential green roof substrates were evaluated in a field study for their thermal characteristics and their impact on Lantana camara dry-weight accumulation. The substrates were a) a sandy loam soil (S), b) the sandy loam soil amended with urea formaldehyde resin foam (S:F) at a proportion of 60:40 v/v, c) the sandy loam soil mixed with peat and perlite (S:P:Per) at a proportion of 50:30:20 v/v, respectively and, d) peat amended with urea-formaldehyde resin foam (P:F) at a proportion of 60:40 v/v. Measurements included the monitoring of the substrate temperature and the determination of the dry-weight accumulation rhythm of shoots, leaves and roots. In addition, the lateral spreading of the roots was recorded and the shoot/root ratio determined. Temperature fluctuation within the substrates was found to depend on the type of the substrate, plant growth, and season. Temperature fluctuation was high in S, moderate in S:F and S:P:Per and low in P:F. Shoot and root dry-weight accumulation was reduced in P:F, while root growth was promoted in S substrate. Differences in dry-weight accumulation and the lateral spreading of the roots were affected more by the water holding capacity of the substrate than by the substrate temperature.
Article
The Energy Independence and Security Act of 2007 (EISA) includes Section 438, an important stormwater-related provision titled, 'Storm Water Runoff Requirements for Federal Development Projects'. The law applies to all federal facilities with a development or redevelopment footprint of more than 5,000ft2, including all Department of Defense facilities, General services Administration (GSA) facilities, and facilities operated or owned by other federal departments or agencies. The most important aspect of Section 438 is the new standard for stormwater management that is specified by the use of the term, 'to the maximum extent technically feasible (METF). Meeting METF will require site engineers to approach stormwater management in a completely different fashion. The final guidance document was planned for release by April 2009 and will meet its intended aim as a framework for designers and engineers working on federal projects to use in selecting and designing effective stormwater management systems.
Article
This three-year study evaluates the quantity and quality of runoff from an extensive green roof on a multistory building in Toronto. Laboratory physical, chemical, and leachate analyses of eleven commercially available green roof growing media were also undertaken to help identify the potential influence that the growing media may have on runoff chemistry. Continuous precipitation and runoff data collected over 18 months outside of the winter period indicated that the green roof discharged 63% less runoff than a neighbouring conventional modified bitumen roof. Runoff volumes from the green roof averaged 42% less than the conventional roof in April and November, and between 70 and 93% less during the summer months. Water samples were collected from both roofs during 21 rain events in 2003 and 2004 and analyzed for general chemistry (e.g., pH, total suspended solids), metals, nutrients, bacteria (n = 16), and polycyclic aromatic hydrocarbons (n = 18). Loads of most chemical variables in green roof runoff were lower than from the conventional roof. Exceptions included constituents such as calcium, magnesium, and total phosphorus, which were either naturally present in the media or were added to promote plant growth. Total phosphorus concentrations in green roof runoff were significantly higher than the conventional roof (α = 0.001), and regularly exceeded the Ontario receiving water objective (0.03 mg/L). Phosphorus concentrations fell significantly after the first year of monitoring (α = 0.001), suggesting that the nutrient is being leached from the media. Chemical analyses of green roof growing media showed that levels of most constituents were similar to or lower than typical background concentrations for agricultural soils in Ontario. However, leachate concentrations from several media exceeded receiving water standards for phosphorus, aluminum, copper, iron, and vanadium. This study highlights the importance of engineering green roof media to minimize leaching of nutrients and other contaminants while maintaining their ability to support plant growth.