Green vegetated roofs in the theory of landscape and architecture

Ecological space + a little social space: Green roof on California Academy of Sciences

Ecological space + (a little) Social Space (Green roof on California Academy of Sciences)

In Germany, vegetated green roofs are often classified as:

  • intensive (ie treated as a garden, typically with exotic plants, irrigation, turf and social use)
  • extensive (ie treated as habitat, without irrigation or maintenance)

I prefer to look at green roofs from a more Vitruvian standpoint and consider their roles as:

  • visual space [Delight]
  • ecological space [Firmness]
  • social space [Commodity]

Image courtesy clickykbd.  The California Academy of Sciences designed by Renzo Piano and ‘The Living Roof´s 1.7 million native plants were specially chosen to flourish in Golden Gate Park´s climate.’  There is a small terrace for viewers but the predominant role of the green roof is Ecological Space.

Visual space+ social space: Singapore School of Art Media and Design

Visual space+ Social Space (Singapore School of Art Media and Design)

” This 5 story facility sweeps a wooded corner of the campus with an organic, vegetated form that blends landscape and structure, nature and high-tech and symbolizes the creativity it houses.”  The  green roof is open to the public and, like the roof of Australia’s Parliament Building in Canberra, is surfaced with mown grass. Image courtesy teddy-rised It is  not ecological space:  the grass is irrigated and  mown. The building was designed by CPG Consultants.

Ecological space only: University of Illinois at Springfield

Ecological Space only (University of Illinois at Springfield)

The Springfield Illinois green roof is ecological space, only. It is not visual space or social space. Image courtesy jeremywillburn,

Visual space + Social space: Roof on the HQ of the American Society of Landscape Architects

Visual space + Social space:Roof ( HQ of the American Society of Landscape Architects in Washington DC)

The green roof on the American Society of Landscape Architects is visual space and social space but not ecological space – at leasst not as shown in this photograph (the roof has other eco-friendly characteristics).  Image courtesy drewbsaunders.

36 thoughts on “Green vegetated roofs in the theory of landscape and architecture

  1. Tom Turner Post author

    The marketing imaage of the Dalian Shide Stadium is fantastic and I would classify it as a green wall, rather than a green roof – it is an easier choice than for the Singapore School of Art and Media.

  2. DAN

    After all this rain over the weekend – it seems that there is so much that all the water could be irrigating deeper rooted planting if stored. Why have mains water irrigation systems atop of rooves? Crazy! What a waste of the clients money? And if there is no rain then a drought tolerant species to be used.

    The Singapore roof top is almost unecological surely? It requires very high maintenance, is about as good as a car park for flora and fauna, and with little aesthetic value – however the buildings sculptural qualities are at least mimicked in its shaven haircut…

  3. David Matzdorf

    “Why have mains water irrigation systems atop of rooves?”

    Because it vastly increases the range of plants that can be grown and, if properly managed, scarcely uses any water at all. My green roof has under-soil irrigation, but I use it exclusively as “death control”, which basically means that it runs for 15-20 minutes at a time, 3-4 times a year. I have an extensive green roof that is a horticultural space as well as an ecological one – it’s a garden as well as a carbon sink and a flood control measure. 21st century London can experience droughts of 4-6 weeks at any time between March and October. Having an occasionally-used irrigation system means that I can grow 1.0-1.5m tall shrubs and low-growing leafy perennials in my 0.1m deep soil, which greatly increases the mass of vegetattion, as well as making a more varied garden.

    If I could pump rainwater back up there to do the job, I would, but I need all of the 450 litres of rainwater that I can collect to water my “terrestrial” garden and the water savings would not merit the expense of the pump .

  4. Gordon Evans

    Extensive green roofs are favoured with permitting advantages – directly translateable into Euros and Cents for the Developer – in North Germany not because of their intrinsic ecological value (the Sedum species that form the core of most of them are useful suppliers of nectar in the flowering period, but otherwise offer very little in the way of habitat) but because of the way that they delay water delivery into the public drainage network, thus reducing handling costs at the macro scale during peak rain events. All of the published roofs here will do this and would therefore fall into the category of “Ecological Space”. In themselves, though, their ecological value is marginal. I too would dearly love to see photos of David’s 1.5 metre tall shrubs on only 10 cm of planting substrate, something that up until now I would have held for unlikely.

  5. Tom Turner Post author

    My own green roof was installed 30 years ago. The roof was covered with a 25mm layer of sand, laid in 1926 when we bought the house and I replaced it with a 50mm layer of sand-peat mix. When it rains after a dry spell it is over 1 hour before any water is discharged so, yes, I am in full agreement about the SUDS/LID value of green roofs. But I disagree about its lack of habitat value. It is very popular with birds. They love harvesting my hard fescues for nest building: they are so much more suitable than poa and lolium species. The birds also find a lot to eat amongst the thin cover of grass and I think they rather like the seed of the chives which are spreading across the roof. One can imagine that chives go well with raw slugs. Lastly, the roof cannot be regarded as a cost. It lies on a built-up felt roof with a design life 15 years. Because the covering has protected the roof from expansion and contraction and loss of hydrocarbons, by evaporation, the roof has now kept the water (or most of it, to be exact) out for 83 years. It must also have kept the house warmer in winter and cooler in summer. I call that an excellent investment.

  6. christine

    Tom, Erika Tokarz in her paper ‘CEER Green Roof Project’ suggests that green roofs extend the life of a traditional roof by up to forty years by sheltering it from ultravoilet degradation. So given you installed your green roof 30 years ago perhaps that means you can expect another ten years? Or do we need to factor in the age of the existing roof at 53 years when you installed the green roof to get a realistic estimate?

    Interestingly Tokarz believes the first green roof was the Hanging Gardens of Babylon.[ ]

    And I suppose although the Hanging Gardens are considered to be gardens rather than architecture she may have a point regarding the technology necessary to construct and maintain the gardens. [ ]

    It is also interesting to note that according to Tokarz green roofs had been a sign of status in Germany since the 18th century. Perhaps explaining why the German’s now lead the world in the modern development of green roofs.

  7. Tom Turner Post author

    It has only had vegetation for 30 of its 53 years, but I put the age of my green roof at 83 years.

    Little is known about the ‘Hanging Gardens of Babylon’. Dalley suggested they were at Nineveh and fed by an aqueduct but this now seems unlikely. There is good evidence for a garden on the slope of a tel in Nineveh and it is possible that the Babylon garden was of the same type.

  8. Gordon Evans

    Germany’s expertise in green roofs has three underlying causes: 1. The populace is predisposed to like them, 2. The Germans commonly build underground structures that extend outside of the footprint of the above ground structures, often to the site boundary and 3. Coercion. Planning regulations often demand them and also lay down minimum standards. Without item 3 there would be a lot less green roofs in Germany.

  9. Gordon Evans

    And on a technical note, the use of either sand or peat on anything other than the domestic scale for an extensive roof is not advisable. Organic materials decay over time, reducing the original depth and discharging sludges. Sand drains badly over large areas and blocks up the pipes. All of the roofs I have seen or been responsible for in Germany have used a stable soil substrate based on lava, expanded clay and selected rubble. It is entirely lifeless and intended to stay that way for as long as possible. Where ecosystems do develop – usually as a result of birch seedling colonisation – they need to be eradicated at once before the roof begins to drain badly and the sedum skin dies off. As I say, the ecological value of this kind of roof is really very limited. Except for the bees.

  10. Tom Turner Post author

    Thank you for the information, but I am puzzled, particularly by the phrase ‘not advisable’. Surely (1) it is normal and natural for organic material to decay over time (2) sharp sand is sufficiently permeable to water and, in any case, part of the reason for having a green roof is to act as a detention facility for rainwater (3) upstands and filters can be used to stop drains being blocked by sand (4) it is desirable for an ecosystem to develop and the plants will provide another means of dispersing water: evapo-transpiration.
    I have been praising Germany’s green roof policy for years, without much investigation, and would be sorry to learn that the ‘green’ roofs I admire are intended to be near-lifeless brown deserts!

  11. Gordon Evans

    I am describing here the large scale, extensive roofs that go over office developments. Weight is a key element. Anything over 2 KN per square metre is too heavy, this rules sand out. The extensive substrates have particle sizes from 20 to 5 mm, nothing smaller, definately no fines. The spaces between the particles drain relatively quickly and never become waterlogged. This is the other thing that rules sand out, it always becomes (more) waterlogged and this leads to a very patchy roof. Decaying organic matter also leads to waterlogging. The actual water retention in the substrates comes from the properties of the lava, expanded clay and rubble elements themselves, all work like small sponges. Sedums are a must in the mix, because most drought-tolerant grasses will only be green in spring and autumn and because sedums are the least maintenance intensive and visually most effective plant for this kind of substrate, and they rarely fail, establishing a closed mat within 2 years. Filters and upstands around drains are maintenance issues completely avoided on sedum roofs. There are many other kinds of German green roofs with their accent more on habitat creation and ecological diversity, but the sedum roof is by far the commonest. A usual specification is for two one day maintenance visits a year after establishment. The “Sedum Habitat” is by no means a brown desert, the many species are extremely colourful both during the long flowering season and again for many varieties in autumn when the foliage turns red. But, habitat diversication is definately not wanted and the aim of the maintenance visits is expressly to stop it taking place.

  12. Tom Turner Post author

    Thank you for the clarification – but you have put me off sedum roofs (despite an affection for bees). My current interest in green roofs arises from the design of a new building for the University of Greenwich School of Architecture and Construction (to be built in Stockwell Street, Greenwich and scheduled for completion in 3-4 years). About 50% of the roof will be vegetated roof and I would like it to be visual space, ecological space and social space – as well as a SUDS/LID facility and feeding ground for bees. The architects are

  13. Gordon Evans

    I will be interested to hear how you get on. The sedum roof is the answer to the kind of cost pressures that you will be up against. Visual, ecological and social spaces are much more expensive, both in their realisation and their maintenance. I hope you have a QS and a client who appreciate this! Good luck.

  14. David Matzdorf

    Thanks for the responses. Here are some responses to your responses.

    “David I would love to see photographs of your roof garden if you have them.”

    Here are some links to threads on an international exotic-gardening website where I am one of the moderators, which include numerous photos at various seasons:

    June 2008:
    Autumn 2008:
    May 2009:

    “I too would dearly love to see photos of David’s 1.5 metre tall shrubs on only 10 cm of planting substrate, something that up until now I would have held for unlikely.”

    You can see the Leycesteria formosa (now about 1.5m tall), several Aloe striatula (approaching 1m) and three Dasylirion wheeleri (the largest is about 0.8m) on those photo-threads. There are also numerous Cistus x pulverulentus ‘Sunset’ of varying ages and one Cistus x skanbergii – they eventually reach over 0.5m and spread to about 1m. I have some small plants of Yucca aloifolia that I propose to experiment with next Spring, as well as some more species of Agave and some columnar (someday) Cacti and I am gradually widening the range of Bromeliads. The substrate is nothing more sophisticated than sandy loam soil, amended locally with extra grit and sharp sand when I plant xerophytes. I try to take full advantage of London’s substantial urban-heat-island effect in my plant selection – the lowest Winter temperature here since 2000 has been -5.5ºC.

    “I disagree about its lack of habitat value. It is very popular with birds…”

    Here I agree with both Tom and Gordon. My experience is that the defining issue, when considering the biodiversity and habitat potential of an extensive green roof, is the degree to which it is a monoculture. If you plant a lot of different plants, whether they are native or exotic, invertebrates, birds and small mammals will figure out a way to use it for food and shelter. There’s a great deal of fussiness about, often from the “native plants tendency”, which assumes that nature is something that we have to coddle and that requires precisely calculated conditions. My experience is much simpler: plant a lot of different stuff and other organisms will colonise it. If you plant nothing but Sedum, I suppose you will will get nothing but bees (not a bad thing in itself, as the bee population is in trouble). But why plant nothing but Sedum?

    Stefan Breneissen in Switzerland has done some interesting work that indicates that the other defining issue for creating a biodiverse green roof is to *vary* the depth of the substrate locally (I am working on this, although there are some structural-engineering issues). Stefan says that, if you have an extensive green roof where the depth of the substrate is the same throughout, you will be reducing its biodiversity value, whatever you plant. If you provide areas as deep as 200mm and areas as shallow as 60mm, different species will colonise different areas.

    “Lastly, the roof cannot be regarded as a cost. It lies on a built-up felt roof with a design life 15 years. Because the covering has protected the roof from expansion and contraction and loss of hydrocarbons, by evaporation, the roof has now kept the water (or most of it, to be exact) out for 83 years. It must also have kept the house warmer in winter and cooler in summer…”

    The notion that protecting a roof membrane from UV will extend its life is intuitively sensible. Mine (which has a PVC membrane) isn’t old enough yet for me to know (2000). The notion that a green roof will function as thermal insulation is also intuitive, but I think it’s not so, certainly not in a wet-Winter climate such as London’s. When I need my roof to help keep my house warm, it’s almost always damp and damp soil isn’t very good insulation. My observation is that the green roof definitely helps to keep my house cooler in Summer (if only when the soil is damp), but has no perceptible effect in Winter. At the same conference where I heard Stefan Breneissen present his findings, there were also studies from America presented that reinforce my observations.

    For those in the UK, there is a 4 page feature on my green roof in the current (28 November 2009) issue of “Amateur Gardening” magazine. Not exactly a learned journal, I know, but who am I to be choosy?

  15. Tom Turner Post author

    Thank you David for all this additional information. I am sure there is a photo of my green roof (in Greenwich) somewhere on this website but I can’t find it! Its character is visual and ecological space, not social space. I have not made any measurements but my perception is the same as yours: it helps to keep the house cool in summer but does nothing for its winter temperature. If correct, this is no small thing: I believe that air conditioning in hot countries uses more energy (per unit) than central heating in cold countries.
    Gordon: I can’t speak for the QS but I work for the client and will have opportunities to express my views. If costs are cut at some point, and we end up with a sedum roof, you can be sure that it won’t be a consequence of lack of input on the roof from one of the client’s spokespeople!

  16. Gordon Evans

    Great photos of your labour of love, David. The challenge is to find a system capable of industrialisation that lands somewhere between your roof garden and the sedum roof. I am a big fan of nutrient free substrates, partly because you have to be when working professionally on a roof whose loadings have to be precisely measured, and which you have to personally guarantee for 10 years. Here is a photo of a herbaceous garden (in its second year)on a roof in Hamburg that I did in 30 cm of extreme lightweight substrate:
    Geraniums seem to hate substrates, other things love it. Perhaps it might be the answer to your structural problems. Whatever, if green roofs are to become ecologically richer on the mass scale, substrates will be part of the solution. Perhaps you could do some experimenting?

  17. Christine

    Thankyou David and Gordon for the photographs.

    David do you also have links to photographs of the building so I can understand the roof in its context?

    Gordon are you aware of the use of the double height(?)space beneath the roof garden? I am not sure in terms of topology whether this example would be strictly a green roof or rather a podium style space? [ ]

    The Roppongi Hills project in Japan might be of interest. [ ] It includes a 4000m2 traditional Japanese garden.

  18. Tom Turner Post author

    Gordon: what about putting substrates on roofs in hot arid climates? I would have thought:

    – they would protect roof membranes
    – they would improve thermal insulation
    – even if vegetated only for 2 weeks/year the vegetation would be worth having
    – they could provide some cooling, and a lot more vegetation,if waste water was applied to the roof
    – vegetated roofs would reduce the dustiness of hot arid cities and make them less noisy
    – it is regrettable that the ancient practice of putting awnings on roofs in West Asia has been so largely neglected in recent times

    My outsider’s view is that as great a roof opportunity has been wasted in hot arid cities as in wet temperate cities – quite apart from the obvious fact that there should be great morphological differences between cities in these climate zones.

  19. Gordon Evans

    Christine: The space below the roof garden is an exhibition hall for trade fairs. The space had to be free of columns, hence the above-grade trusses and the stringent weight limits. The roof measures 140 by 50 metres, and I think can be classified as a “garden”.

    Tom: Here in the Middle East, green roof technology is not yet at German standards, although there are plenty of roof gardens. Many technical issues have been solved with the application of money and it is not uncommon for roofs to be simply loaded up with 1.50 m of earth and planted. The big problem here is irrigation, roof gardens in a hot climate are extremely water-intensive and fresh water, desalinated water, brackish ground water and treated sewage effluent are all running out fast. With their current high rates of consumption (ca. 275 l per day) one person here can irrigate 13 square metres of planting with TSE. If water consumption is scaled back to more sustainable levels (Hamburg average consumption is 120 l per day) this area should decrease. While the idea of green roofs in this arid climate is very inviting, the reality is I fear that they will prove to be completely unsustainable within the next 20 years.

  20. Tom Turner Post author

    Gordon: I can understand the popularity of roof gardens, but what about the potential for extensive green roofs – which are only green with vegetation for a few weeks/year?
    From what you say about water consumption, it sounds as though the Gulf states will, one day, be as densely populated as Xanadu

  21. Gordon Evans

    The European model of the extensive roof, not or hardly irrigated and with up to 20 cm of substrate is not an option here, the planting wouldn’t survive for more than a couple of days. Even xeriscape planting schemes require between 2 and 5 litres of water per square metre per day for most of the year. Adding in things like car washing and other non-recycleable uses – which generate no TSE – the actual per capita water consumption per day is around 500 l in the UAE, the highest in the world. To avoid the Xanadu scenario there are fortunately serious moves afoot to turn this around, and irrigated roof areas have already been identified on the permitting level as something to avoid.

  22. Christine

    Thankyou David. Would I be right in describing the general feel as a green oasis in a medium density urban setting?

    Gordon I asked the question about the use of the space beneath the garden because I was wondering if the garden/building relationship had been conservatively designed.

    If the integrity of some spaces are compromised by problems with a garden there are less consequences if the space is carpark or as in the space you describe a single temporary exhibition use. It also simplifies maintenance when and if things do go wrong.

  23. Gordon Evans

    Christine: The extension of the Hamburg Trade Fair was politically an extremely hot potato because it expanded into Hamburg’s premier public park, Planten un Blomen. We felled 140 trees, many of them over 100 years old and demolished Karl Plomin’s terraces, built for the 1963 International Garden Exhibition. The location and design of the building were therefore led by landscape considerations and an immense pressure from the Hamburg parks authority, we were required to restore the parkland in full for public access. There was a very absorbing series of dicussions and models with the architects and structural engineers where we sacrificed roof loading for the sake of thinner and more elegant trusses. We had to continually test the market during this stage to ensure that the substrate weights we were proposing were actually achievable, causing two of our five substrate tenderers to drop out of the competition. The final structure was so sensitive that we had to load the roof prior to the interior fitting out as it was calculated to sink 7 cm with the external loading. With its 8,000 square metres, I believe that this is the largest herbaceous garden on a roof anywhere. The planting plans were done by my old friend Brita von Schoenaich (, my office (since closed) did everything on the roof from the waterproofing up, including the insulation, and the whole lot was built by the landscape contractor Osbahr GmbH (

  24. Tom Turner Post author

    A fascinating design account – and particularly interesting that the pressure came from the Hamburg Parks Department. UK parks departments are spineless and run by leisure services committees with more interest in sport than leisure or service or amenity. So if a similar outcome were to be achieved in the UK the pressure would have come from the British public, who see the preservation of POS as roughly equivalent to preserving Magna Carta. I wish people had an equal interest in the creation of new public open spaces.

  25. Christine

    Thankyou Gordon for your account of the constraints and unique processes surrounding the procurement of the Hamburg Trade Fair extension. It certainly provides invaluable insight into what is likely to be a very challenging area of endeavour for some time into the future as all the difficulties of designing and implementing innovative projects are ironed out.

    It is also a uniquely interesting account for highlighting the difficulties which still exist as the various disciplines attempt to collaborate on projects while maintaining their ‘traditional’ interests and concerns.

    The balls that need to be kept in play as these projects proceed are only increasing. The solutions which need to be found are often not this OR that, but rather BOTH this AND that. (The goal should be to achieve the project completion without prejudice to anyone’s business through extraordinary delays etc.) Perhaps it would be worth advocating that high risk experimental projects of this type should have some sort of government insurance scheme to protect those who are unduly effected during the procurement process?

    A term used by Horst Rittel to describe problems of this type is a ‘wicked problem’: meaning a problem characterised by multilayering, complex social contexts and a demand for an integrated approach.

    This suggests to me that there are considerable new challenges evolving in the sustainability environment also for project management professions.

  26. PG

    I think we have to ask the question , are cities relevant or necessary today , and also to take into account all the forms of pollution , and that should include raw material consumption . With modern means of communication does everybody need to be in the same place at the same time , would it not be better to modify the current industrial or organisational models .
    Government publicity shows Singapore to be a grean city . Apart from the fact that no city is ecological , Singapore is no example to follow
    Singapore has’nt had an original idea in its 40 years . They should have been using the sun the generate energy and hot water for years . Building orientation is an important factor , and the lack of use of double glazing and good insulation is flagrant in Singapore , along with the excessive use of Air conditioning and lack of double doors to stop the cooled air from escaping.
    Green also means manufacturing and using the minimum of natural resources possible , that includes energy .
    Singapore along with Malaysia has a very bad habit in building , keeping it for a few years , then knocking down and rebuilding , instead of renovation . The government likes the current policy as it gives them material for land recalamation , which again is not ecological , having destroyed large areas of coral , and modifying sea currents . No where is the cost and pollution taken into account for producing large amounts of concrete and gravel etc , and even more running costs .
    And going into todays events , should’nt the governments at the Copenhagen meeting be talking about finding money to do research into new techniques and ideas to help the planet , and reduce consumption and maybe population . Or is this not inline with current financial policy , it does not increase GDP !
    There have to be big new ideas for the future , not just the old ideas repainted .

  27. Tom Turner Post author

    Thank you for an interesting comment on Singapore. The word ‘green’ has two meanings in this context (1) vegetated (2) environmentally responsible – as in ‘green’ politics.
    Could it be that Singapore can claim to be green in the sense of ‘vegetated’ – without even claiming to being ‘green’ in the political/environmental sense?
    I think this is often the case with intensively managed green walls. They look beautiful but they consume resources: electricity, peat composts, maintenance visits, water etc.
    Re the need for research, the answer is a very big YES. It is perfectly possible that if research funding was poured in the problem could be solved. The historic examples of this approach are the development of the Atom Bomb and the Appollo Programme. These were primarily US research programmes. Just think what might be done with a worldwide research programme looking at every aspect of green energy, green architecture, green transport and, of course, GREEN LANDSCAPE ARCHITECTURE.

  28. PG

    I have often thought about cities and their necessity today . In history , cities were fortfied and protected areas with everything all together because of transport and communication problems. This is not the case today , transport and communications are good , and we have to transport everything to cities , either for production or use . This creates road jams and pollution and is far from ecological due to large areas of concrete and tarmack or even efficient .
    Look at continental European countries , a lot of people try and have jobs where they can work from home ( less pollution , less lost time , cheaper , better quality of life), they don’t need or want city life . Also large companies ( banks , finance government ) are all in cities , do they really need to be ?
    I think we have to have new ideas about countries and cities , finance and industry , but how to get governments and populations to react ?
    One thing Asian populations have forgotten today is quality of life , the meaning of ecology , leaving the planet in a better state than when we started .

  29. Tom Turner Post author

    My (limited) experience of Asian cities confirms your view. They are needlessly terrible and produce an often-grim quality of life.
    You are right though, that much could be saved if more people worked from home. South Korea has one of the best broadband networks in the world and people COULD work from home if they were trusted to do so.
    Another interesting point is that in periods of stress modern cities seem to be able to adapt and survive with far lower levels of resource input, as in the Gaza strip during the past two years.

  30. PG

    Does not matter what planners and architects do or say , cities will never ever be ecological , and the people who live their even less.
    Planners architects,politians and civil servants have run out of ideas and creativity , they have become city dwellers with city dwellers minds , unless something is done to change ideas and ways of thinking , everything will get worse , just like the financial and banking systems


Leave a Reply

Your email address will not be published. Required fields are marked *