Toronto requires green roofs for new building projects. Grasses and shrubs are typically grown, and they survive just fine with minimal care. I suspect the only reason that large trees are not typically used is that there's fear that tree roots will damage the roof of the building. Oh, and the weight of the soil required for larger plants is an issue.
Yep, looking at the bylaw, the growing medium - soil - is only required to be four inches deep. Weight of soil and drainage starts to be a problem - you have to figure that the soil may be 100% soaked...
It's going to be tough to grow trees in four inches of soil.
So, nutshell, tree survivability not a problem, but engineering a roof to hold enough soil (and therefore water) to grow a large tree is expensive, and root damage is a problem, and therefore - no large trees on skyscrapers. Still, there's nothing magic about it, just engineering problems. I could easily imagine a high-end residential tower with a forest on the roof.
Just to add some more information to your comment, I'm an Architect and I worked on a building with trees growing on top of a structure, but not at high level. I just happen to know that, according to the tree specialist, a 5-8m tall magnolia tree can probably survive in a cube of soil 1.5m x 1.5m x 1.5m encased in a concrete pit lined with a geotextile to prevent root damage to the structure. It needs to be hooked up to a fairly sophisticated hydroponic system to keep it alive and their growth will be stunted by the restrictions placed on the root ball. Another issue with some of the trees shown is that, because they are under the structure, they will need to be turned every couple of years and/or illuminated with special lights to keep them from being distorted by growing towards the sun and to ensure they receive enough light to photosynthesise.
More generally, the computer renderings you see in the article are often generated by students or outside design agencies based on sketches and other early information from the architects. They are done in a few weeks often to very tight deadlines and are definitely not based on technical drawings so you can get all kinds of weird stuff in there depending on how imaginative the computer graphics person is and how late they had to stay up to beat the deadline. A real skyscraper takes years to design, the details of how tree pits work wont be fully bottomed out until a few years after the fancy renderings are done. Ultimately, trees often disappear from the finished product once it has permission from the city as the developer will be looking for ways to cut costs; things like this often get the chop unless there are specific conditions to keep them in the design that have been placed on the permission granted by the local authority.
My first development job was at an Architecture studio funded by the LA Community College District. The Architecture interns there were allowed a lot of artistic license as they used the renderings in their portfolios.
Usually the renderings mimic "magic hour"[1] lighting conditions and are taken from the best vantage point. Being a talented bunch, sometimes their embellishments ended up in the final work and looked great... and somethings not. Sometimes their renderings ignored the actual position of the sun in order to make it look better at the vantage point the picture was "taken", since the CAD software would add lens flare and glare. Sometimes a rendering has one time of plant or another simply due to the whims of the intern or because of what was available in the asset library (i.e. Coconut or Palm Trees instead of Magnolia tress).
Can you provide any reference/elaboration for your claim that tree survivability is not a problem? It's directly counter to the arguments made in the article. Grasses and shrubs are a lot more hardy than trees, as anybody who's even been on a mountain can attest to.
Your other points are well taken and they are sort of addressed in the article in the form of the infrastructure issues. I think the author also would agree that it's sort of feasible in theory, but extremely expensive and impractical.
But you need a meter of soil, two meters would be better. This is 10-20x as much soil as is currently available on most green roofs (or mountainsides).
There are already pictures in this thread of buildings with trees on top, so... the idea that somehow trees are subject to instant death when lifted off the ground is silly. The conditions are not any harsher than trees living in an asphalt jungle in the city are subjected to every day.
It called my attention that in South Finland there are pine trees which apparently are able to grow with their roots barely covered with soil. Some sit directly on hard rock with all their roots visible. I took two photos of this on 2006:
http://pix.hamoid.com/finland2006/nature_is_a_huge_canvas.jp...http://pix.hamoid.com/finland2006/going_back_to_my_roots.jpg
Those trees survive winter winds at -22F (-30C) with very little sunlight, although sometimes winds knock them down making a vertical wall out of the exposed roots.
You are right, if one sees roots it doesn't mean there aren't other roots.
I think the difference is, South Finland was compressed by 1 km of ice during the last ice age, and the ground is pure granite that is blown up using explosives whenever a new building or road is built.
In those places I talk about, the dirt layer is extremely thin or often nonexistent. The photos I linked don't show that, but many trees are clearly sitting on pure red granite rock with just a little bit of dirt and moss surrounding the roots, in the middle of pale rock.
Yes, and the article directly addresses this: Still, plants in these environments aren’t usually tall and graceful. In other words, not the tall trees we see in architectural drawings.
Though I'm personally very partial to the gnarly mountain trees, too.
> Can you provide any reference/elaboration for your claim that tree survivability is not a problem?
Mountane ecology start a 1500ft at the very lowest (e.g. in Scotland), and a characteristic of them is that they're generally forested. Elevation is not an issue. Wind and exposure mostly mean that no all species will handle it (duh) and you will have the usual loss as not all saplings will survive, but trees do grow, even in windy environments with no groves to buffer them.
Conifers are extremely hardy trees and will grow without issue. Hell, a mere 500ft might be too low and warm for them.
I would go out on a limb here and say that the weight of the soil and drainage issues are not relevant. We're talking 100 pcf or so for saturated organic soil and drainage is pretty easily taken care of with geotextiles. 300 psf isn't overly large for a dead load when you note that live loads should not be applied over the same area as it isn't a space that should have traffic during a significant load combination.
Your second point about root damage has more merit but still doesn't seem like a hard problem.
Even if the dead load isn't a problem, the space taken up by a one- to two-meter-deep soil level is going to be significant. (Oh, you were planning on occupying the floor under the tree? Tough.) Also, even if draining the water isn't a problem, pumping all that water up there in the first place doesn't sound cheap.
The comment you link doesn't even remotely state it's impossible though. Mostly, that you won't have the same tree up there than you natively have on the ground. Which is kind-of obvious really, especially if "on the ground" is flat lowlands.
It goes further than that: air near the surface of the earth is doing a lot more interacting with the earth than the other air, and is getting and keeping a lot more heat and dissolved gases. That includes air at sea-level or air at a mountaintop. Thus, there is no natural analogy to a tree growing atop a 500 foot building. Even bristlecones at 10000 feet still benefit from air that has been in prolonged proximity to a surface.
Yep, looking at the bylaw, the growing medium - soil - is only required to be four inches deep. Weight of soil and drainage starts to be a problem - you have to figure that the soil may be 100% soaked...
It's going to be tough to grow trees in four inches of soil.
So, nutshell, tree survivability not a problem, but engineering a roof to hold enough soil (and therefore water) to grow a large tree is expensive, and root damage is a problem, and therefore - no large trees on skyscrapers. Still, there's nothing magic about it, just engineering problems. I could easily imagine a high-end residential tower with a forest on the roof.