There are around 250 tall towers currently planned for London's skyline. From afar, they'll probably look great. But unless they're planned carefully, they'll start throwing shadows across ever-growing swathes of London.
So one local architecture firm is offering a solution. Using computer modelling, they've figured out a design which would reduce the shadows cast by two theoretical towers by as much as 60 per cent.
The architects, from NBBJ, used computer modelling to position two curved towers in such a way that, when one blocks the sun from the public plaza below, the other reflects light back down.
Oh, and before you ask, this wouldn't be a Walke-Talkie death ray situation: the towers' shape would ensure that the sunlight doesn't focus on a single point and fry anything.
We spoke to the towers' design director, Christian Coop, who says the idea grew out of a recognition that cities will need greater density in coming years – but that it'd be great if they didn't have to sacrifice daylight at ground level to achieve it. "The key to getting tall buildings right is at the skyline level and the base," he says. The design ensures that the area between the towers is bright and pleasant, so is more likely to be used as a public space:
These diagrams show roughly how it would work:
Factoring cosmological factors (i.e. the sun) into building design is relatively new, and so to drive the point home, NBBJ placed their concept designs in Greenwich, near the Prime Meridian. Coop says he feels the "cosmological dimension" should be a routine consideration for architects, just as sustainability or aesthetics are.
So far the concept is still just, well, a concept, but Coop hopes that architects will use this type of modelling to see how their designs will affect the light levels around their buildings. He says the method could also be adapted to control how much sunlight is reflected at different times of year: in cities like Dubai, the design could allow less sunlight to be reflected in the hot summer months, and more in winter.