Overall it looks to be an interesting project, but the work is not far enough along to be published yet.  These guys need to spend some time at the Columbia U Spatial Design Lab and learn how to make an effective map.  Just because some software will let you do something doesn’t mean that you should. 

Many comments have pointed out the correlation of energy use with building height.  They already have this information but they are not using it. It is in the same file they got the taxlot sizes from, which includes building square footage, number of floors, building class (around 200 types), building age and renovation year, and for large buildings–a further breakdown of square footage by type of use.

It’s tricky to transform the zipcode level data to taxlot resolution.  You would need to use some real world data from individual buildings to develop the model.  Maybe when the methodology is published their map will make more sense.

A measure per square metre of usable floor space might be better.

Not sure why this would be useful, since it would make an 8000-square-foot McMansion or office space appear just as green as an 600-square-foot New York apartment, because they might both use the same number of watts per sq-ft.

More floor space = more energy, but this is a bad thing, and not something that should be removed from the equation by dividing by it.

The number of people living in the space is the key metric. If four adults can live in NY and use 200 KWH/month, then they are far greener than 2 adults living in a huge suburban house using 1500 KWH/month.

The building where I live, for example, is part of a complex of four identical buildings (that I’m fairly confident use very similar amounts of energy) with two buildings on each of two blocks.   Yet one block shows up much darker than the other, presumably because it’s smaller. 

My guess is that if you recalculated this by built square footage instead of raw land area, many areas of the image would be reversed, since a typical modern 50 story office tower makes much more efficient use of light, heat, etc. per square foot than a 6 story 100-year-old apartment building.

of course power utilization is going to become obscene when you’re measuring power density by lot size in a major metro, instead of by floor space.

quickly calculating a few dark reds they’re at ~500kwhr per m^2 per year
with massive amounts of servers and at a gentle 72F year round

my modest house operates at ~230kwhr per m^2 per year with a thermostat at an average 58F during the winter and using AC during the 2 hot weeks of summer.

calculating it by lot size i use 13kwhr per m^2 per year…..

their Galaxies are spread out between NYC, NJ, Conn and PA,  just to start locally to NYC

this is why you never heard a single word about wallstreet being unable to function after 9/11.

ontop of that the exchanges operate with a built in latency in stock tick. this was developed to “level” the playing field so that no trading house had an advantage by placing their trading cluster right next door to the NYSE. this built in latency is what allows them to all extend their clusters into galaxies for load balancing and DR in NJ PA & Conn

Can you grow coca indoors?

Pretty sure they just import the cocaine.

There’s certainly a lot of IT equipment and data-center-ish type installations in downtown NYC, but not as much of the trading-floor type as one might assume. 

Then again, it’s Wall St.  Can you grow coca indoors?