A place for experimentation, detached from the real world -- this is the sort of image that the word 'laboratory' conjures. It has been said that we are engaged in a planetary-scale experiment through our greenhouse-gas emissions, even if this experiment is unplanned and undirected. As we are of course very far from understanding the mechanisms of global climate in anything approximating completeness, in some sense any change we cause is an experiment; it's just that the more intense the experiment, the harder it is to pull the 'stop' lever. This notion of undirected experimentation is frequently invoked in the context of discussions about geoengineering, as a rebuttal to the argument that geoengineering is dangerously artificial. So this line of reasoning goes, experimentation is happening anyway.
As I've written about in many different facets here before, whatever is going on is certainly going on to a greater degree in urban areas. After all, aside from emissions resulting from land-use change (e.g. deforestation), most causes and effects of anthropogenic climate change are concentrated there -- a direct result of population density. And the differences are not trivial: the magnitude of the urban heat island of a typical large city is on the order of 2-4 C, in comparison to about 1 C of global-average temperature change since 1900. In fast-urbanizing areas like China, large cities are not only warmest, but are warming fastest; as a global average, though, cities have not materially affected the existence or measurement of large-scale warming. Taking all this into consideration, we can say that cities are ahead of the curve in terms of previewing the effects of a more intensely developed world, which can be conceived of as an "urban dome" environment. Within this experimental 'dome' -- distinct from an urban heat island as it encompasses variables other than temperature -- conditions can differ in many respects from those outside it. For instance, in a case study in Baltimore, city-center CO2 was about 100 ppm (or 20%) higher than in nearby rural areas. Precipitation is generally enhanced over and immediately downwind of cities, a result of atmospheric instability caused by urban heating. Compared to temperature and precipitation effects, air pollution is even more tightly concentrated, as shown in maps like this, wherein major urban (as well as industrial) areas light up among a low-level rural background. In fact, documented cases of concern over air pollution (primarily from burning wood, and, later, coal) go back to ancient times, and extend through medieval ones, all well before widespread industrialization. John Evelyn wrote, "One day, as I walking in Your Majesty's palace... a presumptuous smoke did so invade the court... that men could hardly discern one another for the cloud"; he continued, "And what is all this, but that hellish and dismall cloud of sea-coal... a fuliginous and filthy vapor" which (couched in terms engineered to tug on regal 17th-century heartstrings) "endanger[s] the health of the king... and sullies the glory of this imperial seat."
One experiment begets another: Buckminster Fuller's proposed dome over midtown Manhattan, intended to protect the citizens of those neighborhoods from the environmental consequences of the activity necessary to support their energy-intensive existence; their neighbors outside the dome are presumed to have to fend for themselves. Source: The Gothamist, "The 1960 Plan to Put a Dome Over Midtown Manhattan."
Another way that cities are a natural experiment is with respect to ancillary changes in the rhythms of life. It's not just people waking up at 5am for an hour-and-a-half commute that's unnatural; even vegetation experiences discernable and significant changes in phenology. Eli Melaas from Harvard found differences of up to several weeks in leaf-out between street trees and urban woodlands in Boston -- evidence of "micro urban heat islands" -- plus another week or so between urban woodlands and rural woodlands. Temperature and streetlights combine to fool trees into acting as if they were several hundred miles equatorward of where they are. [This interchangeability between urban heating and large-scale temperature gradients will be explored more in a post in the near future.]
In many respects (and not coincidentally), the changes predicted worldwide in the coming decades and centuries are already manifest in cities. One example is warming that is strongest at night and in winter. This means a much longer growing season and a poleward expansion of arable areas. Another robust prediction of climate models is a global intensification of precipitation where it occurs (the old "wet get wetter, dry get drier" mnemonic) that is also already borne out by urban precipitation records, as briefly mentioned above. On the demand side, it would be reasonable to expect a continued decreasing of economic intensity of emissions as urbanization and energy-sector efficiencies progress, as has been already observed both over time for a given region as it develops (left interactive image below), in step with the close relationship between development level and emissions intensity (right image below).
Of course, having a preview of possible conditions that generally confirms our model predictions, while helpful, doesn't constitute a prescience that one should place all their bets on. For one, having a warm city immersed in an unchanging countryside is qualitatively different than having an entire region that's warming, with an even hotter city within it. And the future form of cities is not necessarily essentially the same as today's -- to give one example, driverless cars could enable drastic reductions in paved surfaces, and thus increases in urban albedo, which together with green roofs could alter in large part the city-countryside contrast that currently does so much to drive urban heat islands. Even with perfect knowledge of the inner workings of the climate system, throwing in human behavior makes for a meta-system far too complex and dynamic to confidently predict its direction with anything more than broad strokes. But having urban areas as laboratories at least enables an inkling of what to expect if the remainder of the 21st century, and the 22nd, proceed as the 20th and the 21st so far have done. The lessons we can draw from this experience is that 'business as usual' is climatologically starting to take us down the path of the unusual.