A large part of applied urban climatology consists of identifying climate-related hazards for existing urban areas, and devising solutions that either mitigate the problem (“mitigation”) or increase the community’s ability to cope with it (“adaptation”). Rarely, however, do we take the evaluation back to its natural starting point — that is, the location and form of urban areas in the first place. Perhaps this is because the major aspects of cities like size and location are largely outside the control of any one generation of policymakers, so that they are essentially unchangeable, especially on the approximately decadal timescales that are relevant for most climate planning. This unchanging character means a substantial amount of lock-in of future emissions due to the development patterns that exist today (that is, future emissions even if all future development was planned to be perfectly carbon-neutral). This concept could perhaps be extended to include locked-in societal risks as well. Unfortunately, the price tag and level of political wherewithal necessary to meaningfully change these patterns are all but unachievable. In certain ways, this is a positive: strong-arming people into relocating to places of the government’s choosing, even if those places are objectively identified, does not seem palatable. What is more feasible is preventing or disincentivizing additional high-risk development, rather than attempting a post hoc ‘cure’ for poorly planned cities. As with most prevention/cure dichotomies, prevention is much easier — and given that the costs of these decisions are ultimately manifest on a societal level when they balloon to a certain size, these choices should be of interest to all.
There are several levels of bad decisions about choosing the location for a city or other new development. There are the harmless mistakes (siting Lima in the summer, not knowing about the constantly cloudy winters), the constrictions that become apparent only as a city grows to appreciable size (water availability in Cape Town), and the constrictions that should have been apparent from the very start (the smidgen of naturally high ground in New Orleans). Of course, the recent troubles with the above-mentioned places could largely have been avoided with better long-term planning, and responsibility for the crises of Hurricane Katrina and the 2016-18 Cape Town water shortage should not rest solely on the shoulders of the cities’ founders. After all, in many such cases it took centuries for the problems to reach levels where they could cause a significant negative impact on daily life. Modern-day development that expands cities, however, cannot so easily claim ignorance. Paving large sections of the Texas Gulf Coast, or building homes in claustrophobic California canyons, is enabled by the absence of open-market forces (e.g. lower-than-fair insurance premiums) in combination with a propensity to underestimate the associated risks. Harveys and wildfires, in other words, were not sufficiently considered or prepared for. Some meetings I attended last fall focused on the rapid population growth of the ‘wildland-urban interface’ and the tragedy-of-the-commons hazards and public-services demands that it raises. And that’s in the comparatively data- and resource-rich United States. Whether in the hills of Nepal or the floodplains of Nigeria, urban sprawl still proceeds largely unchecked in most of the developing world, with little consideration for or knowledge of the eventual societal-level costs when these homes are damaged or destroyed (and many of them are highly vulnerable, even in the absence of anthropogenic climate change). In these places, the residents are left destitute if climate-related disaster strikes, even if the strike is fairly routine or predictable; in the developed world, residents are financially saved but typically only by the failsafe backstop of the government, and thus any kind of development with a non-negligible chance of falling into this category is really a public concern. As with many of the most interesting aspects of urban climatology, this is where science and policy blur together indistinguishably.
On the policy side, the principal question revolves around how to appropriately incentivize development that's aware of the risks and vulnerabilities imposed by its location. A commonly voiced but as-yet-undone step in this direction would be to reform government-run mortgage and insurance programs to increase the financial penalty for developing areas that are known to be high-risk. Multiple overlapping evaluations of potential environmental consequences at scales from individual homes to metropolitan areas would also be wise, particularly in cases where problems may only emerge for certain types or densities of development. These could consider multiple development options at the neighborhood scale in combination with a simple climate simulation to address questions like, for instance, whether the overall risks of flooding along a beachfront are smaller with several apartment towers or a large number of homes. Detailed climate information could be used for analysis if local high-resolution simulations are unavailable or prohibitive. Of course, historically and straight through to the present, it has often been the case that the residents of high-risk areas were forced there by circumstance and would leave if they could; this represents another necessary dimension of any comprehensive policy solution. In other words, smart development requires a true proactive dialogue among actors before major urban-planning (development, zoning, etc) decisions are made. While urban-climate tools are rapidly gaining in their usefulness for analyzing the effects of past development, it would be even more gratifying to use them strategically to help guide these kinds of choices in the first place.
How important is urban climatology, really? The question seems important as the field grows in size and scope, drawing in more and more resources. While certain statistics about the urbanization of the world are now widely known (e.g. that more than half the global population lives in cities, or that megacities are growing much faster than small cities or rural areas), there is also a danger that smaller municipalities will be outshone figuratively as well as literally by the bright lights of big cities, and that other worthy climate impacts will be overlooked in the process.
A post last year from Marshall Shepherd commented on one demonstrated aspect of “urban bias” in the context of forecast accuracy; specifically, he noted that weather reports in the media tend to focus on urban areas to the exclusion of everywhere else, and that this focus is disproportionate even to their larger populations. I would argue that to a considerable extent this is a consequence of media navel-gazing, a sort of availability bias. After all, urban bias makes an appearance in other fields as well — how many TV shows are set in New York City, home to just 2.6% of the country’s population?
Of course, this blog being titled what it is, it’s worth noting why cities are the object of such intense newfound attention. Much of it is well-deserved, a rightful recognition of the significant concentrations of people and assets that urban areas contain; the ten largest metropolitan areas in the US produce 35% of the GDP despite having only 26% of the population. There are also an enormous variety of interesting interactions on a variety of scales, which are just beginning to be able to be studied systematically (i.e. climatologically) thanks to advances in computing power. Cities are also often positioned on rivers, along coastlines, or in valleys, adding an extra topographic level of complexity to their microclimates. As a result of all of this, climate in a large and diverse metropolitan area like New York can span a wide range of temperatures, not to mention precipitation, etc. As shown in the figure below, the annual-average temperature at LaGuardia Airport is equivalent to that in central Missouri, while the temperature in the southern Catskills is like that of Minneapolis, 10 deg F colder and 400 miles north.
All fields of science seem to exhibit a tendency to stampede from topic to topic, leaving topics that are not fashionable sitting dust-covered on the side of the trail. This is unavoidable — part of human nature — but good to be aware of so that the most can be made of it in terms of building up understanding and impacts applications. For example, organizations like CCRUN have focused largely on the major cities of the Northeast, due to their having the dedicated funds and personnel for climate issues. At meetings I’ve attended there have been discussions around the possibility of working with small and medium-size cities, though here (as elsewhere in the country and indeed the world) such work is hindered by the inherent difficulties of scale, the greater number of municipalities and their smaller discretionary funds being chief among them. While 55% of people currently live in cities, only 23% live in cities of at least 1 million. In fact, in 2030 rural areas will still account for 40% of the global population. An additional positive outcome from the push to improve urban climatology would be a conscious downward movement of resources and methods to these smaller localities. Whether by direct focus or extrapolation from results for larger places, small cities, towns, and villages deserve to not be forgotten even in an urbanizing world.
While the drought in the Cape Town area has been ongoing for several years, it wasn't until earlier this year that it became a global news story, in conjunction with predictions that the city's pipes would run dry around the middle of this year. While it's now thought that this dramatic situation will be (barely) avoided, at least in 2018, the mere specter of a moderately-wealthy city that shows up on every globe bumping up against a hard resource limit has jumpstarted many conversations about the seriousness with which the management of essential natural resources must be taken. In the 21st century, the developed world has generally speaking distanced itself far enough from needing to constantly assure the supply of critical requirements for life that it's jolting to be reminded that their constancy is the product of human engineering and innovation, rather than any immutable law of nature. Just as we now worry little about cholera in the water, we are extremely confident that there will be water in the first place, regardless of the fluctuations we see going on around us. In Cape Town's predicament -- a consequence of high natural variability in rainfall, warmer temperatures that evaporate more water, and the lack of a serious backup plan such as desalination or larger reservoirs -- is a vivid reminder that sometimes you get dealt multiple poor hands in a row. The only upside of such a "Day Zero" is the action that it incentivizes, and the rethinking of previously unquestioned norms and assumptions.
Total water volume stored in the reservoirs that supply Cape Town, showing the seasonal and interannual variations as well as the steady drop from the wet year of 2014 to the dry years of 2017 and 2018. The slower drop in water levels so far this year is due to restrictions placed on both urban and agricultural water users. Source: "Water Outlook 2018" report at https://coct.co/water-dashboard/
The Cape Town situation is a complex stew of factors, ranging from inequality between the poor and the wealthy, between native Africans and white/Asian immigrant groups, and between agricultural and urban water users to poor planning, population growth, and sheer bad luck. It took many or all of these for the current situation to develop; after all, this is only one instance in multiple decades, and one city out of hundreds or thousands of its peers. However, in these overlapping, interacting, and sometimes self-reinforcing aspects it serves as a premonition of the kinds of tensions that could occur around the world in the future over shifting availability and constancy of resources. California, for instance, is expected to see stronger seasonal and interannual "whiplash" between wet and dry conditions, making an already-challenging water-management task that much harder. I would argue that climatic change is a bit like economic change in that way -- not in the sense that it's inherently good or bad, but that the most-consequential changes involve the way that they disproportionately benefit certain groups or places while harming others. To take a non-extreme example, comfortable (dry and mild) weather is expected to shift poleward over the coming century, with sharp decreases in such days over developing countries in the Americas, Africa, and South Asia contrasting with increases in Canada, northern Europe, and various sparsely populated highland regions. Despite the grand burden-sharing promises of e.g. the Paris climate accord, it remains a very open question as to whether there'll be any mechanism in place to prevent these benefits from flowing directly to the already wealthy and leaving the global poor in the lurch, suffering from a problem they didn't create in the first place.
From a decadal-climate-change kind of perspective, these localized Day Zeros may serve as a blessing in disguise if they force the underlying problems (which are global in nature) to be reckoned with before the impacts grow to truly uncontrollable sizes. As the map below shows, the Cape Town drought is remarkably concentrated in the city, its immediate suburbs, and the nearby mountains where its reservoirs are located; a drive of just an hour or two reaches unaffected areas. This makes the problem, while of course quite severe in its local impacts, qualitatively different than say large-scale water stress due to glacial melting and the loss of their summer-storage capabilities. One can imagine a slew of other Day Zeros, ranging from the first locations to hit unsurvivable wet-bulb temperatures to the advance of "warm-weather" insect pests into boreal forests. Due to their concentration of people, assets, transportation links, and media outlets, many of these impacts will be felt first in cities, or at least reported on there. This makes knowledge of when an urban area's Day Zero might occur, and how to prepare for the case that it does, of great importance for keeping global society within comfortable bounds that allow basic requirements to always be met -- as we've come to mundanely expect, without giving much thought as to what it takes to ensure this. Between still-growing resource demands, ever-present natural variability, and the shifts in both means and standard deviations due to anthropogenic climate effects, those kinds of unquestioning assumptions bear some significant revisiting.
Drought status as of Aug 2017 for municipalities in the Western Cape region. While Cape Town and its immediate neighbors were and continue to be in severe hydro-agricultural drought, other areas within 100 km of the city have secure water supplies. Source: https://www.westerncape.gov.za/text/2017/August/western_cape_drought_map_risk_and_declared.jpg
Two major features stand out with regards to recent population changes: the dramatic increase in the last 100 years, and the worldwide ongoing migration from rural to urban areas. Both of these demographic factors affect the ultimate climate impacts, just as much as any climate oscillation or long-term change itself would. In this post I'll focus on how current and projected interregional population shifts (due to migration and to natural population increase/decrease) amplify or curb human exposure to climatic hazards. A more-complete accounting of societal impacts would also consider non-population-dependent regional effects, such as those on agriculture, fisheries, or water resources.
A pioneering study in this branch of demography quantified the surprisingly large percentage of the world population living in low-elevation areas vulnerable to future sea-level rise. Their main findings: the majority of megacities are on coasts, and the median person lives at an elevation of only 194 m. Thus the economic effects of even a modest amount of sea-level rise would be considerable, in addition to the disruptive migrations and political headaches. Estimates for these regional economic impacts (taking into account all aspects of climate change) indicate that the spatial and temporal benefits and costs will continuously vary, creating a complex mosaic of 'winners' and 'losers' as climatic averages and probabilities of extreme events shift simultaneously. For example, a simple review piece shows that cool temperate areas will gain comfortable mild days, while much of the tropics and subtropics will see formerly pleasant dry-season days become hotter and drier. Maps of other major extreme events are shown in the below image gallery; comparing these to global population density, and to global projected population change, provides a better sense of where future impacts will be most negative or positive, and what combinations of impacts will conspire to affect a given region differently than in the past.
Above: Global maps of the climatological distribution of (clockwise from top left) tropical storms; extreme heat; dust storms; extreme cold (NH winter); thunderstorms; and wildfire.
A detailed analysis of the intersection of population and climate change is well beyond the scope of this blog post. Some simple observational remarks:
-Densely populated East and Southeast Asia are heavily exposed to tropical storms, extreme heat, and thunderstorms (leaving aside geological phenomena like earthquakes and volcanoes)
-Rapidly growing northern India is exposed to extreme heat, dust storms, and thunderstorms. Not only that, but fertility rates are highest in precisely the most climate-vulnerable areas
-While Eastern Europe and the US Great Plains are the loci of strong projected increases in extreme heat, the health effects (though not the agricultural ones) are mitigated by the population decreases in those areas
-Accelerated urban warming will further compound cumulative exposure to extreme heat, as already observed in Chinese cities, among other locations
Further study of specific regions and connections is warranted -- after all, it's the areas where the population is densest, or economic value highest, that are the most critical for supporting and protecting human society. The existence of many climate feedbacks of course strongly argues for the protection of sparsely populated areas like the Arctic and the Amazon basin, but to zeroth order, local problems have to be dealt with before remote ones. Avoiding a devastating heat wave or storm in India should, in my view, take precedence over preventing sea-ice loss on the grounds that it probably bears some connection with mid-latitude extreme events. Another way to frame this argument is that specific problems take precedence over non-specific ones (i.e. those whose exact form, location, and timing are not yet known).
On decadal timescales, population projections like those in the above map are wrapped up with economic ones in the Shared Socioeconomic Pathways dataset -- like the Representative Concentration Pathways projections for greenhouse gases, but for the more-nebulous human aspect of the future. This brings up another reason for periodic revisitation of studying the population-climate interface: it's always changing. The above map doesn't include (as it couldn't foresee) the exodus of people from Syria due to civil war, as reflected in this 2015 map of population change. Capital investment, too, is always moving around in search of profit; the economic impact of flooding in Thailand would have been much less than the actual $6 billion 20 years ago, before a surge in foreign investment and the country's incorporation into global supply chains. Anders Levermann makes the argument that adaptation to climate risk in this arena is much overdue.
Finally, a nice map of the US sums up at a glance current populations and future growth. Comparing it with the types of extreme events experienced in each region is instructive as to where and what kinds of events we should expect more of in the future. For the last decade Texas has added the most people in absolute terms, but is also highly exposed to many of the most damaging weather extremes. Florida is beginning to see sea-level rise affect its valuable coastal properties, in addition to the habitual challenge of tropical storms. Of the six categories of events below, all are expected to remain constant or increase in frequency or severity in the generally more energetic atmosphere of the future. The longstanding trends toward population and economic consolidation on one hand make adaptation easier (as we have to protect smaller areas), but on the other make it harder (as we have more to lose if a concentrated area is heavily affected). Globally, nationally, and locally, policies will have the power to shape climate-relevant decision-making to some extent, but realistically advocating for preparation for the future is the main tool we have as scientists. Climate remains low on people's list of concerns, meaning that its effects -- pure effects as well as those tangled up with societal issues, e.g. 'climate justice' -- mainly flow from decisions made for other reasons. But this doesn't make plotting and understanding climate-related problems any less worthy of scientific study or any less valuable for our collective future.
Billion-dollar weather events for each US state for the 1980-2016 period, by primary categorization. In this methodology, if a state was affected by a given event, it is included in the count, even if its portion of the total damage was much less than $1 billion. Source: https://www.popsci.com/natural-hazard-risk#page-2