When I was a kid growing up in the middle of North Carolina, my grandmother would always encourage me to get out into the field and do my work early in the morning before the sun was too high in the sky and it was too hot to get any work done. Like most kids, I wouldn’t listen and would spend most of the morning goofing off. When I finally got around to hoeing the corn or picking beans, it would be the middle of the afternoon and the heat and humidity would be absolutely oppressive. Your sweat-soaked clothes would cling to your skin and no amount of water would ever make you feel like it was enough or that you would ever feel what it felt to not be thirsty again.
And to think, it’s only going to get hotter. A recent study published in the journal Science gives us a detailed portrait with numbers.
And it’s not good.
The study, “Estimating economic damage from climate change in the United States,” comes from the research group the Climate Impact Lab (cited as Hsiang et al. 2017), a collaboration of more than 20 climate scientists, economists, computational experts, researchers, analysts, and students from several institutions. The publication has quickly made the rounds given its stark findings that the impacts of climate change–felt through effects on the economy, crime, and food production–are going to be incredibly, disparately distributed across the US. With the South facing the most dire outcomes.
The major findings from this research are that the US will lose 1.2% of its GDP for every increase of 1 degree Celsius in surface temperature, with a huge burden being placed on the poorest third of the counties in the US—particularly those in the South. Researchers project a huge transfer of wealth from the southern US towards the north and the west, resulting greater increases in income equality.
The research group use a composition modeling approach with a county by county analysis of how projected changes in surface temperatures, precipitation, and carbon dioxide concentrations in the atmosphere will affect agriculture, human mortality, coastal storms, crime, labor and energy demand.
It is a complex, but fairly elegant approach where they are combining a bunch of separate models together in a composite manner and looking at the effects of climate on a county by county basis. Prior models that have attempted to forecast the effects of climate have done so globally or for the entire country. What is lost in these large-scale attempts, is the differential effects of climate. The real novelty here is this spatially explicit approach. Most of the coverage on this work has been on those big numbers about the decrease in GDP, but let’s look under the hood a little and look closely at some of the results.
For projections of how climate impacts agriculture, the group look at maize (corn), wheat, soybeans, and cotton specifically. The underlying methodology here is that for a commodity such as a maize, that there is a slight increase in production with increasing temperature till a certain point or threshold, then a huge crash. Some plants like it warm, but at a certain point hot is too hot, and plants die. Some plants like maize like it really, really hot, and some plants like wheat, not so much.
Precipitation is also not a linear, 1 to 1 relationship, but is a bit more complex. Plants like the right amount of water—kind of a Goldilocks thing. Too little, and the plant dries out, too much, and the roots can’t get enough oxygen and die, followed soon by the whole plant.
But what about increasing carbon dioxide in the atmosphere? That’s the life-blood of the plant, right?
Well, it is complicated. Sure plants like carbon dioxide. They take it from the atmosphere and turn it into fuel in the form of sugars and then turn that into more plant. However, to make use of that extra carbon dioxide the plant needs more of other stuff as well—water and nutrients. Without increases in those resources, we actually see a reduction in photosynthesis in plants as they have to shut down to avoid losing too much water from their leaves in a high carbon dioxide world. Think of it as they want the carbon, but just can’t get to it without enough water and nutrients. In some plants, particularly wheat, there is evidence that the nutritional value of the food is compromised when carbon dioxide levels are higher.
If we examine the map closely, we see major negative impacts here to Nebraska, Missouri, Illinois, Iowa, and Arkansas. There are concurrent gains in Oregon and Montana—presumably from increases in wheat production. While there is a geographic limitation here on where the negative impacts are, it should be noted that those red areas represent a significant amount of agricultural production relative to the country as a whole.
Some caveats. A lot of the changes we are seeing in climate are non-linear and are very dynamic. What I specifically mean here as that we are seeing instances of places where there are increased drought frequency and severity while seeing concurrent increases or no change in total precipitation. Rain may still be falling, but it we are seeing that rain fall less frequently and when it does fall, it is more severe.
Quick Facts from the USDA
– Agriculture and food related industries contributed $992 billion dollars to the US GDP in 2015
– 21 million full- and part-time jobs were related to agriculture—11.1% of the US workforce in 2015
Change in mortality in deaths per 100,000 people. From Hsiang et al. 2017
And while this type of work is almost always interpreted politically, the authors do a good job of threading the needle. But there is an underlying issue about environmental justice that is inherent in how climate will impact the most vulnerable.
“Because losses are largest in regions that are already poorer on average, climate change tends to increase preexisting inequality in the United States.” –
– Hsiang et al. 2017
If we look closely at the units on this map, they are in change in deaths per 100,000 people, with the darkest red being between 60 – 80 more deaths a year for every 100,000 people in the population. In 2015, Georgia had 10.2 million people. Based on these models, by 2080 about 8,000 more people will day per year solely due to climate change in the state of Georgia.
However, here I don’t think this map tells the whole story. Research into heat-induced mortality in urban areas show that not all cities are respond in kind (1, 2, 3). Cities are spatially variable across the country and even within their own limits. Hot cities like Atlanta or Phoenix are a bit more resistant than cities like Boston or Seattle to acute heat, like a heatwave. Plainly, what is a heat wave in Seattle maybe a nice, balmy day in Phoenix. This is borne-out in mortality data as well where mortality increases following a heatwave, but the temperature at which things can be dire differs by cities—lower temperature thresholds for the north and higher for the south.
There are large centers of population in the northeast and this map does not necessarily show the health burden that is in store for those areas.
I am working in northern Michigan for the summer where it got up to 86 F yesterday and it was pretty hot for here. But had I been in Virginia, where I live most of the year, I would be begging for 86 F. The temperatures there for July have been at or near 100 F with humidity over 90% some days.
Inherent in any work that attempts to make a forecast far into the future, are the underlying probabilities. That’s what this is paper really is. A probabilistic model. All of the outcomes presented are the median outcomes by county based on running the model thousands of times. That means that half the time things are worse than what is presented, but half the time not quite as bad. But it does not mean that there is equal risk on either side of the median either. The model here and all of its working parts are based on the best knowledge we have as of June, 2017 about how all the little pieces here work.
There is always a statement about how we are not tied to this course, that we can change things. However, real solutions are going to have to come from high-up the food chain. But localities are already stepping up to the plate. The city of Miami, facing increased pressure from sea-level rise that is already leading to frequent flooding at high-tide, is taking steps to increase the city’s resilience to climate change. Many hope though that we can work together in a unified response, and not have to respond patch-work, only where there is political and social will even when the knocking at the door is getting louder, and louder.
*Full caption: Impacts are changes relative to counterfactual “no additional climate change” trajectories. Color indicates magnitude of impact in median projection; outline color indicates level of agreement across projections (thin white outline, inner 66% of projections disagree in sign; no outline, ≥83% of projections agree in sign; black outline, ≥95% agree in sign; thick white outline, state borders; maps without outlines shown in fig. S2). Negative damages indicate economic gains. – From Hsiang et al. 2017