The human cost of global warming: Reply to Lynas
Richard Parncutt 

October 2023
 Richard Parrncutt ICMPC 2012

Was keiner geglaubt haben wird / was keiner gewusst haben konnte / was keiner geahnt haben durfte / das wird dann wieder / das gewesen sein / was keiner gewollt haben wollte (Erich Fried, "Dann wieder")

What no one will admit to having believed / what no one believes they could have known / what no one will remember having suspected / that will again / have been that / which no one admits having wanted

Our emissions are killing future people. Whereas many now agree with that statement, there is a remarkable reluctance to look at the detail. How many people are we killing? What are the ethical, legal,  economic, and military implications? What about human security? I can't think of more important questions, which is why I am addressing them.

Mark Lynas, author of Our Final Warning: Six Degrees of Climate Emergency, recently reviewed my 2019 Frontiers article on the human cost of global warming, criticizing it in interesting ways that I would like to address. Lynas referred to an interview in which Roger Hallam claimed that, according to me, “a billion people are about to be killed”. In fact, I had claimed that the total death toll attributable to global warming -- starting when anthropogenic global warming first started, and lasting until well into the 22nd century (ignoring future millennia) -- would be roughly one billion, if warming is limited to 2°C (including temporary overshoot). If you divide the amount of fossil carbon burned since the start of industrialization to reach 2°C (about a trillion tonnes) by the billion deaths that burning that carbon will probably cause, you get the 1000-tonne rule: burning roughly 1000 tonnes of fossil carbon causes a future death.

The 1000-tonne rule can hardly be wrong when considered as an order-of-magnitude estimate. As I will explain below, the projected death toll of one billion at 2°C lies between reasonable best- and worst-case scenarios. Consider the projected population growth in Africa, from currently 1.4 billion to 4 billion in 2100. Agriculture cannot possibly keep up with that despite global warming. Food from other continents will not be forthcoming if other (perhaps all) countries are experiencing unprecedented difficulties feeding their own populations. Moreover, global warming will indirectly kill in other ways. Spread out over a century, a billion people could die in Africa alone, and from starvation alone, even if warming is limited to 2°C.

A death toll of that magnitude would be attributable to a combination of factors including
On that basis, a proportion of those billion deaths would be attributable to global warming in the sense that they would not have happened if there had been no global warming and other factors had remained the same. Combining those deaths with other deaths attributable to global warming, both on Africa and on other continents, we might reasonably expect a total global death toll from global warming (limited to 2°C) of one billion, spread across a century.

From another perspective, the World Food Program estimates that about 9m people are currently dying of starvation each year, although much more food is produced globally than can be eaten. The problem could be solved by a combination of improving food production, reducing waste, and improving distribution. If global temperature increases by 3.7°C by 2100, global food production will fall by about 1/3 due to changing temperatures and rainfall patterns (link). In a conservative estimate, that will mean premature death for 1/10 of future global population, or one billion people. But the starvation death toll will be considerably higher if the international food trade is disrupted by conflict, and we also expect a similar number of deaths from extreme heat. If warming is limited to 2°C, these death tolls will be greatly reduced, but the total of all death tolls in connection with global warming will still approach one billion, or 10m/year for a century.

Causes of premature death attributable to global warming

The number "one billion" is an estimate of the total long-term death toll attributable to global warming of 2°C. I arrived at this estimate by a method that could be called semi-quantitative triangulation. By "triangulation", I mean looking at the same problem from different perspectives and honing in on a possible solution -- similar to the hermeneutic circle in the humanities. Like a humanities scholar, I am assuming that the reader is acquainted with the detailed context of the question -- in particular, the many ways global warming will kill. By "semi-quantitative", I mean something between quantitative and qualitative: very approximate, like an order-of-magnitude estimate. Like a scientist, I am assuming the reader has some background in quantitative empirical methods and statistics, and approaches to dealing with quantitative uncertainty.

Looking at the problem systematically, it helps to divide causes of death into categories. In the following, I use the labels primary, secondary, and tertiary. Primary causes are associated with larger numbers of deaths (hundreds of millions altogether over 1-2 centuries, or millions per year) and/or higher probabilities. Tertiary causes kill fewer people (millions altogether, or tens of thousands per year) and/or they are associated with lower probabilities.

Primary causes of death

These factors will cause millions of deaths per year, or hundreds of millions altogether across a century, with high probability. Within given geographic areas, they will be inescapable for most people, depending on their mobility or financial means. That will lead to enormous and unprecedented death tolls -- barely imagineable for us today.

Humid heat. If wet-bulb temperature exceeds 35°C in a given region, millions could die on a single day. It hasn't happened yet, but when it starts to happen, there will be no reasonable way of stopping it from happening again (given the dangerous side-effects of geo-engineering solutions). People will try to adapt to a world in which such events happen with gradually increasing frequency and intensity. Northern India is one of the areas at risk. Even now, 100,000 people are dying every year from direct heat in Europe alone, whose population is about 1/10 of global population. So far, only a small fraction of that number can be attributed to climate change, but the fraction is increasing.

Starvation. Worldwide, agriculture will be affected by species extinctions (e.g., insects), disturbed (unpredictable) weather patterns, and deteriorating soil quality. The nutritive value of food will fall. Efforts to reduce greenhouse emissions of agriculture will push up food prices and restrict food supplies. Fishing often depends on coral reefs, which provide protein for one billion people and will hardly survive 2°C. Oceanic heatwaves will cause species extinctions in parts of the food chain, affecting other parts (example). In these and many other ways, hunger and starvation will steadily increase, eventually putting billions of people in mortal danger.

Water distribution. Changes in global water distribution, caused by global warming, are causing existing death rates to rise in various ways. Drier areas are becoming drier and wetter areas wetter, affecting agriculture (and hence food security). Almost two billion people rely on gradually disappearing mountain glaciers and snowpack for drinking water, at the same time as higher temperatures increase the demand for water. In addition, according to the UN, 3.6 billion people lack safe sanitation at home, and 1.8 billion live in homes without safe drinking water. Each year, 829,000 people including 400,000 children die from diseases attributable to unsafe water, inadequate sanitation, or poor hygiene. Current UN plans to address these issues will likely be overtaken by diminishing water supplies (drought) and disappearing glaciers. Water wars are on the horizon. Conversely, flooding causes water contamination from latrines and septic tanks, leading to cholera outbreaks and proliferation of vector-borne diseases. Water problems will turn untold millions into climate refugees.

Disease. In a conservative estimate, WHO anticipates that global warming will cause 250,000 deaths per year starting in 2030; this estimate is limited to undernutrition, malaria, diarrhoea and heat stress. DARA estimated that hunger and disease was causing 400,000 deaths per year, rising to 700,000 in 2030. Infections, parasites, AIDS, tuberculosis, and childhood diseases are also relevant; global warming is generally increasing their incidence (at least indirectly). Global warming is also changing the geographic distribution of vector- and rodent-borne diseases including arboviral (dengue, chikungunya, West Nile, and malaria), threatening previously unexposed human populations. New global epidemics, comparable with or worse than covid-19 (which killed 7m people), are possible, caused by disruption of complex ecosystems. In the arctic, old diseases to which humans have little resistance could emerge from the melting permafrost. The probability of a new epidemic triggered by animal-human contact or melting permafrost may not be high, but in both cases there could be hundreds of millions of premature deaths.

Children are especially at risk. According to UNICEF's Children’s Climate Risk Index, 240 million children are currently exposed to coastal flooding, 330m to riverine flooding, 400m to cyclones, 600m to vector borne diseases, 815m to lead pollution, 820m to heatwaves, 920m to water scarcity, and one billion to dangerously high levels of air pollution. "An estimated 850 million children – 1 in 3 worldwide – live in areas where at least four of these climate and environmental shocks overlap. As many as 330 million children – 1 in 7 worldwide – live in areas affected by at least five major shocks." Global warming is causing such death rates to increase gradually -- often in combination with chemical pollution (of water, food, household furnishings, clothing, cosmetics, transport, medications).

These are probably the main ways in which global warming of 2°C will kill. If the death rate in connection with poverty (the number of people who die prematurely because they cannot afford the necessary food or medical treatment) is currently ten million per year (conservative estimate; including three million children dying of hunger), and global warming of 2°C doubles that rate for a century, it will cause roughly a billion deaths altogether. Note that although many relevant death rates fell steadily in the late 20th century due to economic growth in developing countries and the success of international aid programs, since roughly 2015 death rates from hunger and poverty have been rising again, and global warming could ultimately be the main reason.

The current rate of deaths attributable to global warming, considering all the listed effects (both above and below), is probably roughly one million per year. That sounds like a lot, but it is far fewer than the 9 million currently dying every year from effects of pollution (including 6.7m from air pollution) (further details). Given the general trend, it is reasonable to predict that with 2°C of warming the death rate caused by global warming will rise to 10 million per year by 2100, and remain there for several decades, even in the most optimistic political scenario. Globally, there are currently 143 million births and 67 million deaths every year, so the prediction is that the the death rate will increase to 76 million per year, of which 10 million will be due to global warming.

Secondary causes of death 

Migration and conflict attributable to global warming will probably cause hundreds of thousands of premature deaths per year, and tens of millions altogether across a period of a century. Many people will manage to avoid both migration and conflict, being lucky enough to make that choice. For that reason, these death tolls will be lower than those linked to direct heat, hunger, and disease.

Migration. Sea-level rise and/or the threat of starvation will drive many to risk migrating, knowing that they may never arrive at their destination. Today, over 200 million people live less than one meter above sea level. Global warming of 2°C will probably mean a one-meter rise in sea level by 2100, forcing many to migrate. In the long-term, 2°C of warming will inundate the homes of 700 million people (Strauss et al., 2021). Above the rising tide, large areas of North Africa, the Middle East and Asia will become uninhabitable due to humid heat. Given current levels of poverty, roughly half of all climate refugees could die prematurely in an attempt to find a new home. So far, we have seen only the earliest warning signals: over 2000 refugees drown in the Mediterranean every year, and about 4000 die on land migration routes. As those numbers gradually rise, democratically elected far-right governments will respond with increasing callousness and even violence. 

Conflict. Global warming will cause conflicts over access to liveable land and dwindling water supplies in the midst of mass migration. The Middle East has about 1% of the world's fresh water and shares it among 5% of global population; the water is decreasing as population increases. Civilian water systems have already been attacked in wartime in Yemen, Syria, and Iraq. In other conflicts, the African Sahel has seen many recent national coups, in countries where the effects of global warming are especially apparent; "devastating floods, droughts, and heatwaves decimate access to water, food, and livelihoods, and amplify the risk of conflict. This will ultimately force more people to flee their homes" (source). The civil war in Ethiopia (Tigray, 2020-2022) killed 600,000 civilians while hundreds more died of hunger every day. Again, global warming played a role. In Iraq, the US/UK invasion of 2003 eventually caused half a million deaths -- mostly violent, and more civilian than military. Most Iraqi land is threatened by desertification, as the already hot climate gets even hotter. Drought is forcing people to move from the country to the cities (more), as also happened in the leadup to the Syrian civil war (more), which also killed about half a million people. In recent decades, many other conflicts have been exacerbated by global warming. In 2022, well over 200,000 people were killed in conflicts worldwide, and this global rate is increasing.

Tertiary causes of death 

The following additional causes of death due to global warming, although devastating by themselves, will be less serious than the others, given their lower probability, or the lower death toll of each event. They will probably cause tens of thousands of deaths per year, or millions per century.

Earthquakes, tsunamis, volcanoes. The incidence of earthquakes, tsunamis, and volcanoes is probably rising due to global warming. "Between 1900 and 1950, the Earth recorded an average of 3.4 earthquakes per year with a magnitude greater than 6.5. That figured doubled to 6.7 a year until the early 1970s, and was almost five times that in the 2000s" (source). Tectonic plates may expand slightly like railway tracks in the sun, but only if close to the surface. More likely, the effect will be indirect, with global warming causing aquifers to empty or glaciers to disappear, which can in turn trigger earthquakes (more). Earthquakes are low on probability but high on consequences. They are the most deadly natural disaster, killing 750 000 people globally from 1998 to 2017 (more than half of all deaths related to natural disasters). That includes the 2004 Indonesian tsunami, which killed over 200,000 people. Volcanoes currently kill only 500 people per year, on average; but a single volcanic eruption can kill 10,000 people, and 800 million people live within 100km of one of 500 active volcanoes (more).  

Extreme weather. So far, extreme weather represents a relatively small contribution to the global death toll from global warming. One UN report said that extreme weather has caused two million deaths in 50 years. But extreme weather can also cause death indirectly. For example, flooding can lead to cholera outbreaks.

Landslides. Landslides currently kill about 1000 per year.  So far only 1/10 of landslides can be attributed to global warming, but the proportion is rising as the number of landslides also rises. Causes include extreme rain, rising sea levels, and wildfires.

The analytical approach: Adding contributions

One way to predict the total death toll from global warming is to add estimates based on the above points (bottom-up approach), keeping in mind the following issues:
 Despite these caveats, a tentative quanitative analysis of predicted causes of death can clarify the scope of the problem. If and when global warming reaches 2°C, the annual global death toll that can be attributed to warming might break down as follows:
Needless to say, these numbers are very approximate. They can be regarded as hypotheses that can be tested by looking at the problem in different ways from different angles:
The holistic approach: Seeing the big picture

Consider the range of reasonably likely outcomes. Things could turn out much better or worse than expected. A holistic estimate of the most likely number of lives lost to global warming must take both worst- and best-case outcomes into account, assuming reasonable probabilities for different outcomes.

In a first pass at the problem, we can try to nail down the outer limits of the range of possible outcomes. If global warming is already killing roughly one million people per year (when most ways in which global warming can indirectly kill are taken into account, as explained above), and if by some miracle the situation does not get worse in coming decades, the total death toll after a century will be roughly 100 million. That can be regarded as a fixed lower limit, and of course it is only an order-of-magnitude estimate. A fixed higher limit is the entire future human population: 10 billion.

These limits suggest that the distribution of possible outcomes will be a log-normal, that is, normal (bell-shaped) relative to a logarithmic axis. Wikipedia (consulted on 24 October 2023) offers the following examples of log-normal distributions from everyday life: the length of comments posted in Internet discussion forums, the time users spend looking at online texts, and the duration of chess games.

The middle of such a curve -- the most likely outcome, or expected value -- is one billion, midway between the lower and upper limits on a logarithmic scale. It is possible to anchor two further points on the same distribution, which I will call reasonable best- and worst-case predictions. These are relatively extreme outcomes that are nevertheless reasonably likely.

A reasonable best-case outcome is one in which scientific predictions turn out to be exaggerated, or people adapt in unexpectedly creative and successful ways to global warming, saving many millions of lives. In such a scenario, the yearly global death toll from climate change might be 3 million rather than 10 million per year. A rate of 3 million per year corresponds to the current rate at which children are dying of starvation, so another way of describing this outcome is to say that global warming would double that rate.

In a reasonable worst-case prediction, we might anticipate a total death toll of 3 billion, or 30 million per year for a century. Some tipping points could be reached as soon as 1.5°C (Greenland ice sheet collapse, West Antarctic ice sheet collapse, tropical coral reef die off, boreal permafrost abrupt thaw). Even before global warming reaches 2°C, they could interact with the Atlantic Meridional Overturning Circulation, the El Niño-Southern Oscillation, and/or the Amazon rainforest tipping point, with catastrophic consequences for humanity. The earth could move toward a hothouse state of 4 to 5°C of warming. In the ensuing climatic chaos, one in three people worldwide might die prematurely, spread across several decades. That is possible even if "only" one trillion tonnes of carbon are burned, which without tipping points would cause 2°C of warming. 
A reasonable worst-case scenario also involves the temperature rise that we anticipate even if all emissions suddenly stop. If all emissions stopped tomorrow, mean global surface temperature would still rise by another degree: half a degree caused by the disappearance of air pollution, allowing more sunlight to enter, and another half degree as the system reaches a new equilibrium. In that way, burning a trillion tonnes of carbon could cause 3°C rather than 2°C of warming. These points are often left out of scientific predictions, raising doubts about their validity.

Summing up, the problem of predicting future death tolls can be tentatively solved by a technique called "top-down prediction". The process involves getting a feel for the detailed extent of the problem, considering the big picture, hazarding a first approximation, and checking whether this educated guess corresponds to available knowledge. Many readers will be unhappy about the inherent uncertainty of such an approach, but it may be unrealistic to claim more accuracy. If five different points on a standard and theoretically reasonable probability distribution are anchored by qualitative and quantitative arguments, we can be reasonably confident about the location of the curve's peak.

Alarmism without exaggeration 

This text has an alarmist feel about it. The text can also be seen as an attempt to present the most important facts, directly and objectively. In any case, alarmism can be seen as a rational response to an increasingly desperate global situation. It is surely ethically problematic to know how serious the situation is and not to be alarmed.

I am therefore proud to be an alarmist, and invite readers to be likewise. Whereas the climate alarm bells are now well and truly ringing, and every year the ringing gets louder -- if humanity is to survive global warming, or at least maintain reasonable living standards for most people, then many more alarm clocks will need to go off.

If alarmism is seen as necessary, interesting ethical questions about climate change communication arise. Is it ok to exaggerate things in order to achieve climate goals that are clearly in the public interest? Roger Hallam is an example: sometimes he exaggerates, but in that way he also managed to co-create Extinction Rebellion, which is surely one of the most important social movements of all time.

Global warming of 2°C will cause a billion premature human deaths in the long term. That is a shocking statement, and it is easy to fall into the trap of exaggerating it or distorting its meaning. The predicted death toll will largely be due to increases in existing death tolls from hunger, disease, and poverty. Doubling those existing rates could cause a billion deaths over a period of a century. Moreover, the catastrophe will be spread out across a long time period, reducing its subjective impact. 

Exaggeration is counterproductive if it reduces confidence in a source of information. But alarmism does not imply exaggeration. The best way to raise the alarm may be to state the facts as simply and directly as possible, without exaggeration. My 2019 article included a section on avoiding bias.

It can help to clarify that we are talking about real people and not mere statistics. The number of children under 10 years of age in the world is roughly one billion, and the vast majority of them live in the global South. These children really exist, right now. Every decade, a new generation in this age-group comes along. If global warming is going to cause premature deaths at a total rate of one billion per century, or 100m per decade, then 10% of all children under 10 today are going to die prematurely for that reason, at some time in the future. We are effectively putting the children of the world on aeroplanes, knowing in advance that 10% of those planes will crash. As an order-of-magnitude estimate, this prediction can hardly be wrong. The percentage of people who will die prematurely due to global warming is smaller for people in older age-groups. Whereas global warming is a deadly threat for everyone, the threat is bigger for younger than for older people.

One might argue that 10% is not that many. Back in 1662, demographer John Graunt estimated in this "Natural and Political Observations Made upon the Bills of Mortality" that one-third of London children died before their 16th birthday. But things have changed since then. With modern medicine, and a modern concept of human rights, we today reasonably expect and demand a long and happy life for every child, and are shocked if that universal birthright is not fulfilled.

If global warming is ignored, current global trends suggest that, by 2100, the number of children in the world will fall by about 40%. Reasons include better education and employment for women, better access to contraception, and less fear of poverty in old age. At the same time, the number of people over 80 will soar due to improved health services. But such predictions are problematic. The global chaos created by global warming will slow, stop, or reverse many of the positive developments that have led to lower birth rates and higher life expectancies. For that reason, order-of-magnitude predictions of future death tolls from global warming are unaffected relatively little by such demographic changes.

The future death rates that I am proposing are conservative relative to those of Hallam and some well-known climate scientists who have implied that several billion would die or only one billion would survive a few degrees of warming. In my view, those respected colleagues were describing worst-case scenarios. My estimates are instead consistent with related well-known work by John Nolt (a philosopher) and Danny Bressler (an economist). They are also consistent with the recent work of climate scientist Tim Lenton on the human climate niche, if we assume that roughly half of all future people who find themselves out of the niche will not be able to afford solutions (e.g., air conditioning; imported food and water) and will die prematurely as a result. Given high rates of poverty in the global South, this estimate is reasonable. Readers who, after considering these various sources of evidence, still doubt that 2°C of global warming will altogether cause a billion premature human deaths, or 10 million deaths per year on average for a century, are invited to read this.

The present text will mainly be read by middle-class citizens in rich countries. How should they (you) respond to the news that global warming of 2°C will kill a billion people? Again, it is important not to exaggerate. Most victims of global warming will have limited financial means, and they will mainly live and die in the global South. For most readers of this text, it is unlikely that family, neighbors, friends, colleagues, or politicians will die prematurely due to global warming. Instead, readers will experience falling living standards (quality of life) and increasing global chaos, possibly leading to social and economic collapse.

If that is not enough to motivate people to get politically active, another motivator is altruism. Most people reading this text identify as responsible adults and as good people. We believe that our heads and hearts are in the right place when it comes to moral or ethical issues. If that is really true, we should be thinking about the mortal danger that our emissions pose for countless millions of other people. We should be realizing that every conscious human on the planet has the same intrinsic value. If we care about long-term human survival, we will need to start caring about people in the global South -- people with less privilege, luck, or money than ourselves. I offer this not as some kind of holier-than-thou preaching but rather as a logical conclusion.

Why focus on humans?

For the purpose of evaluating the importance of global warming for humans, and comparing the importance of different contributions to global warming or different attempts to mitigate global warming, I am arguing that the value of a conscious human life is the most important value that we humans have. As such, it represents the foundation of all human values. Every other value can be compared to it, as a kind of standard.

For example, we might ask about the value of the Great Wall of China, the Roman Colosseum in Italy, the Taj Mahal in India, or the Great Pyramid of Giza in Egypt. The construction of such architectural wonders cost enormous numbers of lives. Imagine a dreadful thought experiment in which we have to choose between destroying one of those wonders and killing a group of people. How many people would we be prepared to kill? A possible answer is to say that we should not kill anyone at all, because the value of a human life is greater than any such material value.

A related issue is the value of non-human lives. I strongly support the humanist, utilitarian, moral philosophy of Peter Singer. I agree with him that the world should become vegetarian (to prevent unnecessary animal suffering, but also to mitigate global warming), and that richer people should donate enough to poorer people to eliminate poverty, or at least enough to save lives without sacrificing their own well-being (drowning child analogy). If someone proposed making it illegal to kill any animal at all, I would support it. Unlike Singer, however, I see a big difference between (self-) consciousness, which humans have much more of than non-human animals, and sentience, where the difference is smaller and some would argue there’s no difference at all. That sounds speciesist, but I also believe there are good reasons for adopting such a position.

From an empirical psychological perspective, consciousness is an eco-social phenomenon. Like language, it is acquired gradually by infants and children during interactions with carers and others. The claim is difficult to confirm directly due to the "hard problem of consciousness": strictly speaking, we are only directly aware of our own consciousness, and all other people could be zombies. We assume that other people have consciousness because it would be crazy not to. In a physicalist scientific approach, psychological indicators such as functional network connectivity in the brain and the development of attention and multimodal integration seem to indicate developing consciousness in infants, but at this level the distinction between perception, sentience, and consciousness is unclear. Behaviorally, we can observe how infants and young children gradually get a feeling for relationships between past, present, and future, and learn to imagine what is happening in different places (perception of occlusion and containment); such abilities seem closer to what adults intuitively understand by "consciousness".

Pet dogs acquire snippets of language and consciousness in similar ways, but to a much smaller extent, which puts them in an entirely different category. Besides, our thinking about non-human language and consciousness may be biased by out tendency to anthropomorphize our pets, just as we anthropomorphize human infants. When a baby smiles and coos, our baby schema is activated, which motivates us to care for the baby and to treat it as if it had language and consciousness. A similar effect can be observed between pet owners and pets. But empirical psychological studies show that a baby hardly has any (self-) consciousness, just as it hardly has any language. In that sense, a baby is comparable with a pet dog. The enormous difference between humans and non-humans is also clear in an evolutionary approach. For the past 100,000 years or so, humans and other primates have behaved in radically different ways that are presumably due to radical differences in the extent and sophistication of human and non-human language and consciousness.

These arguments are consistent with the idea that we can measure human value in units corresponding to conscious human lives. If we are to carefully pursue utilitarianism, systematically promoting human well-being and avoiding human suffering, we need a good measure of human value that can be systematically maximized. The conscious human life is a good candidate for such a measure on the assumption that the number of premature deaths caused by a natural or unnatural disaster is proportional to the amount of suffering for both those who die and those who survive. 

The path to human extinction

The path to extinction is clearly one that we must avoid. To do that reliably, we will need to jump out of our comfort zone and try to understand what that path would be like to actually walk along.

I suggested in my 2019 article that the relationship between the predicted death toll and the global temperature change is roughly linear in a first approximation. Something like this: one billion deaths at 2°C, 2bn at 3°C, 3bn at 4°C and so on until human extinction (10bn deaths) is reached at about 10°C. If we consider only wet-bulb temperatures, the death rate will increase suddenly -- not linearly. But direct heat is only one of many ways that global warming will kill. It is part of a more complex picture. Future people will respond creatively to the diverse difficulties and invent new ways to survive in a complex and chaotic world, which could result in a roughly linear relationship.

Seen another way: we humans, and the complex ecosystems upon which we depend, evolved to withstand global mean temperature fluctuations of about one degree Celsius, plus or minus. We are now outside that range, and our ability to survive is falling in proportion to distance from the range boundary. I am presenting that idea in the the style of a physicist, advancing a simple linear model as a first approximation to a complex problem. Behind such a model is a series of assumptions, such as this one: In an extreme worst-case scenario, with warming approaching 10°C, relatively rich people (if there is still money) may try to save themselves using technology, but that could be impossible without a human workforce to support it.

What if humanity throws all warnings to the winds and stubbornly burns all reasonably available fossil fuels? That is possible, given the past few decades in which scientific warnings were repeatedly ignored at the highest level. Think of the current political state of the COP conference series. The series' failure to achieve its main goal of reducing emissions suggests that most national governments have been captured by fossil fuel industries and are now incapable of responding to the facts of science and the will of the people. Many key players are still acting as if short-term profit for a rich minority is more important than long-term survival for everyone.

The amount of fossil carbon in the earth's crust that is reasonably available for extraction and burning has been estimated at five trillion tonnes. If burning that causes ten billion deaths, then the amount of fossil carbon corresponding to one future death is 500 and not 1000. From that perspective, the 1000-tonne rule is a conservative estimate. It is certainly not exaggerated.

Ignoring the world's most important issue

What is the most important task, of any person, ever? Such an open question can be answered in many different ways. Consider the following claims:
 If we accept all of these claims, it follows that global warming is the biggest risk that humanity ever took. Therefore, preventing it is our most important task ever. 

That being the case, it is remarkable that so few people are trying to predict future death rates from global warming, considering the massive ethical, political, and legal implications. Of all the groups of people involved in the climate debate -- scientists, activists, politicians, CEOs, deniers, the general public -- none seems willing to focus on future death tolls.

Is this about racism? We seem to know or assume that most victims of global warming will live in hotter countries and/or have dark skin. Are we unwilling to admit that intrinsic racism exists in most people and hence in ourselves, despite the clear psychological research findings?

Or are we simply overwhelmed? It's not easy to include the whole world in our thinking. We have known for all our lives that millions of people are dying in the global South from hunger and disease, and we also guess the North could save those lives with a simple redistribution of wealth that would still leave the North rich and the South poor. We feel bad about that, so we try to suppress it.

Psychologists have documented many biases to explain odd behavior. The availability bias (or heuristic), for example, is a tendency to overestimate the likelihood of things that are easier to recall. People find mortality statistics disturbing, which means they are seldom repeated, which in turn makes them hard to recall. 

I am reluctant to hunt for relevant psychological biases, because the act of doing so is itself biased, looking for biases that will confirm my opinion (confirmation bias). The following biases may nevertheless be relevant:
Conclusion

The bottom line is that all greenhouse gas emissions must urgently be stopped so that none of this every happens. On that point, Lynas and hopefully anyone reading this commentary will agree. I would add that the main thing limiting the rate at which fossil fuels should be stopped is the number of deaths the energy transition will itself cause. Economic factors are secondary, if (as I hope) human lives are considered more important than money.

In closing, I should clarify my qualification. Lynas suggested that this topic is outside of my field of expertise as a music researcher. In fact, I have postgraduate qualifications in physics and psychology. My approach to order-of-magnitude estimation and probability distribution are borrowed from basic physics. My research on music and my research on global warming are interdisciplinary in similar ways, both combining physics and psychology with other disciplines in humanities and sciences. The basic ideas are expressed in a non-technical way to promote public discussion.
The opinions expressed on this page are the authors' personal opinions.
Suggestions for improving or extending the content are welcome.
Back to Richard Parncutt's homepage