Climate Death Toll Model
A model of human deaths due to climate change as a function of carbon burned and temperature change

Richard Parncutt
28 August 2014

Richard Parrncutt ICMPC 2012

"Do the math" is the catchcry of climate activists drawing attention to the world's carbon budget. If we want to avoid warming beyond 2°C, we have to avoid burning more than 1 trillion tons of carbon altogether; both the degree of warming and the amount of burned carbon are calculated relative to the start of industrialisation in the 18th Century. We have already burned half of that and heated the planet by almost 1°C. Climate science is saying that, given the international agreement to limit warming to 2°C, we can burn only another half a trillion tons. The world's carbon reserves add up to about 4 trillion tons, so about 3/4 of that will have to stay in the ground. That is perhaps the main thing that politicians and the general public need to know about climate in coming years. Are we going to achieve this goal or not?

This article takes "doing the math" a step further. My math is not about carbon - it's about people. I ask: How many people will climate change indirectly kill? And instead of merely guessing, I will develop a simple model, based on a small number of clearly stated assumptions. Of course the predictions will be very approximate, but an approximate beginning is better than nothing. I hardly need mention that the predictions are absolutely terrifying. So fasten your seatbelts. But if we are going to solve this problem we have to keep a clear head and look carefully at the detail.

First some background. According to accepted usage, climate change will become "dangerous" when it reaches 2°C relative to pre-industrial temperatures, which will happen if CO2 concentration stabilizes at 450 ppm (it is currently near 400, and it was 280 before the industrial revolution). To understand what "dangerous" means, please read this report from the World Bank. The way we are headed right now, the chance of exceeding 2°C of warming is high. Even if the main emitting countries achieve current emission reduction targets, scientists agree that the final temperature will exceed 2°C warming with high probability. On top of that, the major emitters have a history of not achieving their own targets. Consider the case of Canada, which withdrew from the Kyoto Protocol in order to avoid paying the bill. Canadian politicians and fossil fuel corporations then tried to assuage their guilt by engaging creatively in gentle art of climate denial. If behavior like this continues, our children can look forward to a very grim future.

The dismal failure of the Copenhagen and Warsaw climate conferences was part of a familiar pattern. Progress on climate negotiations is regularly destroyed in the background by the climate denial lobby. The net result of all this denial is a lot of talk and not much action. Our response to the climate crisis has so far consistently been far too little too late. In fact, it may be too late already. But we must cling to a certain naive optimism, for fear that people will give up altogether.

So far we have almost 1°C of warming. The situation is not yet "dangerous", although there are already many serious adverse affects of climate change, which are causing tens of millions of human deaths. At 1.5°C, which could be reached even if CO2 levels stabilize near 400 ppm, there will be more serious consequences. Consider rising seas, which is just one of many dangerous consequences of climate change. At 1.5°C, many low-lying islands will become uninhabitable. Large areas of Florida will be destroyed (it's happening already, but Florida's  politicians are still denying the climate change that is happening right under their noses). Hopefully that will alert Americans to the true dimensions of the crisis. Better late than never! (Perhaps even Canadians holidaying in Florida will take note.) At 1.5°C the effect on river deltas such as for example in Bangladesh will already be serious and large numbers of people will  be trying to migrate. As the temperature continues to rise, the sea will continue to rise - inexorably, with every decade for at least a century, even if all carbon emissions suddenly stopped now.

The global situation will become increasingly tense and there will be no relief on sight. Needless to say, nothing like this has every happened in human history. Take the Second World War for example. The war ended in Europe when the Russians, Americans, French and British defeated the Nazis. A decade later, things were getting back to normal.
(I am leaving out the Holocaust here, after which nothing could really be "normal" again.)  There will be no such deliverance from global warming. Our descendents will have to wait for centuries before there is a chance of things getting back to "normal", whatever that means exactly. Within their lifetimes, things will only get worse.

Estimating the death toll from global warming

As we reach and probably exceed
2°C of warming later this century, we can expect to see steady increases in global death rates associated with poverty: hunger, preventable disease and curable disease in developing countries. The current total death rate from these preventable causes is already deeply shocking, at about 10 million per year. A global warming scenario that stabilizes at 450 ppm CO2 and 2°C pf warming is likely to double this death rate, as I have estimated and explained elsewhere. If you add deaths from wars over resources and mass migration of climate refugees, the total could easily approach one billion. That is about 10% of the projected maximum human population of 10 billion. 

Humanity will presumably survive 2°C of warming. I am assuming that about 90% of people will survive, and they will also find ways to adapt. That raises the question of how much global warming would be necessary to wipe out humanity. The ice ages of the past millennium were minor affairs compared to the warming that is predicted in the coming century. About 66 million years ago, climate change wiped out the dinosaurs. For the purpose of argument, I will assume about 10°C of warming would be enough to make the human species extinct, in the following ways:
Any one of thse points could lead to extinction, but a combination of different deadly threats is more likely.

The climate change leading to the death of the last human would be partially due to human activity and partly natural.The first few degrees of warming would be mainly due to human emissions. That is what we are already experiencing. Later, natural positive feedback processes (less reflection of sunlight from ice, more forest fires, release of stored methane) would kick in, pushing temperatures even higher, even if human emissions had completely stopped. These natural processes would cause temperatures to continue to rise even after humans and most other species had become extinct.

Based on these assumptions, the following graph shows the number of human deaths that we might reasonably expect for each degree Celcius of warming. The number of deaths is assumed to be added up (integrated) over a long period. What could "long" mean in this context? I do not have a clear answer to that question, but an answer of some kind is necessary. For the purpose of argument, I will assume here that a "long time" is about 70 years, the typical duration of a human life. So we could be talking about the total number of people who die anywhere in the world as an indirect consequence of climate change between the years 2030 and 2100. Of course the predicted numbers of deaths are only very rough estimates, but any estimate is better than none if we are to conquer denial and look the danger in the eye.

I am assuming in a first approximation that the death toll will increase linearly with temperature. That is unlikely, of course. But without a strong theoretical basis, I prefer not to say anything specific about nonlinearity. An argument for non-linearity might go like this: Right now, global warming is just starting, so we expect a slow start to the death toll, accelerating later on - something like an exponential curve. This argument is misleading, because the death toll from global warming is not starting from nothing. Even without global warming, we already have about 10 million deaths per year from hunger, preventable disease and curable disease in developing countries. That's one billion per century. I am assuming that global warming of 2°C will double this rate. A century (or perhaps more appropriately, one lifetime of about 70 years) may seem like a long time for such a calculation. But we are dealing with very long-term climatic changes. It could take centuries or even thousands or millions of years to reverse such changes. Relative to the duration of a human life, the climatic changes we are talking about are effectively permanent. The higher death tolls to the right side of the graph (approaching human extinction) might happen in shorter time spans: decades or even years. In that case, the death ratewould accelerate non-linearly with increasing temperature. But the relationship between the total number of deaths and the  temperature reached might still remain a linear, as shown in the graph.

Even if there is a good argument for linearity, this graph still seems like a wild guess. But the graph is based on a small number of plausible assumptions. These are:
According to the text accompanying the National Geographic Channel film "Six degrees could change the world",

An increase of 4ºC would see the oceans rise drastically. Then comes the twilight zone of climate change, if the global temperature rises again by another degree. Part of once temperate regions could become uninhabitable, while humans fight each other for the world’s remaining resources. The sixth degree is what is called the doomsday scenario as oceans become marine wastelands, deserts expand and catastrophic events become more common.

According to the Executive Summary of the 2012 report "Turn Down The Heat: Why a 4°C Warmer World Must be Avoided" by the Postdam Institute for Climate Impact Research (PIK) for the World Bank,

A 4°C world will be one of unprecedented heat waves, severe drought, and major floods in many regions, with serious impacts on ecosystems and associated services. (...) The 4°C scenarios are devastating: the inundation of coastal cities; increasing risks for food production potentially leading to higher malnutrition rates; many dry regions becoming dryer, wet regions wetter; unprecedented heat waves in many regions, especially in the tropics; substantially exacerbated water scarcity in many regions; increased frequency of high-intensity tropical cyclones; and irreversible loss of biodiversity, including coral reef systems (...) Thus, given that uncertainty remains about the full nature and scale of impacts, there is also no certainty that adaptation to a 4°C world is possible. A 4°C world is likely to be one in which communities, cities and countries would experience severe disruptions, damage, and dislocation, with many of these risks spread unequally. It is likely that the poor will suffer most and the global community could become more fractured, and unequal than today. The projected 4°C warming simply must not be allowed to occur--the heat must be turned down. Only early, cooperative, international actions can make that happen.

On this basis, it is reasonable to predict that 6°C warming will mean premature death for half of 10 billion people.
I am also assuming that some humans will survive in the cooler parts of the planet until 10 or 11°C. Given the above quotes, my assumptions are rather optimistic and conservative.

How much carbon do we have to burn to cause one future death?

Humans have already burned half a trillion tons of carbon, and the global mean temperature is already almost 1°C higher than during the preindustrial period. If we burn another half a trillion tons of carbon, the warming will be 2°C (more). Total known carbon reserves are about 4 trillion tons. Let's assume in a first approximation that the earth's temperature rises about one degree for every additional half a trillion tons that we burn. Think of the greenhouse effect: adding an extra half a trillion tons is like putting an extra blanket around the earth. For purpose of argument, I will again assume a linear relationship: every such "blanket" will increase the temperature by about the same amount, at least to begin with. So 2 trillion tons of carbon might lead to a 4°C rise. However there are also positive feedback processes to consider. Presumably they will become increasingly important as the temperature rises. Let's say the 3rd trillion tons causes 3°C of warming: 2°C due to greenhouse gases and an extra 1°C due to positive feedback. That is a pretty conservative estimate - it could be much worse. In the same way, we might assume that the 4th trillion tons will cause 4°C of warming: 2° due to greenhouse gases and 2° due to positive feedback. The resultant graph would then look like this: 


We can now combine the above two graphs into one, and consider the number of deaths as a function of the total amount of carbon that has been burned:

For the purpose of this graph, I have assumed that the death toll has already begun (for details see the Human Impact Report of the Global Humanitarian Forum). The graph is curving upwards, which means the more carbon we burn, the greater the number of future deaths for a given amount of carbon. This now allows us to estimate the number of tons of carbon corresponding to one death, depending on the degree of warming. I am interested in this question, because it enables us to estimate the number of future deaths that are being caused by specific fossil fuel industries today.

At 2°C of warming, we will have burned 1 trillion tons of coal, causing 1 billion deaths altogether over a long time period. That's one thousand tons of carbon per death. At 4°C it will have been 2 trillion tons for 3 billion deaths, which is 670 tons per death on average: the more people die, the smaller the amount of carbon that is needed to kill them. At 7°C, according to this model, it's 3 trillion tons for 6 billion deaths or 500 tons per death. At 11°C, at which point I am assuming that last human will already have died, it's 4/10 = 400. These calculations are summarized in the next graph.

If the numbers are getting confusing, it may help to look at it this way: If humans burned all 4 trillion tons of carbon at their disposal, global temperature would rise by 8 degrees without considering positive feedback. Positive feedback, which climate scientists agree will become serious beyond about 4°C of warming (if not before), would add several more degrees. The result would almost certainly mean human extinction. Remarkably, many intelligent, well-read people are still wondering what will happen after we have used up all the fossil fuel reserves. Have you ever heard someone idly mention that oil, coal and gas reserves will one day run out, and then ask what are we going to do then? What indeed. If people are still thinking this way, despite 20 years of high-profile public information on global warming, we should assume that they are serious about collective suicide. We should calculate the human cost of fossil fuel burning as if humanity really is going to destroy itself. That means assuming that one future person will die for every 400 tons of carbon burned, which is simply 4 trillion divided by 10 billion.

I have argued elsewhere that you kill one future person every time you burn 1000 tons of carbon. This last graph shows that this is a conservative estimate, based on the rather optimistic assumption that global warming will be brought under control. If it is not brought under control, we will be killing one future person every time we burn much less than 1000 tons.  I'm not a hysterical climate activist (as some hysterical deniers have claimed), so in the following I will stick to the more conservative estimate of 1000 tons per death.

How accurate are these numbers? Not very, of course. They can be regarded as order-of-magnitude estimates, which are often (tacitly) assumed in physics to
have an uncertainty of 30% (since log102  0.3). For example, if "100" is an order-of-magnitude estimate, it lies between 30 and 300. The number 1000, for the scenario in which warming is brought under control, is calculated on the assumption that 1 billion people will die as an indirect result of 2°C of warming over a period of some 70 years. That in turn is based on the assumption that the first graph above is linear, the linear trend will begin at 1°C of warming, and humanity will (would) be extinct when (if) warming reaches (reached) 11°C. Given the general consistency of the model and diverse sources of information to which I have referred, I would say that the uncertainty in the number 1000 is about 30%; in any case I am not aware of any strong arguments that my estimate is too high or too low. The number 400 is the number of tons of coal to kill one person in a human extinction scenario. This is the quotient to two numbers, 4 trillion and 10 billion. I have assumed that if we burn all 4 trillion tons of carbon that are currently known to be available, humanity will die out. Let's say that the number 4 trillion has an uncertainty of 25%; the expected maximum world population of 10 billion could be out by 10%. In that case, the quotient of these two numbers has an uncertainty of about 30% (the percentages do not add; if the distribution of errors is assumed normal, we square the uncertainties, add, and take the square root).


If we assume that global warming will be brought under control and apply the more generous and conservative estimate of 1000 tons of carbon per death to the world's leading fossil fuel industries, there is only one possible conclusion that we can come to. They should all be closed down as soon as possible.

I am Australian, so let me start close to home (more). Much of the electricity consumed in Melbourne, a city that I know and love because I grew up there, a city of three million people, is generated by burning brown coal at Yallourn power station. That power station produces 22% of all electricity consumed in the state of Victoria. On their website I read that "Every hour 2,400 tonnes of brown coal is used to boil water into superheated steam to drive four turbine generators." Brown coal is about 40% carbon, so that is like burning 1000 tons of carbon per hour, which according my calculations is killing one future person every hour.

According to the IEA Clean Coal Centre, a staggeringly sophisticated example of climate denial describing itself as "The global resource on the clean use of coal", there are about 2300 coal-fired power stations in the world. My model suggests that each of them is killing several future people every day. Incidentally, speaking of the IEA: regardless of how efficient a coal power station is or what kind of "CO2 mitigation" technologies are being developed, if you generate electricity by burning coal, you always produce massive amounts of CO2. The amount ultimately depends on the amount of electricity you generate. You can't change the basic laws of physics and chemistry.

Victoria produces 68 million tons of brown coal per year. I guess that's about 20 million tons of pure carbon, which translates to 20 000 future deaths per year. That shocking enough, but it is only the start, because this is only a fraction of Australia's coal production. The port of Newcastle in NSW is exporting 100 million tons of black coal per year, and they are planning to export even more in the future. That translates to 100 000 future deaths per year. Total coal exports from Australia are about twice that. A massive new mine - one of the biggest in the world - was recently approved by the Australian government for construction at Carmichael in Queensland. It will produce another 60 million tons of coal per year, which I guess will mean at least 30 million tons of pure carbon per year, or another 30 000 more future dead per year. As if there was no tomorrow - and perhaps there isn't. For more shocking facts about Coal in Australia see the wikipedia page.

Australia, with only 20 million people, is the world's largest coal exporter. No wonder the economy is doing well. No wonder climate denial is thriving. On the website "Australians for coal", I read that the Australian coal industry employs 200 000 people (incidentally, 200 000 is also the number of future deaths caused by Australian coal every year according to my model, if we include both domestic and export). The coal industry contributes some AUD 60 billion per year to the Australian economy. On that basis, one might conclude that the Australian coal industry is happy to keep burning coal in spite of the future deaths that are being caused, provided there is enough money in it. This money adds up to about AUD 300 000 for each future death. What does this mean for people working in the coal industry? Of course it is not easy to change jobs, but people who have this opportunity should do so. Please get out as quickly as possible! My calculations suggest that that someone working in the Australian coal industry can save one future life every year by getting out. If they stay, they have one future life per year on their conscience. For those who cannot easily change their career, we need a political solution, and we need it fast. The solution is to move the entire industry into sustainable energy. The transition can be financed by environmental taxes, including the carbon tax that Australia repealed in 2014. Given the amount of urgent work to be done in sustainably energy, and the many largely untapped sources of tax revenue (e.g. wealth taxes, Tobin tax), it should be possible to re-employ all of those 300 000 people while at the same time maintaining electricity supplies throughout the country.

Coal is still the main way of generating electricity in the world, and it is a major cause of global warming (see Wikipedia: coal). Coal exports alone mean that Australia's per capita contribution to the destruction of the planet is enormous. So next time you catch yourself thinking that Australians are nice friendly people, think again. Australians have been aware of their enormous contribution to destroying the planet for many years. The whispering in their hearts is telling them that their prosperity is being won at the cost of the planet and future generations, but most people are evidently not listening to their hearts. In 2013, Australians elected a prime minister in an election in which climate change was one of the main issues. Before the election, the future prime minister aligned himself solidly with climate deniers, and millions of Australians voted for him in a stunning display of selfishness or stupidity (it's often hard to know which is which). (The previous sentence is intended as a logical conclusion - not as an insult or political statement.) In defence of my fellow Australians, I should also say that an ever-increasing number are fighting hard to stop this madness, but they are having limited success. Democracy is going downhill as the wealth gap widens. I guess the country is increasingly in the grips of billionaires and their sidekicks. Please sign my global wealth tax petition.

According to Living Planet Report, which takes into account a range of factors (carbon, grazing, forest, fishing, cropland, built-up land), the ten countries with the highest per capita ecological footprint are Qatar, Kuwait, UAE, Denmark (meat production), USA, Belgium (high population and small area), Australia, Canada, Netherlands (high population and small area), and Ireland (high agricultural emissions). Anyone who has the courage to consider the probable death toll from global warming will agree with me that these countries are committing a horrifying crime, while pretending to be innocent.

When we realise that every 1000 tons of burned coal is causing a future death, our response can no longer be a question of reducing carbon emissions by 10% here and 20% there. Carbon emissions must completely stop as quickly as possible. Of course it is practically impossible to suppress the fossil fuel industry overnight. So the next question we must ask is: How fast can this change happen? What criteria can or should determine maximum speed of change? I wish to argue that the main reason is not (or should not be) economic. Instead we should consider those people whose health depends crucially on energy supplies. Many people would die as a result of a sudden global ban on the use of fossil fuels (if such a thing were politically possible). We should therefore base our strategy on human rights and consider human lives to be the most valuable thing we have. Closing down fossil fuel industries quickly will have its own death toll, which should be minimized, even if closing down these operations is simultaneously saving far greater numbers of future lives. No matter which way you look at it, questions of life and death are more important than questions of money and profit. 

I would be grateful if a reputable climate scientist would read this text carefully and comment on it. Please direct me to literature that would help me improve the quantitative estimates in the model, and correct any other errors.

The opinions expressed on this page are the authors' personal opinions.
Suggestions for improving or extending the content are welcome at
Back to Richard Parncutt's homepage