La lune


Documentation > Greenhouse effect > Risks > Do we know if there is a "threshold of danger"?

Is there a well identified threshold in climate change ?

september 2003

website of the author : - contact the author :


Thresholds in climate change

A process is said to have a threshold if there exists a given level of perturbation - called a threshold - under which the effect is proportionnal to the perturbation, and over which things evolve in a totally different way.

A good example is a rubber : let's pull it a little, and it lenghthens. Let's pull it a little more, and it lengthens a little more. We can prolongate this effect a little while : any additionnal little traction applied to this rubber will result in a little additionnal lengthening, until....the rubber suddently breaks. We then have passed beyond a threshold : a little increase of the perturbation has generated a brutal - and sometimes hardly foreseeable in a precise way - change of the nature of the consequence (breaking instead of lengthening), and, in addition, in the chosen example, the change is not reversible (we cannot stick back together the two bits of the rubber).

Our climate is full of such processes with thresholds, for which a little change of the perturbation may lead to abrupt and massive changes of the nature of the consequence, and we generally don't know precisely - sometimes not at all - when and where a component is susceptible to "break down", that is to start to behave in a very different way from what we are accustomed to. We now start to know that it could be the case for marine currents, for the marine life depending on corals, or the response of polar ice caps to the ultimate temperature rise. But for most climate processes the thresholds are not necessarily known, and when they are qualitatively, there are seldom precise quantitative estimates.

In addition a given temperature increase can be without major consequences after a century, but lead to massive effects later on : to know whether, for this or that component of the climatic machine, we have passed beyond a threshold or not, we also need to know what time horizon is associated to the question.

As a conquence, it is not possible, today, to say that there exists a temperature rise below which we would be guaranteed against any major problem, but over which we would suddently start to be subject to considerable risks.. No scientist can state that it is reasonnable not to go over a temperature rise of 1.63 °C, that as long as we don't cross this threshold we are guaranteed against any catastropha, but that as soon as we reach a 1.64 °C the Apocalypse is for tomorrow, or that a CO2 concentration of 435 ppmv is harmless, but not of 436 ppmv.


Thresholds in the capacity of adaptation

What we call a risk does not only result from a change in environmental conditions, but depends also on our faculty - or absence of faculty - to face the said change. A stone that falls on our head does not mean the same risk whether we wear a helmet or not. If we go back to climate (my favourite subject here, did you guess ?), a limited diminution of the agricultural output resulting from a climate change is not a major "risk" for a country like France, that produces way over its own consumption, but will possibly be qualified as such for India or China if it happens there today.

More generally speaking, if we try to evaluate the risks attached to a climate change that might happen in 50 or 200 years, we need to be able to describe not only the involved climate modifications or their direct consequences (are we talking of the sea level ? of the marine currents ? of possible diseases ?) but we must also guess what means will be at the disposal of humanity at that time (how much energy will remain out of the present abundance ? will there still be forests or natural resources that might held to adapt ? etc). Guessing the latter is much harder than speculating on the future of the climate !

In particular, what fundamentally governs our aptitute to change the world, as it has done so in the past and will do so in the future, and therefore conditions our aptitude to face a given evolution, is abundant energy. It's this abundance that allows us to build roads and use trucks to carry food (grown with a lot of energy, also !) from places where it grows, that allows to resist to cold or to heat, that allows to produce vaccines, tools, medications, that enables us to be mobile in mass, etc. And all these things resulting from abundant energy concur to make us resistant to many adverse evolutions.

But whether energy will still be abundant - or cheap, which is about the same - in 50 years is a vast debate ! For fossil fuels, for example, a simple prolongation of present trends over 50 years leads to a total exhaustion of all proven reserves (coal included), without even mentionning the associated deterioration of the climate that would result from the freeing in the atmosphere of all the corresponding CO2.

Let's come back to agriculture, without which it would not be possible to have a couple billion people on earth (earth could definitely not bear several billion men living only from hunting and picking). If nothing grows any more in a particular place, the situation is not equally serious whether we still have abundant energy to import massively food from elsewhere, or if this possibility is no longer available.

In the same respect, the ability to resist new diseases will not be the same in a country endowed with important sanitary infrastructures, and in a country that isn't equipped, etc.


And at last a catastropha is a variable-geometry notion : nobody puts the threshold at the same place !

A "crossing of threshold" is often called, in common language,....a catastropha. For example, a landslide resulting in deaths (crossing of a threshold concerning the viscosity of earth) is a catastropha, an outbreak (crossing of a threshold concerning the exposition to pathogen agents) is a catastropha, a storm (crossing of a threshold concerning the speed of wind) is a catastropha, etc.

Suppose we are able to foresee all the precise consequences of a given climate change we could - or have - put in motion (which is not the case), in order to know whether we have already gone beyond the threshold of danger there still remains to define what is a catastropha. With such a definition, as long as we don't get any catastropha, we can consider that we are still below the threshold of danger, and if we do get catastrophas in the future, we are already over the threshold.

Well, there starts the difficulty : a catastropha is not an objective notion, and it all depends on who speaks.

Let's take a concrete case : the possible disappearing of corals if the global temperature rises by more than 2 °C. Is it a catastropha ? The answer will probably not be the same whether you ask a travel agent specialized in diving trips or a Norvegian farmer, taking only two occidentals for this comparison.

Suppose climate change brings three hurricanes per winter in France (identical to the ones we have experienced in 1999). Is it a catastropha ? For me perhaps, but for a french roofer surely not, and for an inhabitant of Chile probably not.

Suppose climate change kills 2 billion individuals, that would starve to death. Is it a catastropha ? I will take the risk to be very politically incorrect : it depends on the place where people starve, the possible repercussions "elsewhere", and who's talking....


By way of conclusion...

Knowing whether we are already beyond the threshold of danger requires to have visibility on :

the future evolution of our climate system for any given emission scenario of greenhouse gases,

the "means to resist" that will still be at our disposal for the time horizon we are talking about,

what is personnally or collectively acceptable among the possible consequences.

Consequently, given the ignorance on what our adaptation possibilities in several decades - or centuries - will be, the necessary subjective appreciation of what constitutes an unbearable consequence, and the inertia of the involved physical processes, it is not possible to say when we have possibly "crossed the threshold of tolerable risk", and hence no more possible to say if it is already the case.

Back to climate change index
Back to the top