Let me start with something to dispel the confusion about what models are for. When you deal with complex, adaptive systems, models are NOT meant to predict the future. As John Gall said in his book on complex systems, “systems always kick back” – to which I may add, “and sometimes they kick back with a vengeance“. (another way to express this concept is “forecasting always fails.“)

But if dynamic models cannot predict the future, what are they good for? Simple, they are about being prepared for the future. Think of the Paris climate treaty of 2015. It was the result of millions of runs of various climate models, none of which claimed to predict “the” future. But these models are tools to prepare for the future; they tell you what may happen, depending on what you do. They are tools to shape political decisions. Out of all those runs, a goal was extracted, a setpoint, a number: “we don’t want temperatures to rise of more than  2 °C and, for that purpose, there is a limit to the amounts of fossil fuels we can burn.” It was a political decision that took into account not just what the models say, but what could be concretely achieved in the real world.  No model would give you that number as an output. The Paris agreement was a masterpiece of diplomacy and of communication strategy because it concentrated so much noise into a simple, stark, number: a goal to reach.

And there we stand: with Paris, we set the goal, but how do we get there? This section of policy planning was poor in Paris, where the best that could be done was to line up the INDCs, the intended nationally determined contribution; that is how single countries think they could reduce emissions. That’s not planning, it is a first stab at the problem; it shows the good will to do something, but no more. As they stand, the INDCs won’t get us far enough.

So, we are again at the task of getting prepared for the future. We know that we need to reduce carbon emissions, but how fast? Besides, it is not just a question of reduction, it is a question of substitution. We need to maintain the essential energy services to the world’s population: surely, as a society, we can shed a lot of fat and keep going, but without a minimum of energy input, the system collapses. At the same time, we need to maintain the current input without exceeding the emissions limits. A difficult challenge, although not an impossible one.

Here, we need models, again. No model can tell you exactly how to get there, but models will tell you what is likely to happen given some choices and some decisions. And out of the models, you have to extract a concrete, politically feasible goal: how to invest the remaining resources into attaining the Paris objectives? In other words, what fraction of the world’s GDP need to be invested in the transition to a renewable economy?

Giving an answer to this question is the ambitious task of the MEDEAS project which has now reached a full year of work and set up the basis for an extensive modeling effort. MEDEAS takes an approach mainly based on system dynamics, similar to the one of the well-known “The Limits to Growth” approach. It is not the only ongoing project in this area, others projects take different lines of approach. But in all cases the idea is to build up knowledge on what is needed for the transition. Some data are already available that tell us we need a major effort to replace fossil fuels fast enough. The transition won’t come by itself pushed by purely economic forces. But we need to explore the issue more in depth before these considerations can be turned into a number that can be agreed upon by the interested parties. We need to take into account both what’s needed and what is politically feasible. Then, we will have a goal to reach.

If you want to know more about MEDEAS, you can see the MEDEAS website. There is also a MEDAS newsletter, still in a preliminary phase. And, if you would like to be involved, contact me (ugo.bardi(strangething)unifi.it)