π Space π Science and technology
Satellites, black holes, exoplanets: when science extends beyond our planet

How do we study the climates of other planets?

Isabelle Dumé, Science journalist
On June 1st, 2022 |
4 min reading time
François Forget
François Forget
CNRS Research Director in Astrophysics
Key takeaways
  • At the Dynamic Meteorology Laboratory (LMD), researchers are studying the Earth's climate using satellite observations and numerical models to simulate the atmosphere.
  • Their objective is to predict what will happen on our planet in the future as well as on others in our Solar System.
  • For example, they have developed Dynamico, a tool to calculate circulation in Earth's atmosphere – low-pressure areas, anticyclones, and winds – which they have also used to study Mars and Venus.
  • They are also trying to model the climate on Mars from thousands or even billions of years ago to better understand recent ice ages or even the presence of lakes and rivers on its surface from a long time ago.

Space explo­ration is becom­ing ever more ambi­tious with new mis­sions to dif­fer­ent plan­ets in our solar sys­tem and beyond. My team at the LMD is con­tribut­ing to this glob­al effort by analysing the obser­va­tions made by these mis­sions and by devel­op­ing glob­al cli­mate mod­els to sim­u­late the behav­iour of extrater­res­tri­al atmos­pheres using uni­ver­sal physics equations.

At the LMD most of my col­leagues study Earth’s cli­mate using satel­lite obser­va­tions and numer­i­cal mod­els that sim­u­late its atmos­phere. The aim is to mod­el changes in cli­mate and project what will hap­pen in the future – say, 50 years from now. We have adapt­ed our tech­niques and applied them to the atmos­pheres of the oth­er plan­ets in the solar sys­tem as well as to Titan (a moon of Sat­urn) and Tri­ton (one of Neptune’s moons) and of course Plu­to, for­mer­ly our ninth planet.

Numerical global climate models

Whether they are ter­res­tri­al or extrater­res­tri­al, are a bit like a video game but they are based on phys­i­cal equa­tions that allow us to cal­cu­late all the phe­nom­e­na around a plan­et – its clouds, winds, atmos­pher­ic cir­cu­la­tion, dust storms, frost, snow. We then try to see if, sim­ply on the basis of well-known the­o­ret­i­cal equa­tions, we can rep­re­sent all these phe­nom­e­na. This goal is very ambi­tious – some­times we don’t suc­ceed but, often, the mod­els work aston­ish­ing­ly well. The results also allow us to bet­ter under­stand our plan­et’s atmos­phere by reap­ply­ing lessons learned else­where. It is a bit like in med­i­cine, in which ani­mal mod­els are used to bet­ter under­stand the human body.

We are cur­rent­ly ded­i­cat­ing much of our time to the plan­et Mars and are involved in a num­ber of space mis­sions includ­ing ESA’s Mars Cli­mate Orbiter and Insight, an Amer­i­can mis­sion, for which French col­leagues pro­vid­ed the seis­mome­ter. This geo­phys­i­cal sta­tion is also a weath­er sta­tion. Again, we try to inter­pret the obser­va­tions we make with our numer­i­cal mod­els. We are also devel­op­ing an ambi­tious EU-sup­port­ed project called Mars Through Time, which involves using our extrater­res­tri­al cli­mate mod­els to try and mod­el the cli­mate on Mars thou­sands, mil­lions and even bil­lions of years ago, when its orbit and axis of rota­tion were a lit­tle dif­fer­ent and there were ice ages (recent­ly) and even lakes and rivers (much longer ago) on its surface.


Surprises galore

For this and oth­er projects, we recent­ly applied a new com­put­er pro­gramme called Dynam­i­co, orig­i­nal­ly devel­oped to study the Earth’s cli­mate, to solve the equa­tions of flu­id mechan­ics to cal­cu­late atmos­pher­ic cir­cu­la­tion – that is how depres­sions, anti­cy­clones and winds evolve. When we applied the pro­gramme to Mars and Venus, we found that it did not sim­u­late the sit­u­a­tion on Venus very well. This is because sub­tle effects that the mod­el neglects on Earth can be much stronger on oth­er plan­ets, which means that we some­times have to improve the mod­el by adding cer­tain terms to the equa­tions it con­tains. We encoun­tered a sim­i­lar sit­u­a­tion a few years ago when apply­ing our mod­els to Mars. The extra term we added to our equa­tions in this case not only helped us improve our mod­el so that it bet­ter described the Mar­t­ian atmos­phere, it also helped us bet­ter sim­u­late the mon­soon in India when it was reap­plied to Earth’s atmos­phere. This may seem like a small detail, but it is an impor­tant result if we want to try and under­stand whether there will be a major drought or tor­ren­tial rains in this area of the world in the future because of cli­mate change. So, study­ing Mars’ cli­mate has helped us to bet­ter under­stand Earth’s.

The results from these mod­els, which are used by hun­dreds of research teams from around the world, are also cru­cial for prepar­ing future space mis­sions – espe­cial­ly those designed to land on the sur­face of a plan­et, or which will make use of the atmos­phere to slow down their space­craft. We are fund­ed by space agen­cies and indus­try for these projects. The mod­els can gen­er­al­ly be trust­ed – which, inci­den­tal­ly, is very use­ful for con­vinc­ing some cli­mate scep­tics. Indeed, it is amaz­ing how well our mod­els often rep­re­sent what we observe. It is fas­ci­nat­ing to see that a mod­el will have pre­dict­ed per­fect­ly how, for exam­ple, the winds behave on a plan­et and that these pre­dic­tions are con­firmed by the data sent back from a real probe.

The sit­u­a­tion is even more inter­est­ing when the mod­el does not work. Some­times this is because the sit­u­a­tion is very com­pli­cat­ed or “non-lin­ear”, which means that the cli­mate is extreme­ly sen­si­tive to such a such para­me­ter. The mod­el has to be fine­ly test­ed and adjust­ed to account for this sen­si­tiv­i­ty. Most often, this implies that there is a phys­i­cal phe­nom­e­non that we have not thought of and that is, in fact, present. This phys­i­cal phe­nom­e­non may not be a process that acts direct­ly on the envi­ron­ment by heat­ing or cool­ing it, for exam­ple. It may be a feed­back that dri­ves the cli­mate sys­tem into a cer­tain oper­at­ing regime, that is, a par­tic­u­lar cli­mate. We need to under­stand these phys­i­cal sys­tems, which have mil­lions of degrees of free­dom and com­bine many length and time scales, and appre­ci­ate what makes them enter cer­tain states and func­tion as they do. A real chal­lenge and a beau­ti­ful 21st cen­tu­ry physics prob­lem for us to solve.


Our world explained with science. Every week, in your inbox.

Get the newsletter