Update on K2-141b

Above, some art commissioned by NASA to commemorate the mission of the Kepler Space Telescope. Even as the mission encountered technical issues with its reaction wheels near the end of its journey, good science was extracted. The K2 mission allowed the telescope some drift from its original pointing near the constellation of Cygnus. This permitted the telescope to examine a wider range of stars, though each could not be observed for as long as the original set. Still, close-in short orbital period planets like K2-141b could be detected. This week, Giang updates us on his work to model this lava planet.

By Giang Nguyen

As 2021 marches on, I, along with everyone else in the lab, are also chugging along amidst a global pandemic. However, I’m more fortunate than most as my work solely lies in virtual space. As long as I have access to a computer and some internet, my work on K2-141b progresses. Although the internet isn’t necessary to run my models, I always need to have 5 tabs of Google searches on how to python open.

As a refresher, K2-141b is a planet about 200 light years away located in the Aquarius constellation. Its orbit is so close to the star it orbits that its surface temperature can be hot enough to melt and vapourize rocks, creating a thin atmosphere. K2-141b is also tidally locked which means there is a permanent dayside and nightside on the planet. Almost half of the planet is covered by giant magma oceans that can be over 100 km deep.

My work, previously published in the Monthly Notices of the Royal Astronomical Society, assumed an optically thin atmosphere such that all of the star’s light will reach to K2-141b’s surface unhindered. But now, we are restricting that assumption to account for the radiative transfer that occurs in the atmosphere. This involves calculating how much infrared and UV radiation is absorbed by the atmosphere and the subsequent feedback on the surface’s energy budget.

Infrared, or long-wave radiation, accounts for almost half of the stellar radiation but UV accounts for less that 1%. Although IR radiates much more than UV, an SiO atmosphere is better at absorbing UV than IR, 10,000,000 times better. As the atmosphere and the surface themselves also radiate infrared waves, they can warm each other up which complicates things further. Nonetheless, I pressed on with what I had and crunched the numbers.

If you neglect UV absorption, our results wouldn’t differ much from the results of our MNRAS paper. The atmosphere would be a bit warmer for a larger area but eventually drops down to near absolute zero when approaching the night-side; the winds would be about 33% faster. This makes sense as more energy is absorbed initially but, through sensible heat and radiative cooling, the system finds an equilibrium and temperature drops. Next step, add UV radiation.

Although UV absorption is minuscule initially, the atmosphere's thick optical depth at UV wavelengths ensures that 100% of UV stellar radiation is absorbed almost everywhere. While IR absorption and emission drops exponentially, UV absorption stays steady. Just like the tortoise and the hare, slow and steady wins the race. There comes a point when UV heating becomes the dominant radiative term and IR emission is not strong enough to cool the atmosphere. From there on out, the winds get faster and the temperature gets even hotter.

As you approach the night-side, the temperature gets hotter and hotter, upward of over 13,000 K. The winds go up to 7 km/s (400% increase from the classic no radiative transfer scenario). Unlike every other simulation, the results with UV absorption stayed subsonic throughout the entire atmosphere since the temperature rises faster than the wind’s acceleration. These strange results seem counterintuitive physically…but within the mathematical axioms we’ve built for our model everything checks out.

For now, I am further analyzing these results. Atmospheres like the one we have on Earth are also very good at absorbing UV radiation. However, our stratified atmosphere relegates that job to the top, far away from the surface. This leaves Earth with a temperature inversion, and K2-141b should have one too. The adiabatic profile may no longer be accurate when we introduce complex radiative transfer schemes. But if the results are true, then K2-141b’s atmosphere becomes plasma and flies across the surface at 7 km/s speed, like exhaust from a rocket. That’s pretty metal but I don’t know what the results are yet. I’ll keep you updated when I do. Until next time.