Author: Petra Kohutova and Erwin Verwichte at the University of Warwick.
A little bit of background
Coronal rain is a common phenomenon occurring in active region coronal loops. As the name suggests, it shares many parallels with its terrestrial counterpart; it consists of cool plasma condensations falling from coronal heights towards the solar surface guided by magnetic field lines. Coronal rain is formed as a direct consequence of thermal instability. A coronal loop is likely to become thermally unstable if it is heated predominantly at the footpoints. If the thermal conduction along the loop is not efficient enough, the radiative losses from the plasma at the loop top can overcome the heating input from the footpoints, resulting in onset of a thermally unstable regime and in the runway cooling of the plasma at the loop top. This leads to the formation of cool and dense plasma condensations, which then fall towards the solar surface in the form of coronal rain showers. Recent high resolution observations have shown that coronal rain is much more common than previously thought , suggesting it has an important role in the chromosphere-corona mass cycle. Due to its origin, coronal rain also provides us with physical insight into the atmospheric thermal cycle.
What have we observed?
We analysed observations of a thermally unstable loop taken by IRIS (Fig. 1). The loop shows a significant amount of coronal rain formation. As soon as the coronal rain appears, the loop starts to oscillate in the direction parallel to the plane of the loop. From time-distance plots we can deduce that the coronal rain formation excites a vertically polarised kink mode (Fig. 2). This is not surprising, as abrupt formation of a cool and dense condensation region at the loop top pushes the loop axis downwards. Lorentz force acts to restore the loop shape and the loop starts to oscillate as a result. The lack of observable damping suggests that the excitation is prolonged in time. We further observed a decrease in the loop oscillation period. This is a result of the evacuation of the loop top as the coronal rain drains mass from the loop that then falls towards the footpoints. If the mass of the loop changes, we expect a shift in the period of the fundamental kink mode. Based on the observed change of the period, we can estimate the total mass of the plasma lost due to coronal rain to be a third of the total loop mass (see  for a complete discussion). We also observed a change in the oscillation amplitude. Using a relation between oscillation amplitude and the mass of the condensation plasma derived in , we arrive at total rain mass consistent with the estimate based on the change in period.
Can we simulate this?
Yes! We carried out 2.5D MHD simulations of the above scenario using Lare2d code. The problem is set up to include a realistic stratified atmosphere consisting of a cool chromosphere, a thin transition region layer and a hot corona. We used current-free potential magnetic field configuration representative of a coronal arcade and created a long and thin loop that is denser than the surrounding plasma. Finally, we introduced a cool and dense condensation region at the loop top (see  for a complete description of the setup). As in the observations, the presence of the condensation region is found to excite fundamental harmonic of a vertically polarised kink mode in the loop. We also managed to reproduce the decrease in the oscillation period of the loop, which can be seen in the simulation as the condensations leave the loop top and fall into loop legs (Fig. 3). Considering that the simulation is set up such that the total mass of the plasma in the loop is conserved, the period shift in this case is caused by the redistribution of the mass along the loop leading to the change in the longitudinal density profile.
Why is this important?
This mechanism can explain excitation of vertically polarised loop oscillations without the need for placing specific external drivers below the loop. Continuous rain formation leading to prolonged excitation can explain the apparent absence of damping. Our work therefore shows that we need to be careful when drawing conclusions from the observations of “decayless” small-amplitude coronal loop oscillations commonly seen in AIA [5,6] and linking them to photospheric modes, as they can be excited by thermal instability in the corona. Finally, we showed that the oscillations excited by coronal rain have seismological potential. The evolution of the oscillation period can be used to determine the fraction of the loop plasma mass that becomes unstable. This gives us information about the parameters of the heating function and hence tells us how localised the heating of the coronal loop is.
-  Antolin, P. & Rouppe van der Voort, L. 2012, ApJ, 745, 152
-  Verwichte, E. & Kohutova, P. 2017, A&A, 601 L2
-  Verwichte, E., Antolin, P., Rowlands, G., Kohutova, P., & Neukirch, T. 2017, A&A, 598, A57
-  Kohutova, P. & Verwichte, E. 2017, A&A, 606 A120
-  Wang, T., Ofman, L., Davila, J. M., & Su, Y. 2012, ApJ, 751, L27
-  Nisticò, G., Nakariakov, V. M., & Verwichte, E. 2013, A&A, 552, A57