Author: Paulo Simões, Lyndsay Fletcher, Hugh Hudson (University of Glasgow), Graham Kerr (NASA GSFC), Guigue Giménez de Castro (Centro de Rádio Astronomia e Astrofísica Mackenzie), Matt Penn (NSO)
During a solar flare, rare infrared (IR) continuum observations show strong brightenings in the chromosphere, co-spatial and co-temporal with flare hard X-ray footpoints . Similar to the much better known flare optical observations, the IR gives a direct window into the rapid energisation of the chromosphere – but with better detectability. The first tentative description of infrared (IR) flare continuum  proposed that it was due either to thermal free-free radiation – i.e. bremsstrahlung – from the flare chromosphere, or to blackbody emission from the heated photosphere. Figure 1. shows some recent flare IR observations obtained with the McMath-Pierce telescope, and in anticipation of flare IR observations with the Daniel K. Inouye Solar Telescope, we have used radiation hydrodynamic simulations of a heated flare chromosphere to investigate the generation of the flare IR, and investigate what it tells us about the evolution of the flare chromosphere .
Infrared emission mechanisms in the solar chromosphere
The IR continuum emission mechanisms are relatively straightforward, even in flares. The continuum source function, which essentially corresponds to the intensity emitted over a photon’s mean-free path, is given by the local Planck function, with temperature equal to the local electron (kinetic) temperature. The opacities depend on the electron, proton and neutral hydrogen densities in the flare, and in a flare these must be treated in non local thermodynamic equilibrium (nLTE). The main source of IR opacity in the chromosphere is H+ free-free absorption. At the photosphere, the H– free-free opacity dominates.
Radiation hydrodynamic simulations of flares
In a flare the rapid energisation of the chromosphere leads to heating and, primarily, rapid changes in ionisation. The flare chromosphere is not expected to be in ionisation equilibrium, so we turn to nLTE modelling. We use the RADYN code [e.g. 4, 5] which solves the time-dependent, coupled, non-local equations of hydrodynamics, radiation transport and atomic level populations in a 1D atmosphere, including backwarming by coronal soft X-ray, EUV and UV radiation. Flare heating is modelled using an input beam of electrons, with collisional evolution treated in the Fokker-Planck approximation.
Motivated by the rapid IR flare time variability observed by McMath-Pierce we simulate a short electron beam pulse of 4s with a symmetric triangular time profile. The beam energy flux F as a function of energy E is described by F = F0E-δ above a low-energy cutoff Ec. In Figure 2 we show (a) the atmospheric temperature (b) the contribution function at a characteristic IR wavelength of 5μm, (c) electron density and (d) optical depth for different values of Ec and the spectral index δ , plotted as a function of height above the photosphere, at the peak energy injection time. These simulations are for an energetic flare, with a maximum value of flux of 1011 erg cm-2 s-1.
The details of the atmosphere vary with beam parameters, with low δ and high Ec beams tending to have more effect on deeper layers. But all show an increase in temperature, electron density and contribution function between 0.7Mm and 1.5Mm, i.e. the mid-to-upper chromosphere. The contribution function shows where the emergent radiation originates, and from Figure 2(b) it is clear that both before and during the flare most IR emission comes from near the photosphere, at around 0.1Mm. However, the biggest change in the IR emission, i.e. the IR flare excess, is from the higher altitude chromospheric layer. It is caused by the rapid increase in flare temperature (Figure 2a) and electron density (Figure 2c), the latter due to ionisation primarily of hydrogen (and, to a lesser extent, helium). The chromosphere is optically thin where this emission is formed.
A rapid response
We can also examine the time evolution of the IR emission in response to the energy input. This is shown in Figure 3. Curves of the infrared brightness temperature TB (basically, the intensity) for the different beams, at four different wavelengths, are superposed on the time profile of the energy input. The right-hand axis shows the flare IR contrast, which at wavelengths of a few microns is expected to be in the few to ten percent range, consistent with the McMath-Pierce observations. This should be readily observable above the photospheric background, since the contrast is higher and the background variability due to granulation is much less at these wavelengths.
The IR light curves at 2 and 5 μm peak very close in time to the peak of the energy input, but the peak is delayed because ionisation is still increasing for a short while after the energy input peak. This is because the recombination timescale is longer than the ionisation timescale – visible also in the presence of the ‘tail’ in the TB curves. However, even at short wavelengths the delay is only fractions of a second, making this a very prompt signature of energy input.
Infrared observations of solar flare chromospheric sources look like they will provide useful flare diagnostics, being both a prompt response to the energy input, and directly related to the evolution of ionisation in the chromosphere. We have also been able to show  that as the flare rises to its peak IR intensity, the brightness temperature maps very closely the evolution of electron density where the contribution function peaks, offering the possibility that the IR brightness temperature can be used to track something like the chromospheric ‘total electron content’. Flare observations in the infrared continuum are rare at present, but this will change with the Daniel K. Inouye Solar Telescope which will be able to observe parts of the IR continuum between 1 and 5 μm, initially, though with capability to go out to 10μm in the future.
The research leading to these results has received funding from the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 606862 (F-CHROMA).
References Penn, M., Krucker, S., Hudson, H. et al. 2016, ApJ 819, L30
 Okhi. K, Hudson, H. S. 1975, Sol. Phys. 43, 495
 P. J. A. Simões, G. S. Kerr, L. Fletcher et al., Astron. Astrophys. (2017, submitted)
 Carlsson, M. & Stein, R. F. 1995, ApJ, 440, L29
 Allred, J. C., Kowalski, A. F., & Carlsson, M. 2015, ApJ, 809, 104