6 June 2018 in Astronomy & Astrophysics
Context.The Helioseismic and Magnetic Imager (HMI) provides continuum images and magnetograms with a cadence better than one per minute. It has been continuously observing the Sun 24 h a day for the past 7 yr. The trade-off between full disk observations and spatial resolution means that HMI is not adequate for analyzing the smallest-scale events in the solar atmosphere.Aims.Our aim is to develop a new method to enhance HMI data, simultaneously deconvolving and super-resolving images and magnetograms. The resulting images will mimic observations with a diffraction-limited telescope twice the diameter of HMI.Methods.Our method, which we callEnhance, is based on two deep, fully convolutional neural networks that input patches of HMI observations and output deconvolved and super-resolved data. The neural networks are trained on synthetic data obtained from simulations of the emergence of solar active regions.Results.We have obtained deconvolved and super-resolved HMI images. To solve this ill-defined problem with infinite solutions we have used a neural network approach to add prior information from the simulations. We testEnhanceagainst Hinode data that has been degraded to a 28 cm diameter telescope showing very good consistency. The code is open source.
20 June 2019 in Astronomy & Astrophysics
Context. Spectropolarimetric inversions are routinely used in the field of solar physics for the extraction of physical information from observations. The application to two-dimensional fields of view often requires the use of supercomputers with parallelized inversion codes. Even in this case, the computing time spent on the process is still very large. Aims. Our aim is to develop a new inversion code based on the application of convolutional neural networks that can quickly provide a three-dimensional cube of thermodynamical and magnetic properties from the interpreation of two-dimensional maps of Stokes profiles. Methods. We trained two different architectures of fully convolutional neural networks. To this end, we used the synthetic Stokes profiles obtained from two snapshots of three-dimensional magneto-hydrodynamic numerical simulations of different structures of the solar atmosphere. Results. We provide an extensive analysis of the new inversion technique, showing that it infers the thermodynamical and magnetic properties with a precision comparable to that of standard inversion techniques. However, it provides several key improvements: our method is around one million times faster, it returns a three-dimensional view of the physical properties of the region of interest in geometrical height, it provides quantities that cannot be obtained otherwise (pressure and Wilson depression) and the inferred properties are decontaminated from the blurring effect of instrumental point spread functions for free. The code, models, and data are all open source and available for free, to allow both evaluation and training.
1 March 2021 in Astronomy & Astrophysics
Context.The evolution of the photospheric magnetic field plays a key role in the energy transport into the chromosphere and the corona. In active regions, newly emerging magnetic flux interacts with the pre-existent magnetic field, which can lead to reconnection events that convert magnetic energy into thermal energy.Aims.We aim to study the heating caused by a strong reconnection event that was triggered by magnetic flux cancelation.Methods.We use imaging and spectropolarimetric data in the Fe I6301& 6302 Å, Ca II8542 Å, and Ca IIK spectral lines obtained with the CRISP and CHROMIS instruments at the Swedish 1-m Solar Telescope. These data were inverted with the STiC code by performing multi-atom, multi-line, non-local thermodynamic equilibrium inversions. These inversions yielded a three-dimensional model of the reconnection event and surrounding atmosphere, including temperature, velocity, microturbulence, magnetic field, and radiative loss rate.Results.The model atmosphere shows the emergence of magnetic loops with a size of several arcseconds into a pre-existing predominantly unipolar field. Where the reconnection region is expected to be, we see an increase in the chromospheric temperature of roughly 2000 K as well as bidirectional flows of the order of 10 km s−1emanating from there. We see bright blobs of roughly 0.2 arcsec in diameter in the Ca IIK, moving at a plane-of-the-sky velocity of the order of 100 km s−1and a blueshift of 100 km s−1, which we interpret as ejected plasmoids from the same region. This scenario is consistent with theoretical reconnection models, and therefore provides evidence of a reconnection event taking place. The chromospheric radiative losses at the reconnection site are as high as 160 kW m−2, providing a quantitative constraint on theoretical models that aim to simulate reconnection caused by flux emergence in the chromosphere.
11 September 2019 in Astronomy & Astrophysics
The topology and dynamics of the solar chromosphere are greatly affected by the presence of magnetic fields. The magnetic field can be inferred by analyzing polarimetric observations of spectral lines. Polarimetric signals induced by chromospheric magnetic fields are, however, particularly weak, and in most cases very close to the detection limit of current instrumentation. Because of this, there are only few observational studies that have successfully reconstructed the three components of the magnetic field vector in the chromosphere. Traditionally, the signal-to-noise ratio of observations has been improved by performing time-averages or spatial averages, but in both cases, some information is lost. More advanced techniques, like principal-component analysis, have also been employed to take advantage of the sparsity of the observations in the spectral direction. In the present study, we use the spatial coherence of the observations to reduce the noise using deep-learning techniques. We designed a neural network that is capable of recovering weak signals under a complex noise corruption (including instrumental artifacts and non-linear post-processing). The training of the network is carried out without a priori knowledge of the clean signals, or an explicit statistical characterization of the noise or other corruption. We only use the same observations as our generative model. The performance of this method is demonstrated on both synthetic experiments and real data. We show examples of the improvement in typical signals obtained in current telescopes such as the Swedish 1 m Solar Telescope. The presented method can recover weak signals equally well no matter what spectral line or spectral sampling is used. It is especially suitable for cases when the wavelength sampling is scarce.
4 August 2022 in Astronomy & Astrophysics
Context. Our knowledge of the heating mechanisms that are at work in the chromosphere of plage regions remains highly unconstrained from observational studies. While many heating candidates have been proposed in theoretical studies, the exact contribution from each of them is still unknown. The problem is rather difficult because there is no direct way of estimating the heating terms from chromospheric observations. Aims. The purpose of our study is to estimate the chromospheric heating terms from a multi-line high-spatial-resolution plage dataset, characterize their spatio-temporal distribution and set constraints on the heating processes that are at work in the chromosphere. Methods. We used nonlocal thermodynamical equilibrium inversions in order to infer a model of the photosphere and chromosphere of a plage dataset acquired with the Swedish 1-m Solar Telescope (SST). We used this model atmosphere to calculate the chromospheric radiative losses from the main chromospheric cooler from H I, Ca II, and Mg II atoms. In this study, we approximate the chromospheric heating terms by the net radiative losses predicted by the inverted model. In order to make the analysis of time-series over a large field of view computationally tractable, we made use of a neural network which is trained from the inverted models of two non-consecutive time-steps. We have divided the chromosphere in three regions (lower, middle, and upper) and analyzed how the distribution of the radiative losses is correlated with the physical parameters of the model. Results. In the lower chromosphere, the contribution from the Ca II lines is dominant and predominantly located in the surroundings of the photospheric footpoints. In the upper chromosphere, the H I contribution is dominant. Radiative losses in the upper chromosphere form a relatively homogeneous patch that covers the entire plage region. The Mg II also peaks in the upper chromosphere. Our time analysis shows that in all pixels, the net radiative losses can be split in a periodic component with an average amplitude of amp̅Q = 7.6 kW m−2 and a static (or very slowly evolving) component with a mean value of −26.1 kW m−2. The period of the modulation present in the net radiative losses matches that of the line-of-sight velocity of the model. Conclusions. Our interpretation is that in the lower chromosphere, the radiative losses are tracing the sharp lower edge of the hot magnetic canopy that is formed above the photosphere, where the electric current is expected to be large. Therefore, Ohmic current dissipation could explain the observed distribution. In the upper chromosphere, both the magnetic field and the distribution of net radiative losses are room-filling and relatively smooth, whereas the amplitude of the periodic component is largest. Our results suggest that acoustic wave heating may be responsible for one-third of the energy deposition in the upper chromosphere, whereas other heating mechanisms must be responsible for the rest: turbulent Alfvén wave dissipation or ambipolar diffusion could be among them. Given the smooth nature of the magnetic field in the upper chromosphere, we are inclined to rule out Ohmic dissipation of current sheets in the upper chromosphere.
17 March 2023 in Astronomy & Astrophysics
Context.Observations of center-to-limb variations (CLVs) of spectral lines and continua provide a good test for the accuracy of models with a solar and stellar atmospheric structure and spectral line formation. They are also widely used to constrain elemental abundances, and are becoming increasingly more important in atmospheric studies of exoplanets. However, only a few such data sets exist for chromospheric lines.Aims.We aim to create a set of standard profiles by means of mosaics made with the CRISP and CHROMIS instruments of the Swedish 1-m Solar Telescope (SST), as well as to explore the robustness of said profiles obtained using this method.Methods.For each spectral line, we used a mosaic that ranges from the center to the limb. Each of these mosaics were averaged down to 50 individual spectral profiles and spaced by 0.02 in theμscale. These profiles were corrected for p-mode oscillations, and their line parameters (equivalent width, line shift, full-width at half-maximum, and line depth) were then compared against literature values whenever possible.Results.We present a set of 50 average profiles that are spaced equidistantly along the cosine of the heliocentric angle (μ) by steps of 0.02 for five continuum points between 4001 and 7772 Å, as well as ten of the most commonly observed spectral lines at the SST (Ca II H & K, Ηβ, Mg I 5173 Å, C I 5380 Å, Fe I 6173 Å, Fe I 6301 Å, Ha, O I 7772 Å, and Ca II 8542 Å).Conclusions.The CLV of line profiles and continua are shared in the CDS as machine readable tables, providing a quantitative constraint on theoretical models that aim to model stellar atmospheres.