Periodic Reporting for period 4 - SUNMAG (SUNMAG: Understanding magnetic-field-regulated heating and explosive events in the solar chromosphere)
Okres sprawozdawczy: 2022-07-01 do 2023-06-30
The main scientific questions that the SUNMAG project has addressed are:
1. What are the amount and distribution of energy that is released in the chromosphere of magnetically active regions and flares?
2. What are the physical mechanisms responsible of releasing this energy?
3. What is the physical state of the solar photosphere and chromosphere when flares are triggered and how is flare triggering related to the chromospheric magnetic field configuration?
Solar activity and flares modulate space weather, can disrupt satellite communication and can affect the power grid on earth.
Our team has successfully developed and implemented all the software developments that were contained in the proposal.
This includes a multi-resolution spatially-coupled inversion method and its implementation into the STiC code.
The calculation of the energy losses by radiation has been implemented into STiC and tested using simulated datasets were the ground truth is known.
We have used STiC to reconstruct depth-stratified model atmospheres of the solar outer layers from a set of spectropolarimetric observations.
From the inferred models, we have calculated the first high-resolution maps of chromospheric radiative losses, which pose an estimate of the energy budget that must be provided by heating mechanisms.
Our results show that in plage observations, the radiative losses have a time-dependent component that correlates very well with the passage of shock waves in the chromosphere. The amplitude of this variation is roughly 1/3 of the total loss. Our interpretation was that the static component could be sustained by Ohmic dissipation or ambipolar diffusion, whereas the dynamic component is sustained by the slow-mode of magnetohydrodyamic waves. The mean value of the radiative losses was estimated to be on average 26 kW m^-2, with peak values over 80 kW m^-2.
The study of the radiative losses in a flux-emerging region undergoing strong magnetic reconnection yields concentrated losses of up to 160 kW m^-2. By analizing the geometry of the target, and the values of other parameters we concluded that the strong loss of energy by radiation was being sustained by the conversion of magnetic energy into heat and particle acceleration though magnetic field reconnection.
Our study of the radiative losses during a C-class flare and the consecuent comparison with numerical simulations of similar events was very useful to a) characterize the radiative losses during the flare peak and b) to set constraints on the parameters of the electron beam that transports energy from the corona into the chromosphere.
1. What are the amount and distribution of energy that is released in the chromosphere of magnetically active regions and flares?
2. What are the physical mechanisms responsible of releasing this energy?
3. What is the physical state of the solar photosphere and chromosphere when flares are triggered and how is flare triggering related to the chromospheric magnetic field configuration?
Solar activity and flares modulate space weather, can disrupt satellite communication and can affect the power grid on earth.
Our team has successfully developed and implemented all the software developments that were contained in the proposal.
This includes a multi-resolution spatially-coupled inversion method and its implementation into the STiC code.
The calculation of the energy losses by radiation has been implemented into STiC and tested using simulated datasets were the ground truth is known.
We have used STiC to reconstruct depth-stratified model atmospheres of the solar outer layers from a set of spectropolarimetric observations.
From the inferred models, we have calculated the first high-resolution maps of chromospheric radiative losses, which pose an estimate of the energy budget that must be provided by heating mechanisms.
Our results show that in plage observations, the radiative losses have a time-dependent component that correlates very well with the passage of shock waves in the chromosphere. The amplitude of this variation is roughly 1/3 of the total loss. Our interpretation was that the static component could be sustained by Ohmic dissipation or ambipolar diffusion, whereas the dynamic component is sustained by the slow-mode of magnetohydrodyamic waves. The mean value of the radiative losses was estimated to be on average 26 kW m^-2, with peak values over 80 kW m^-2.
The study of the radiative losses in a flux-emerging region undergoing strong magnetic reconnection yields concentrated losses of up to 160 kW m^-2. By analizing the geometry of the target, and the values of other parameters we concluded that the strong loss of energy by radiation was being sustained by the conversion of magnetic energy into heat and particle acceleration though magnetic field reconnection.
Our study of the radiative losses during a C-class flare and the consecuent comparison with numerical simulations of similar events was very useful to a) characterize the radiative losses during the flare peak and b) to set constraints on the parameters of the electron beam that transports energy from the corona into the chromosphere.
We reached the following milestones of the project:
1. We acquired most of the data that are required for the project. The observations were performed in yearly campaigns between 2018-2023. All observations were coordinated with NASA's IRIS satellite that co-observed our targets.
2. We developed a spatially-regularized weak-field approximation method for quick inference of chromospheric magnetic fields from observations.
3. Our postdocs and PhD students have finished the analysis of datasets acquired in plage and a flux-emerging region.
4. We have modified our codes to allow for the computation of chromospheric radiative losses, which is ultimately an estimate of chromospheric heating rates.
5. We have characterized and studied the spatio-temporal distribution of chromospheric radiative losses from a plage, a flare and a emerging-flux observations. In all cases, we found peak values in the radiative losses that are much larger than previously assumed, which means that the amount of energy that must be accounted for can be larger than previously assumed.
6. The STiC code includes now a full multi-resolution inversion approach that allows processing datasets acquired at very different spatial-resolution.
7. We have also made publicly available a simplified version of the multi-resolution inversion code (pyMilne) that can be easily used for proof-of-concept studies or to train scientists into these techniques.
8. We have worked into an acceleration technique for the STiC code based on Newton-Raphson and Newton-Krylov solvers. This was not originally planned, but due to the computational demands of the code, such acceleration was needed in order to allow for the processing of large FOVs.
Our team members have regularly attended international conferences where our results have been disseminated every year among the solar and stellar physics communities.
Additionally, we have attended other PR meetings that are oriented to a more general public ("Astronomdagarna" in Sweden), we have published three dissemination articles in newspapers and we wrote a PR article of the project for the Spanish Astronomy Society (SEA, https://www.sea-astronomia.es/sites/default/files/bv2023_espectroscopia.pdf ).
1. We acquired most of the data that are required for the project. The observations were performed in yearly campaigns between 2018-2023. All observations were coordinated with NASA's IRIS satellite that co-observed our targets.
2. We developed a spatially-regularized weak-field approximation method for quick inference of chromospheric magnetic fields from observations.
3. Our postdocs and PhD students have finished the analysis of datasets acquired in plage and a flux-emerging region.
4. We have modified our codes to allow for the computation of chromospheric radiative losses, which is ultimately an estimate of chromospheric heating rates.
5. We have characterized and studied the spatio-temporal distribution of chromospheric radiative losses from a plage, a flare and a emerging-flux observations. In all cases, we found peak values in the radiative losses that are much larger than previously assumed, which means that the amount of energy that must be accounted for can be larger than previously assumed.
6. The STiC code includes now a full multi-resolution inversion approach that allows processing datasets acquired at very different spatial-resolution.
7. We have also made publicly available a simplified version of the multi-resolution inversion code (pyMilne) that can be easily used for proof-of-concept studies or to train scientists into these techniques.
8. We have worked into an acceleration technique for the STiC code based on Newton-Raphson and Newton-Krylov solvers. This was not originally planned, but due to the computational demands of the code, such acceleration was needed in order to allow for the processing of large FOVs.
Our team members have regularly attended international conferences where our results have been disseminated every year among the solar and stellar physics communities.
Additionally, we have attended other PR meetings that are oriented to a more general public ("Astronomdagarna" in Sweden), we have published three dissemination articles in newspapers and we wrote a PR article of the project for the Spanish Astronomy Society (SEA, https://www.sea-astronomia.es/sites/default/files/bv2023_espectroscopia.pdf ).
The SUNMAG project aims at producing realistic 3D model atmospheres of the solar photosphere and chromosphere in order to study chromospheric heating mechanisms.
In order to produce such models, we have combined three techniques that were not used before in similar studies. These techniques are:
1. the inclusion of many chromospheric and photospheric spectral lines in one simultaneous data inversion.
2. the inclusion of vertical and horizontal regularization in order to constrain the parameters of the inferred models.
3. a proper 2D treatment of spatial instrumental effects in order to be able to combine data from different facilities acquired at different resolution.
Techniques 1. , 2. and 3. are already implemented in STiC (our analysis code) and in a fast inversion code that has also been made openly available to the community (pyMilne).
We have acquired very high quality spectropolarimetric datasets with the new CHROMIS instrument at the highest spatial resolution that is available at the moment (~75 km at the surface of the Sun) in combination with diagnostics from the CRISP instrument at the Swedish 1-m Solar Telescope and with NASA's IRIS satellite in the ultraviolet. We have observed several flaring active regions, which are excellent targets for the SUNMAG projects.
We have also developed a technique for de-noising of polarimetric data using deep-learning methods. This method uses spatial coherency in the polarimetric images in order to increase the signal-to-noise ratio of the observations.
The latter is crucial for the inference of the magnetic field vector on the surface of the Sun with inversion methods.
Our team has improved the "weak-field approximation" technique, which allows for a quick estimate of the magnetic field from a set of observations, by imposing horizontal regularization in the solution.
Our results indicate that this improvement allows extracting information from noisier datasets than with the traditional approach.
Additionally, our team has developed a pipeline that allows computing the chromospheric radiative losses from the inverted models from the STiC code.
We have produced and studied one of the first spatially resolved maps of chromospheric radiative losses in plage, a flare and an active region undergoing strong magnetic reconnection.
We have performed the first spatially-coupled / multi-resolution inversion of an IRIS / SST dataset.
In order to produce such models, we have combined three techniques that were not used before in similar studies. These techniques are:
1. the inclusion of many chromospheric and photospheric spectral lines in one simultaneous data inversion.
2. the inclusion of vertical and horizontal regularization in order to constrain the parameters of the inferred models.
3. a proper 2D treatment of spatial instrumental effects in order to be able to combine data from different facilities acquired at different resolution.
Techniques 1. , 2. and 3. are already implemented in STiC (our analysis code) and in a fast inversion code that has also been made openly available to the community (pyMilne).
We have acquired very high quality spectropolarimetric datasets with the new CHROMIS instrument at the highest spatial resolution that is available at the moment (~75 km at the surface of the Sun) in combination with diagnostics from the CRISP instrument at the Swedish 1-m Solar Telescope and with NASA's IRIS satellite in the ultraviolet. We have observed several flaring active regions, which are excellent targets for the SUNMAG projects.
We have also developed a technique for de-noising of polarimetric data using deep-learning methods. This method uses spatial coherency in the polarimetric images in order to increase the signal-to-noise ratio of the observations.
The latter is crucial for the inference of the magnetic field vector on the surface of the Sun with inversion methods.
Our team has improved the "weak-field approximation" technique, which allows for a quick estimate of the magnetic field from a set of observations, by imposing horizontal regularization in the solution.
Our results indicate that this improvement allows extracting information from noisier datasets than with the traditional approach.
Additionally, our team has developed a pipeline that allows computing the chromospheric radiative losses from the inverted models from the STiC code.
We have produced and studied one of the first spatially resolved maps of chromospheric radiative losses in plage, a flare and an active region undergoing strong magnetic reconnection.
We have performed the first spatially-coupled / multi-resolution inversion of an IRIS / SST dataset.