HIAC
Available Inversion and Quick Look Analysis Schemes
MERLIN | LILIA | Artificial Neural Net | HMI VFISVA major component of CSAC is to act as a community repository for a variety of computer codes to invert spectro-polarimetric data. The codes below are open-source and available for download. We also would like to know if you have a similar code to share.
Milne-Eddington gRid Linear Inversion Network (MERLIN)
MERLIN is a general-purpose inversion code based on the least-squares fitting of the observed Stokes profiles using the Levenberg-Marquardt algorithm. It assumes a standard Milne-Eddington Atmosphere will use single or multiple line Stokes profile to retrieve magnetic field vector, line-of-sight velocity, source function, Doppler broadening as well as the macroturbulence and the stray light filling factor. MERLIN, because of its robust nature, forms the core of the Hinode SOT/SP inversion client.
MERLIN-Based Hinode Inversion Client: Here
MERLIN-Based Hinode Inversion Client Tutorial (PDF): Here
Online Documentation and User's Guide: Here
An example of a MERLIN inversion of Hinode/SP data is shown below.
Status: Publicly available.
Downloads:
LTE Inversion based on the Lorien Iterative Algorithm (LILIA)
LILIA is a general-purpose inversion code based on the least-squares fitting of the observed Stokes profiles. It assumes LTE, 1D plane-parallel atmosphere and hydrostatic equilibrium to retrieve the depth-stratification of the temperature, density, magnetic field vector, line-of-sight velocity, as well as the macroturbulence and the percentage of stray light. This version has been designed to be applied on individual profiles. A more advanced version suitable for application on entire maps is under development.
An example application of LILIA from Socas-Navarro and Norton 2007, "The Solar Oxygen Crisis: Probably Not the Last Word", Astrophysical Journal Letters, 660, L153.
 |  |
| Maps of the temperature and the magnetic flux density at the base of the photosphere in the 3D model, as derived from the SPINOR data and LILIA inversion. Left: Temperature. In those pixels that exhibit sufficient polarization signal we show the average temperature inside and outside the magnetic concentration (weighted according to fill factor of each pixel) is depicted. The cool feature in the lower left corner is a pore that was used to stabilize the adaptive optics, allowing us to achieve the spatial resolution shown in the image. Right: Decimal logarithm of the inferred magnetic flux density in gauss. | The LILIA inferred logarithmic oxygen abundance. Left: From a LTE calculation. Right: From a Non-LTE calculation. Note that the color scales are different in each image, since the LTE abundances are systematically higher. The vertical lines are missing data due to errors in the detector system that was recording the oxygen data. |
Status: Publicly available.
Downloads:
Artificial Neural Net (ANN) Quicklook Inversion Code
Artificial Neural Networks (ANN) have been applied to the inversion of Stokes measurements. The ANN code employed by CSAC uses the Milne-Eddington approximation to compute a training set of Stokes profiles. ANN inversions require very little computational resources in terms of both processor speed and memory storage. ANN is being explored as a potential initialization step to MERLIN (see above).
An example of an ANN inversion of Hinode/SP data is shown below.
For more details on the application of ANN strategies to the inversion of Stokes SP measurements please see Socas-Navarro 2003 or Socas-Navarro 2005.
Status: Publicly available.
Downloads:
SDO/HMI Inversion Code (VFISV)
The Very Fast Inversion of the Stokes Vector (VFISV) was developed by Juan Borrero for the HMI instrument on SDO, is built for speed and (for now) will invert Stokes data from a single photospheric line (Fe I 6302.5nm) and therefore it can be applied to Hinode/SP datasets. Further information is available in the VFISV user guide (see below). VFISV assumes a standard Milne-Eddington Atmosphere to retrieve magnetic field vector, line-of-sight velocity, source function, Doppler broadening as well as the macroturbulence and the stray light filling factor.
An example of a VFISV inversion of Hinode/SP data is shown below.
Status: Publicly available.