MSDP Data Processing at BASS2000
2/ Processings options and limitations
3/ Stokes profile html visualization
4/ LOS field and intensity html visualization
6/ Why are some data not processed ?
Two types of processed data are available in the MSDP mode :
- Stokes profiles,
- Line-of-sight velocity and magnetic fields, and intensity maps at various wavelengths in various conditions.
The Stokes profiles are sampled over 17 points in the line, and the sampling is centered at line center determined by a bissector analysis. Note that if the beam exchange has been used during the observation (for examples I + V followed by I - V in the sequence), you will find two profiles as if they were not the same parameter, as the beam exchange mode has not been selected in our standard processing.
For the LOS field and intensities several data levels are in practice available. At the "q" level, the profiles are the original ones, i.e. not resampled as a function of a zero velocity level. At the "p" level, they are resampled.
The results are organized in several files as described here :
2/ Processing options and limitations
The purpose of this work is to process a large quantity of data in order to provide maps and profiles to a large community. As a consequence, the raw data must be processed using a standard procedure, and data which do meet certain requirements are not processed or may exhibit problems.In this standard mode, we make the assumption that the scans are well documented in the headers, and that calibration files are correct. The basic options include the number of points for the sampling, informations about the telescope configuration for that year,...
After the data processing, we check that the direction of the scan (in x and y) is correct from the look of the maps. If it is not, we run the processing again, see point 6 below. However, it is quite time consuming to check for the orientation of the map (including left-right or top-bottom reversals) and the sign of the LOS magnetic field, especially when the field-of-view is small. As a consequence, you must always check these with other data (such as the mdi magnetograms on-line).
Another problem that may occur is the possibility for emission in the line (especially the Ca8542 line, and sometimes the Halpha line), which leads to wrong bissector computations and saturation effect. They are easy to notice however. To allow an estimation of the field, the sum and differences of maps on both sides of the line center are also provided for these two lines without using the bissector analysis.
More details about the code can be found here
3/ Stokes profiles html visualization
For each wavelength of the sequence and each stokes profile, 3 maps are shown : one at line center, and one on each side in the wings. The full stokes profiles are of course available in the fits files you will get after your request. Only Stokes I and; when available for 5896 λ, Stokes V profiles are shown so far, as Stokes U and V are still very noisy.
4/ LOS field and intensity html visualization
For each wavelength of the sequence, an example of intensity, LOS velocity and LOS magnetic (for Hα and 5896 λ) field maps is shown : 2 positions in the wing after the bissector computation, and at one intermediate position in the wing without the bissector computation (for Hα and 8542 λ). The full data set is available in the fits files you will get after your request.
Please note that LOS B for Halpha is usually noisy and maps should be spatially smoothed.
The input to the processing code are on-line in the prochist file :
- list of the input data filenames, including time of the beginning of each acquisition.
- a file "sequence-par " describing the sequence (including sequence number of calibration files, the date, ...)
- a file "ms.par" including many options. These can be modified at will when working on the data in more details (they will be delivered with the data as well).
- a file describing the telescope and instrument configuration for the current year.
Output, file include :
- a text file "scan.lis" with informations about the run
- difference between intensities on each side of the line center showing the quality of the cospatiality between the beams (in % of the average intensity level). The values should be small if the cospatiality is good.
Note that for 8542
λ, the cospatiality values are valid only where the bissector analysis is valid.
Other files will be delivered with the processed data, such as files describing the geometry of the raw image or intermediate result.
6/ Why are some data not processed
You may notice that some sequences are not processed while other of the same day are processed for example. In several cases the processing was interrumpted. In some cases you may be able to process the data using the available code while some data may be completely corrupted. Here is a list of problems met when processing these data :
- bad scan parameters. You can see examples of bad data and good data (respectivly left and right in the page) for a bad inveri parameter and a bad inverj parameter.
- exotic data (prominences)
- impossibility to derive the geometry of the sequence from the field stop. Most of the time this is due to a low intensity level. It may be possible to compensate for this by increasing the milgeo parameter up to 2000-3000, or tu use the geometry of another day, but it does not work all the time. When another geometry is available that day it has been used, however.
- lack of calibration files for that day
- bad data.
- data with signal lower than the dark current.
We keep a record with the problems met when processing a given sequence, so we are at your disposal if you need this information to help you to process these data.