2. Data types provided by IRIS

2.1. Spectrograph

The telescope feeds light into the spectrograph (SG) box which contains the Czerny–Turner spectrograph. The light from the telescope is focused on the slit assembly. The slit assembly is a prism that has a reflective coating, which also contains the slit. The reflective coating directs the light into the slit–jaw imager path. Light that goes through the slit into the prism is dispersed, directing FUV light in the 1332 – 1407 Å range and NUV light in the 2783 – 2835 Å wavelength range onto separate parts of the collimator mirror. The slit/predisperser prism assembly ensures that both FUV and NUV passbands image the same region on the Sun within the \(1/3\times175\text{arcsec}^{2}\) entrance slit. After the collimator, the FUV and NUV SG beams are fed to separate gratings, camera mirrors, and detectors. The FUV and NUV gratings, have a groove density of 3600 lines \(\text{mm}^{-1}\). The FUV and NUV SG beams have separate shutters and are recorded onto three separate CCDs – two for the FUV and one for the NUV. The two FUV CCDs observe two separate wavelength ranges: one that includes two bright C II lines (1332 – 1358 Å), and another that contains Si IV and O IV lines (1389 – 1407 Å). These two FUV CCDs are controlled by the same CEB and read out as if they were one CCD. The NUV passband from 2783 – 2835 Å is recorded by a CCD that is controlled by a different CEB (which also reads out the SJI CCD). IRIS spectral rasters are formed by scanning across the solar surface using the PZTs to change the orientation of the secondary mirror. Typical observing programs include both the FUV and NUV SG passband. The FUV and NUV spectral bandpasses cover spectral lines and continua that in the solar atmosphere are formed over a range of temperatures log \(T\) [K] = 3.7 – 7.

The brightest lines in the SG are the C II lines around 1335 Å, Si IV 1394 Å, Si IV 1403 Å, Mg II k 2796 Å, and Mg II h 2803 Å. These are the lines that are included in routine, high-cadence raster scans where exposure times are of the order of two seconds. The O IV, Fe XII, and Fe XXI lines are fainter and require longer exposure times.

2.2. Slit-jaw Imager

The slit-jaw images that are reflected off the slit/prism assembly next reach the filter wheel. The filter wheel includes six different filters (four for solar applications, and two for ground testing). The filter wheel can be rotated to place any one of the filters in the beam. The NUV filters are all transmitting, whereas the FUV filters are reflective, ensuring a different path for the NUV and FUV SJI beams. Each of these beams includes separate reimaging and fold mirrors. Both beams encounter the same shutter mechanism and are recorded on the same CCD, with one half observing the NUV SJI images, and the other half the FUV SJI images. The FUV beam includes a fixed FUV bandpass filter to block light with longer wavelengths. The NUV beam includes a Solc filter with a free spectral range of 33 Å to reduce the near-UV bandwidth to 3.6 Å. The Mg II k and Mg II wing SJI filter options are realized by combining a broader interference filter (around 15 Å) in the filterwheel with the narrow-band Solc filter.

The four solar SJI filter options are dominated by emission from, respectively, C II 1334/1335 Å, Si IV 1394/1403 Å, Mg II k 2796 Å, and the wing of Mg II around 2830 Å. The relatively broad passbands imply that contributions from continuum or wing emission are significant and, depending on solar conditions, can be dominant. Nevertheless, the bright lines are expected to contribute significantly to the SJI images. The SJI images were chosen to provide diagnostics over a wide temperature range. The C II SJI filter images may include emission from the Fe XXI line under flaring conditions. To enable solid co-alignment between the various SJI and SG channels, fiducial marks have been added to the slit. These are gaps two pixels long along the slit, one in the top half of the CCD and one in the bottom half of the CCD. These fiducial marks show up as dark features in the spectra and as bright regions in the slit portion of the SJI images.

_images/data.png

Upper row: IRIS slit-jaw images (SJI) 2830 Å, 2796 Å, 1330 Å and 1400 Å of NOAA AR 11817 taken on 14 Aug 2013 at 18:50 UT. These images are sensitive to plasma of 6000 K, 15 000 K, 30 000 and 80 000 K, respectively. They mostly sample plasma of the upper chromosphere (2796 Å) and upper photosphere (2830 Å), the upper chromosphere (1330 Å) and low transition region (1400 Å). The dark vertical line in the middle of the images is the location of the slit. Middle row: IRIS FUV 1 (left panel) and FUV 2 (right panel) spectrum of NOAA AR 11817. The two strongest lines in FUV 1 are C II lines around 1334-1336 Å formed in the upper chromosphere and low transition region. IRIS FUV 2 spectrum shows the strongest Si IV lines, formed in the transition region (65 000 K). Lower row: IRIS NUV spectrum revealing the Mg II k 2796 Å and Mg II h 2803 Å lines, both formed over a range of heights from the upper photosphere to the upper chromosphere. This wavelength range also contains a multitude of photospheric lines. The thin horizontal lines are fiducial marks that allow for easy co-alignment.

2.3. IRIS Data Level Definitions

Raw spacecraft telemetry is converted into Level 0 image files. Level 1 images are reoriented so that wavelength increases left to right.This constitutes the lowest level of scientifically-useful data, however since it is uncalibrated, Level 2 is the correct data product for most analyses.

The type of processing for data Levels beyond 1 is dependent on whether the data is from the slit-jaw imager or spectrographs. Darks and pedestal offsets are removed, and flat-fielding corrections for telescope and CCD properties are applied to generate Level 1.5 data. The data at Level 1.5 have the geometric and wavelength corrections applied and the images are mapped to a common spatial plate scale. Spectral images are remapped to align with an equal-sized array where wavelength and spatial coordinates align with the grid. An array mapping the wavelength axis to physical wavelength is created in this process. As with AIA, equivalent procedures to those used internally to transform level 1 to level 1.5 are distributed via SolarSoft as iris_prep.pro [http://sohowww.nascom.nasa.gov/solarsoft/iris/idl/lmsal/calibration/iris_prep.pro].

Levels 2 and 3 are generated from Level 1 or Level 1.5 data and are reorganized so that they can be analyzed using tools adapted from Hinode/EIS and SST/CRISP. Level 2 data consists of sets of 3D image extensions of each wavelength band stored as (λ,x,y) assembled from rasters of NUV and FUV Level 1.5 data. Level 3 data exist only for spectral rasters, and are 4D datacubes stored as (x,y,λ,t).

Note

Level 1 vs. level 2 data - The spectral data of IRIS is distinct from many contemporary observatories like SDO. IRIS Level 2 data is the equivalent to Level 1 data products of those other observatories. The Level 2 data are fully reduced, calibrated, etc. and packaged such that they are ready for further analysis. On the other hand IRIS Level 1 data MUST be passed through the calibration routines iris_prep.pro by the expert user to reach only level 1.5. The transition from level 1.5 to level 2 is a non-trivial exercise in packaging the data and while the code is available, it is currently not being supported for general use. We strongly recommend that the non-expert or casual IRIS user use the Level 2 data products. For more detail on IRIS data calibration please see Wülser et al., 2018.