MIRI Coronagraphic Observations (Potential Science Use Case In Future)

One to three sentences describing what the page is about. Try to place most important information in the first 160 characters.

On this page

The JWST Mid-Infrared Instrument (MIRI) has both classical Lyot coronagraph and three four-quadrant phase mask (4QPM) coronagraphs.  The coronagraphic imaging mode provides high contrast imaging capabilities for studying faint point sources and extended emission that would otherwise be overwhelmed by a bright point-source in its vicinity.  A broader discussion on JWST coronagraphy discusses terminology and implementation strategies.  

In the case of exoplanets, MIRI coronagraphic imaging will: (1) measure effective temperatures, (2) determine bolometric luminosities, (3) measure the ammonia absorption feature (sensitive to the presence of ammonia clouds and to the atmospheric temperature), and (4) in combination with shorter wavelength measurements, support comprehensive modeling of the atmospheric properties (e.g., Bonnefoy et al. 2013). In the case of debris disks, the capability will let us: (1) explore the virtually unknown region between about 1 and 30 AU, containing for example the ice lines (not resolved with Spitzer, too low in surface brightness for HST or ground-based telescopes), (2) constrain the minerology of planetary debris, and (3) search for structures maintained by larger bodies in planetary systems.

Coronagraphic imaging can be enabled by choosing the coronagraphic imaging template in APT.  


The numbers below can provide you some guidance, but users should ultimately use the Exposure Time Calculator for all sensitivity and saturation calculations.

The coronagraph shows good rejection of the central point spread function (PSF), by a factor larger than 100 on-axis (up to a few hundreds), although the level of rejection is proba- bly limited more by the experimental limitations than by the coronagraph. The scans also showed that the inner-working angle (IWA), defined as the angle with 50% of the maximum transmission, is at 1.3λ=D for the 4QPM units.

Figure 2
Radial transmission of the MIRI coronagraphs (4QPM in red, Lyot spot in blue).
Figure 2
Contrast achieved in the test of the flight 4QPM coronagraphs (red for 10.65 μm, magenta for 11.40 μm, blue for 15.50 μm) in the flight imager. The solid lines are for the unattenuated PSF and the dashed lines show the effect of the 4QPM.

Coronagraph contrast performance 

Boccaletti et al. (2015) describe in detail two models for their analysis: Case A makes relatively optimistic assumptions about the wave front error, while Case B is more conservative.  These cases vary the waver front error (WFE) and offsetting accuracy.

Figure 3
Normalized noise-free contrasts obtained under Case A (smaller WFE, in red) and case B (larger WFE, in blue: only the reference star subtracted final result) in F1140C (left) and F1550C (right) filters on the PSF (solid) and the raw coronagraphic image (dashed). Estimated 5σ contrasts using reference star subtraction are also shown (dash-dotted).

Full system performance 

The second part of the simulation includes the detection noise.  The figure below displays four cases for various filters. Each subpanel shows the averaged contrast for the PSF (blue line) and the coronagraphic image (red solid line) as well as the 3-σ contrast for the reference star subtraction process (red dashed line) and the ideal noise level if only photon noise dominates (red dash-dotted line). Typical contrasts after postprocessing achieve 104 to 105 for separations larger than 0.5′′–1′′. Note that the radial transmissions of the MIRI coronagraphs are not accounted for in these contrast profiles. 

Figure 4
Estimated performance of MIRI in a few selected cases, assuming Case A (smaller WFE) and including the effects of noise. Colored symbols in Figure 10 stand for the intensity of Jovian planets orbiting G and M stars at 10 pc and for different assumed planetary temperatures. The radial transmission of the coronagraph for each modeled planet is highlighted with dotted lines.

The performance presented in this section is strongly related to the end-to-end optical system assumptions, such as the wavefront error and the whole optical train stability of the telescope. The simulated performance may be conservative, particularly with the development of operations and analysis techniques optimized for JWST.

The current model will have to be updated once the telescope is on orbit and commissioned. 

Maximizing contrast

To maximize contrast, most coronagraphic observations will require coronagraphic observations of a “reference star,” whose image is then subtracted from the science target image(s). Typically this will be a star that is of similar or greater brightness compared with the science target; this ensures equal or greater signal-to-noise for the reference star in a similar or shorter amount of total integration time. The procedure for obtaining these observations is identical to that for the science target.

Several techniques and algorithms for improving contrast have been developed that are more sophisticated than the classical reference star subtraction method. One is angular differential imaging (ADI), where the target and reference star are observed at multiple orientations of the observatory. Because the diffraction patterns and detector artifacts are fixed to the telescope reference frame, they move relative to the target when the observatory is rolled. Subtracting two (or more) rolled images removes, or at least reduces, these diffraction and detector features. JWST has limited roll capability (±5°), because it must keep the sunshield correctly positioned relative to the Sun, but will still provide sufficient angular roll for some science cases at large enough angular separations from the star. Larger roll angles can be achieved by observing the objects at different epochs during the year albeit with more significant changes in the wavefront error compared to back-to-back observations. Another class of enhancement is centered on sophisticated image analysis techniques, such as LOCI or KLIP (Lafreniere et al. 2007; Soummer et al. 2012).

Science use cases

Exoplanet imaging

The expected contrasts at a separation of 1′′, even for simple reference star-subtracted coronagraphy, enable detection in F1140C of 1.0 RJ planets with temperature of 500 K (respectively 350 K) orbiting a G0V (respectively M0V) star. The situation is more favorable in F1550C where a 400 K temperature is reached at 1′′ from a G0V star.

Imaging of HR 8799

For the simulations of HR 8799, we assumed an A0V star at 40pc, and 3 hr of on-source integration in each MIRI filter. The planets b, c, and d, were assigned temperatures of 900 K, 1100 K, and 1100 K with identical radii of 1.25 RJ . The corresponding contrasts in each filter, according to Allard et al. (2001), are listed in the table below.

Table 1. HR 8799 Planet's Contrast in Difference of Magnitude



The angular separations are 1.72′′, 0.96′′, 0.62′′ for planet b, c, and d. Some of the planets are close enough to the center of the 4QPM coronagraphs to be attenuated significantly. Planet b is always at a separation larger than 3λ=D in any filter so it is not impacted. Planet c is inward of 3λ=D in F1065C, F1140C, and F1550C, and hence is attenuated by 10, 15, and 30%, respectively. Similarly, planet d is attenuated by 35, 40, 40% in the same coronagraphic filters. 

Figure 1
Simulated observations of the HR 8799 system in the 4QPM coronagraphic filters, assuming Case A (smaller WFE). The planets are labeled b, c, and d. The image is in units of contrast relative to the unocculted star, as displayed by the color scale.

Planets b and c are detected at all wavelengths, while planet d is only visible at 15:5 μm. Signal-to-noise ratios measured in the brightest pixel are given in the table below.

Table 2. HR 8799 Planet's Signal-to-Noise Ratio


These numbers, if integrated over a PSF size, can be as large as 100 for planet b, for instance. The apparent elongation of planet d in the F1550C filter (in which the star residuals are fainter than the planets and the background level) is in fact due to the presence of planet e nearby. These images were used to measure the planet photometry and to compare with the initial planet intensity. In the example of planet b, at F1550C the departure of the measured intensity from the true value is larger and likely the result of residual diffraction on one side and background noise on the other side. This does not mean MIRI will fail to obtain accurate photometry in these cases but rather that a special analysis will be required. 

Figure 2
Spectral energy distribution (expressed in contrast) of the planet HR8799 b, assuming Case A (smaller WFE). The solid line represents the spec- tral model (Allard et al. 2001) degraded to a resolution of R 1⁄4 20. The red and blue points represent the true and measured intensities, respectively, in several MIRI filters (F560W, F770W, F1000W, F1065C, F1140C, F1550C). To get high quality measurements in the first three filters may require implementation of the coronagraph bar observations discussed in § 7. Horizontal bars indicate the width of filters.

The HD 181327 debris disk 

Another example of MIRI coronagraphic capabilities is given for the case of an extended object ring-like debris disk around the star HD 181327 (Schneider et al. 2006). Here, the assumed central star is a type F5V located at 50pc and observed for 1 hr in each MIRI filter. The disk model corresponds to a ∼90 AU-wide belt containing 0:05 M⊕ (up to 1 mm) of icy and porous silicates and carbonaceous dust grains in collisional equilibrium. 

The corresponding star to disk total flux ratio is 154, 48, 3.5, 0.16, respectively, in F1065C, F1140C, F1550C, F2300C filters. The 86 AU ring-like pattern is obvious in all images as well as the north/south asymmetry. The 23 μm image is sensitive to the outer part of the disk while the central cavity is blocked by the Lyot spot. Compared to the HST images obtained in the near-IR (Schneider et al. 2006) and visible (Schneider et al. 2014), the MIRI images have greatly reduced stellar residuals in the center and will allow a cleaner analysis of the dust grain properties based on mid-IR colors. In addition, the ring disk is seen close to face-on and hence would be immune to the gain brought by angular differential imaging in ground-based observations. Here, for simplicity, we assumed that the radial transmission of 4QPMs is isotropic, so we did not model the cross-like pattern of this coronagraph where the field is partially obscured. In practice, for extended objects, observations at multiple rolls (at least two at 45°) will enable “filling in” the gaps caused by the 4QPM boundaries and by the support bar in the Lyot coronagraph. 

Figure 3
Simulated observations of the HD 181327 debris disk in the coronagraphic filters, assuming Case A (smaller WFE). The image is in units of contrast relative to the unocculted star, as displayed by the color scale. The cross effect from the 4QPM is not included in this simulation.



(enter date using //)

Latest updates
  • (enter date using //)
    Briefly describe change