NIRCam Point Spread Functions
Simulated JWST NIRCam point spread functions (PSFs) in each filter have predicted FWHM between 0.02" and 0.16" assuming nominal mirror wavefront errors.
On this page
At wavelengths λ > 2 μm, JWST obtains diffraction-limited imaging with a Strehl ratio = 0.8 and point spread function (PSF) FWHM of ~λ/D radians (JWST's D = 6.5 m mirror). The NIRCam detectors achieve Nyquist sampling or better (FWHM > 2 pixels) above 2 µm in the short wavelength (SW) channel (0.6–2.3 µm) and above 4 µm in the long wavelength (LW) channel (2.4–5.0 µm). Below these wavelengths, the PSF is undersampled. PSF sampling may be improved by performing subpixel dithers between exposures.
NIRCam PSFs
JWST PSFs may be simulated with WebbPSF. Given a wavefront error budget for the mirror alignment, WebbPSF can create multiple realizations of the PSF in a given NIRCam filter. Each of these realizations assumes various contributions to the wavefront error from several optical components, which cause the PSF shape and orientation to vary slightly.
WebbPSF allows the user to select an optical path difference (OPD) map, which are measured regularly and made available through WebbPSF. The figures below show one realization for NIRCam imaging in each filter assuming a representative OPD map. This is the same OPD map used for the Cycle 2 ETC. It was found, during commissioning, that this provides a close match to the empirical PSF, with only small differences (Rigby et al. 2023).
FITS images of these simulated PSFs in each filter for LW and SW channels may be downloaded from the JWST PSF simulation library.
These are a single realization of NIRCam PSF FITS images for all filters simulated by WebbPSF using a representative OPD map. OPD maps are available through WebbPSF. SW PSFs are centered in the A1 detector, while LW PSFs are centered in the A5 detector. Each FITS file contains both an oversampled PSF (oversample = 4) and a version rebinned down onto the detector pixel scale (0.031" and 0.063" in the short and long wavelength channels, respectively). The source spectrum is a G0V star spectrum from the Castelli & Kurucz model library.
Note: Diffraction spikes are oriented relative to the telescope (V3 position angle), not to NIRCam (see Figure 2 in the NIRCam Field of View page).
Figures 1 and 2 show the PSF in each filter on a log scale. Figures 3 and 4 zoom in to the PSF cores.
Click on the figure for a larger view.
NIRCam PSFs in each SW channel filter were simulated by WebbPSF using a representative OPD map. Each PSF image is 322 × 322 detector pixels (9.98" on a side). All are displayed with the same log scale. PSFs are shown in the detector pixel scale (0.031").
Click on the figure for a larger view.
NIRCam PSF in each LW channel filter were simulated by WebbPSF using a representative OPD map. Each PSF image is 159 × 159 detector pixels (10.02" on a side), displayed with the same log scaling as Figure 1. PSFs are shown in the detector pixel scale (0.063").
Click on the figure for a larger view.
These are zooms into the central 34 × 34 pixels (~1" on a side) of each simulated PSF shown in Figure 1. Each PSF is displayed with the same log scaling as Figure 1.
Click on the figure for a larger view.
These are zooms into the central 35× 35 pixels (~2" on a side) of each simulated PSF shown in Figure 2. Each PSF is displayed with the same log scaling as Figure 1.
PSF FWHM
Figure 5 shows the PSF FWHM as a function of the filter pivot wavelength, for the simulated PSFs shown in Figures 1-4. Numerical values for each FWHM, in units of arcsec and pixel, are reported in Table 1. Table 1 also includes the FWHM of the empirical PSF, measured using observations of the standard star P330E from the Cycle 1 Absolute Flux Calibration program.
Click on the figure for a larger view.
This figure shows the FWHM measured for each of the PSFs simulated by WebbPSF, shown in Figures 1–4. Each point, color-coded by wavelength, corresponds to one of the 29 NIRCam filters. The PSF is undersampled below 2 µm in the SW channel and below 4 µm in the LW channel. Note: for the extra-wide W2 filters (F150W2 and F322W2), the FWHM appears to be smaller then expected due to the choice of the star spectrum. Table 1 lists these simulated FWHM values and empirical PSF FWHM values measured from the Cycle 1 Absolute Flux calibration program.
Table 1. FWHM of both the simulated PSFs shown in Figures 1–4 and the empirical FWHM
| Simulated FWHM | Empirical FWHM | ||||
|---|---|---|---|---|---|
| Filter | Pivot wavelength (µm) | PSF FWHM (arcsec) | PSF FWHM (pixel) | PSF FWHM (arcsec) | PSF FWHM (pixel) |
| F070W | 0.704 | 0.023 | 0.742 | 0.029 | 0.935 |
| F090W | 0.901 | 0.030 | 0.968 | 0.033 | 1.065 |
| F115W | 1.154 | 0.037 | 1.194 | 0.040 | 1.290 |
| F140M | 1.404 | 0.046 | 1.484 | 0.048 | 1.548 |
| F150W | 1.501 | 0.049 | 1.581 | 0.050 | 1.613 |
| F162M | 1.626 | 0.053 | 1.710 | 0.055 | 1.774 |
| F164N | 1.644 | 0.054 | 1.742 | 0.056 | 1.806 |
| F150W2 | 1.671 | 0.045 | 1.452 | ||
| F182M | 1.845 | 0.060 | 1.935 | 0.062 | 2.000 |
| F187N | 1.874 | 0.061 | 1.968 | 0.064 | 2.065 |
| F200W | 1.990 | 0.064 | 2.065 | 0.066 | 2.129 |
| F210M | 2.093 | 0.068 | 2.194 | 0.071 | 2.290 |
| F212N | 2.120 | 0.069 | 2.226 | 0.072 | 2.323 |
| F250M | 2.503 | 0.082 | 1.302 | 0.085 | 1.349 |
| F277W | 2.786 | 0.088 | 1.397 | 0.092 | 1.460 |
| F300M | 2.996 | 0.097 | 1.540 | 0.100 | 1.587 |
| F322W2 | 3.247 | 0.096 | 1.524 | ||
| F323N | 3.237 | 0.106 | 1.683 | 0.108 | 1.714 |
| F335M | 3.365 | 0.109 | 1.730 | 0.111 | 1.762 |
| F356W | 3.563 | 0.114 | 1.810 | 0.116 | 1.841 |
| F360M | 3.621 | 0.118 | 1.873 | 0.120 | 1.905 |
| F405N | 4.055 | 0.132 | 2.095 | 0.136 | 2.159 |
| F410M | 4.092 | 0.133 | 2.111 | 0.137 | 2.175 |
| F430M | 4.280 | 0.139 | 2.206 | 0.144 | 2.286 |
| F444W | 4.421 | 0.140 | 2.222 | 0.145 | 2.302 |
| F460M | 4.624 | 0.151 | 2.397 | 0.157 | 2.492 |
| F466N | 4.654 | 0.152 | 2.413 | 0.158 | 2.508 |
| F470N | 4.707 | 0.154 | 2.444 | 0.160 | 2.540 |
| F480M | 4.834 | 0.157 | 2.492 | 0.164 | 2.603 |
Radial profiles
Figures 6 and 7 show the radial profiles for each of the simulated PSFs shown in Figures 1-4. The radial profiles have been normalized to the value of the peak pixels.
Click on the figure for a larger view.
In these plots of the radial profiles for each SW simulated PSF, the pixel size (0.031") is reported in each panel as a dotted line.
Click on the figure for a larger view.
In these plots of the radial profiles for each simulated LW PSF, the pixel size (0.063") is reported in each panel as a dotted line.
Encircled energy
Figures 8 and 9 show the encircled energy curves (the fraction of light contained in a circular aperture) for each simulated PSF in red, and the empirical PSF in blue. Table 2 lists the 50% and 80% EE values for both. Numerical values for the encircled energy as a function of radius for each simulated PSF are provided in the ASCII tables below.
Download the encircled energy measured from the simulated PSFs here:
SW Encircled Energy (.txt)
LW Encircled Energy (.txt)
Click on the figure for a larger view.
In these encircled energy curves for each SW PSF, the radius (in arcsec) at which the encircled energy is 50% and 80% is reported in Table 2. The blue dots show the empirical encircled energy measured with data from the Cycle 1 Absolute Flux calibration program, and the red line is the simulated PSF shown in Figures 1–4. These plots show an average of the encircled energy for PSFs on detectors A1 and B1. The annotated EE values are measured from the simulated (red) curve.
Click on the figure for a larger view.
In these encircled energy curves for each LW PSF, the radius (in arcsec) at which the encircled energy is 50% and 80% is reported in Table 2. The blue dots show the empirical encircled energy measured with data from the Cycle 1 Absolute Flux calibration program, and the red line is the simulated PSF shown in Figures 1–4. These plots show an average of the encircled energy for PSFs on detectors A5 and B5. The annotated EE values are measured from the simulated (red) curve.
Simulated PSF | Empirical PSF | |||
|---|---|---|---|---|
| Filter | radius (EE = 50%) | radius (EE = 80%) | radius (EE = 50%) | radius (EE = 80%) |
| F070W | 0.032 | 0.165 | 0.038 | 0.173 |
| F090W | 0.024 | 0.140 | 0.034 | 0.157 |
| F115W | 0.025 | 0.136 | 0.029 | 0.148 |
F140M | 0.030 | 0.123 | 0.031 | 0.149 |
| F150W | 0.032 | 0.121 | ||
| F150W2 | 0.034 | 0.135 | 0.033 | 0.146 |
| F162M | 0.035 | 0.121 | 0.033 | 0.141 |
| F164N | 0.035 | 0.121 | 0.034 | 0.140 |
| F182M | 0.039 | 0.129 | 0.037 | 0.138 |
| F187N | 0.040 | 0.130 | 0.038 | 0.138 |
| F200W | 0.042 | 0.138 | 0.040 | 0.144 |
| F210M | 0.044 | 0.143 | 0.042 | 0.146 |
| F212N | 0.045 | 0.145 | 0.043 | 0.147 |
| F250M | 0.049 | 0.164 | 0.049 | 0.176 |
| F277W | 0.055 | 0.179 | 0.054 | 0.192 |
| F300M | 0.061 | 0.194 | 0.059 | 0.200 |
| F322W2 | 0.062 | 0.204 | ||
| F323N | 0.069 | 0.214 | 0.064 | 0.213 |
| F335M | 0.070 | 0.216 | 0.066 | 0.221 |
| F356W | 0.073 | 0.228 | 0.069 | 0.230 |
| F360M | 0.075 | 0.232 | 0.071 | 0.234 |
| F405N | 0.084 | 0.256 | 0.078 | 0.253 |
| F410M | 0.084 | 0.258 | 0.079 | 0.258 |
| F430M | 0.088 | 0.270 | 0.083 | 0.269 |
| F444W | 0.090 | 0.276 | 0.085 | 0.276 |
| F460M | 0.095 | 0.291 | 0.090 | 0.290 |
| F466N | 0.095 | 0.292 | 0.090 | 0.292 |
| F470N | 0.097 | 0.296 | 0.091 | 0.294 |
| F480M | 0.099 | 0.303 | 0.094 | 0.301 |
References
Perrin, M., et al. 2014, Proc SPIE, 9143
Updated point spread function simulations for JWST with WebbPSF
Rigby, J., et al. 2023, PASP, 135, 048001
Characterization of JWST science performance from commissioning