Jades Data Achieves Ultra-Deep 155 Ks MIRI Imaging of the Early Universe

Scientists are fundamentally reshaping our understanding of the early Universe with new mid-infrared observations from the James Webb Space Telescope’s (JWST) Mid-Infrared Instrument (MIRI). Stacey Alberts, Daniel J. Eisenstein, and Andrew J. Bunker, alongside Qiao Duan, Kevin Hainline et al., present the latest data from the JADES program , specifically, ultra-deep MIRI imaging of the GOODS-S and GOODS-N fields. This research details a novel data reduction pipeline, optimising the removal of instrumental artefacts and delivering mosaics with exceptional sensitivity, released today as JADES Data Release 5. Crucially, these observations, combined with existing near-infrared data, will allow astronomers to characterise the first galaxies and probe the epoch of reionisation with unprecedented precision, offering vital clues to cosmic evolution.

7μm over approximately 10 arcmin² and medium-depth imaging at 7.7, 12.8, and 15μm covering 36, 25, and 22 arcmin², respectively, within the GOODS-S and GOODS-N fields. This combination provides critical rest-frame near-infrared and optical constraints on early galaxy populations, allowing scientists to probe their composition and evolution with exceptional detail. Experiments show that incorporating observations beyond 5μm with MIRI can refine estimates of stellar mass and star formation rates at high redshifts, potentially reducing previous calculations by factors of 2, 3.

The research establishes that densely-spaced multi-band NIRCam imaging can effectively mitigate these effects in typical high-redshift galaxies, ensuring robust measurements of key properties0.7μm across approximately 10 arcmin², alongside medium-depth imaging, ranging from 5 to 15 ks, at 7.7, 12.8, and 15μm over areas of 36, 25, and 22 arcmin² respectively, within the GOODS-S and GOODS-N fields. Specifically, the pipeline prioritises accurate flagging of detector anomalies and precise background estimation, essential for revealing faint, high-redshift galaxies. This method achieves a significant improvement in image quality, enabling the detection of previously unseen sources and enhancing the reliability of photometric measurements. This technique reveals crucial information about the stellar populations, star formation rates, and dust content of galaxies at redshifts greater than 4, where key spectral features shift into the mid-infrared. Experiments employ this combined dataset to refine estimates of stellar mass and star formation rates, demonstrating a reduction of 2, 3 factors in previous calculations at redshifts of 7, 9, largely due to emission line boosting in 4μm bands. The approach enables detailed analyses of high-redshift galaxies, confirming that densely-spaced multi-band NIRCam imaging can effectively mitigate potential biases in stellar property measurements, achieving consistent results with and without MIRI photometry for approximately 80% of the sample. Super background subtraction, as demonstrated in Alberts et al. (2024b), resulted in mosaics deeper than predicted by the Exposure Time Calculator (ETC), a key achievement in sensitivity. Combining contemporaneous exposures across multiple, spatially separated pointings enabled robust estimates of residual detector and sky background per pixel, mitigating shadowing effects from dither steps that were too small to fully cover extended sources. Measurements confirm that the mid-IR sky background exhibits a temporal component, necessitating a careful approach to data processing.
For datasets spanning long timescales, observations were grouped into timescales of a few months, iteratively balancing the minimisation of background variation against the benefits of combining more spatially distinct pointings. Specifically, for GS-Deep, super backgrounds were constructed separately for October 2022 and October 2023 observations, while GN-Medium was processed as a single unit0.7μm compared to those from October/November, a result further discussed in Section 3.2.3.1.3. Astrometry corrections were applied pointing-by-pointing, achieving a typical accuracy of 0.01′′ (1σ), a tenth of the MIRI native pixel size, for GS-Deep and GN/GS-Medium pointings with sufficient NIRCam overlap.

For the two GN-Medium pointings lacking NIRCam overlap, astrometry was corrected using the CHARGE HST catalog, achieving an accuracy of 0.015′′ (1σ). The final mosaics boast a pixel scale of 0.05998939′′, twice the pixel size of the NIRCam GOODS-S mosaics, oriented with North up and East to the left0.2 AB) at F770W for GS-Deep, representing a significant leap in mid-infrared detection capabilities. These measurements were obtained using circular apertures with radii of 0.42, 0.42, and 0. A new step was introduced to address artifacts caused by persistence from saturating sources.

The findings establish that deep MIRI imaging can significantly reduce the uncertainties in stellar mass and star formation rate estimates at high redshifts (z ∼7-9) by incorporating rest-frame near-infrared and optical data, mitigating the effects of emission line boosting. However, subsequent analyses have shown that densely-spaced multi-band NIRCam imaging can provide similar results to those obtained with MIRI photometry for a significant portion of high-redshift galaxies. The images reveal detailed structures such as a z ∼14 galaxy detected at F770W and compact groupings of quenched and star-forming galaxies at z ∼3.7 termed the Cosmic Rose.

While the inclusion of MIRI photometry has proven beneficial, particularly for ultra-red dusty and quiescent galaxies, the authors acknowledge that further research is needed to fully understand the implications of these findings. Future work should focus on expanding the survey areas and incorporating more diverse populations of early galaxies to provide a comprehensive view of the epoch of reionization. The limitations acknowledged by the authors include the small field-of-view (2.3 arcmin²) and large exposure time requirements, which restrict the coverage area but have yielded significant returns despite these practical constraints. The findings from this study will contribute to ongoing efforts in understanding the early Universe and the formation of galaxies during the epoch of reionization.