WorksPhysics and Astronomy
Euclid Data Release 1 (DR1). Double-source-plane lens candidates
Abstract
Strong gravitational-lensing systems with multiple source planes are powerful tools for probing the stellar and dark matter content of galaxies. Lens modelling can measure the relative Einstein radii of the two source planes, which encode a ratio of angular-diameter distances determined by the lens and source redshifts and the dark energy equation of state, independently of the Hubble constant. Galaxy-scale double-source-plane lenses (DSPLs) are extremely rare. We present a sample of 84 galaxy-scale DSPL candidates in Euclid Data Release 1 (DR1), of which 80 are newly identified. These systems were selected using machine-learning lens-finding models followed by visual inspection by citizen scientists and experts. We then used lens modelling to test the DSPL interpretation. Among our lens candidates, 11 form the 'Gold' sample: each shows two sets of multiple images at different radii, and the observed configuration is reproduced by a DSPL lens model. The remaining 73 'Silver' candidates have at least two lensed arcs from visual inspection, but either lack multiple images for at least one arc, have a best-fitting DSPL distance ratio close to the single-source-plane limit, remain compatible with a single-plane model with multiple sources, or have not yet been modelled. Even if all 84 candidates are DSPLs, the sample falls below the approximately 90 systems expected from extrapolating the Euclid Quick Data Release 1 (Q1) yield and the approximately 200 obtained by scaling the LensPop forecast to the DR1 area, indicating that many DSPLs remain undetected in DR1. Spectroscopy or higher-resolution imaging is required before we can be confident that the Silver candidates are DSPLs.
Introduction
Strong gravitational lensing occurs when the path of light from a distant background source is apparently bent by the gravity of a foreground massive object that lies nearly along the same line of sight. In this situation, the source can appear as multiple images, extended arcs, or, when the alignment is almost perfect, an Einstein ring [12, 44 b]. In some lensing systems, more than one background source is located behind the same foreground lens, and these sources are at different redshifts. Such multiple source planes can produce separate sets of lensed images, often seen as Einstein rings at different angular radii. These systems are usually called compound lenses or DSPLs.
In a strong-lensing configuration, the angular scale of the Einstein ring, usually written as , is called the Einstein radius. This quantity depends on both the projected mass distribution of the lens and the relevant angular-diameter distances. For lenses with a single source plane, detailed lens modelling can constrain the local logarithmic density slope by fitting the deflection field in the region where the multiple images or arcs form [e.g. 38, 4, 30, 22]. However, such inferences are affected by systematic errors in lens modelling. One important example is the mass-sheet degeneracy [20, 33]. In this degeneracy, adding a uniform mass sheet and rescaling the convergence can change the inferred mass-profile shape and the intrinsic size of the source, while the lensing observables remain the same. In DSPL systems, the presence of a second source plane can help break this degeneracy, provided
that the mass of the first source is negligible and that the background cosmology is fixed (Bradac et al. 2004). An example image and the lensing geometry of a DSPL system are shown in Figure 1.

Figure 1. Example and geometry of a DSPL. Left: a DSPL discovered in DR1, with the , , and bands mapped to the red, green, and blue channels, respectively. Right: schematic lensing geometry showing the observer, foreground deflector at redshift , and two background sources at redshifts and . Euclid spacecraft image: ESA/C. Carreau.
The presence of two Einstein rings at different radii makes DSPLs powerful probes of lens-galaxy mass distributions. The second ring provides an additional enclosed-mass measurement, helping to disentangle the distributions of dark and luminous matter and improving sensitivity to dark matter substructure [17, 13]; studies of SDSS J0946+1006 illustrate this potential, although the claimed over-concentrated subhalo remains contested [41, 29, 1, 10, 13] (He et al. 2025). The relative Einstein radii of the two sources also constrain the DSPL distance ratio (Eq. (7)), whose lensing geometry is illustrated in Figure 1, providing a probe of the dark energy equation of state that is independent of the Hubble constant [e.g. 23, 9, 8, 26, 5]. However, a generalised mass-sheet transformation can weaken these cosmographic constraints [32], and breaking this degeneracy with stellar kinematics requires unbiased measurements and accurate assumptions about the lens mass profile, both of which are very challenging.
Galaxy-scale strong lenses with multiple background sources remain rare, because the probability of lensing two independent source planes scales approximately as , compared to for a single source plane. Among the fewer than 1000 strong lenses discovered before Euclid began its survey, fewer than ten were confirmed DSPLs, and only four of these were suitable for cosmographic analyses ([17]). To our knowledge, the following studies account for the full published sample of galaxy-scale DSPLs and DSPL candidates available before this work. Early examples include the Jackpot lens ([23]) and the Eye of Horus ([39]), while systems with sources at similar Einstein radii include J1721+8842 ([11]), the Dark Energy Survey (DES) lens DES J0408−5354 ([28]), and the Cosmic Horseshoe ([3]). Recent wide-field lens searches are now beginning to enlarge this sample. The ASTRO 3D Galaxy Evolution with Lenses (AGEL) survey reports six galaxy-scale DSPL candidates, including AGEL150745+052256 and AGEL035346−170639, which are used for DSPL cosmography ([2, 5]).
Most pre-Euclid DSPLs were initially selected by galaxy–galaxy lens-finding programmes using ground-based spectroscopy or imaging and were subsequently observed with the Hubble Space Telescope (HST), whose higher spatial resolution revealed DSPL features that were not discernible in the ground-based data. The search is more direct with Euclid: its VIS imaging has a point-spread function (PSF) full width at half maximum (FWHM) of approximately ([14]), allowing multiple-source-plane features to be recognised in the discovery data themselves, while its three near-infrared bands (, , and ) help identify additional arcs from high-redshift sources ([16]). Lens modelling can then test and support the DSPL interpretation. The Euclid Quick Data Release 1 (Q1) reports three robust galaxy-scale DSPL candidates over , with one additional object whose DSPL nature remains uncertain.
In this work, we searched for DSPLs in the covered by Euclid Data Release 1 (DR1; [19]; Euclid Collaboration: Aussel et al., in prep.; Euclid Collaboration: McCracken et al., in prep.). We used the strong-lensing discovery engine, a structured workflow that integrates machine learning with citizen-science classifications followed by expert grading, as described in the ECLIPSE catalogue and discovery-pipeline papers (Euclid Collaboration: Lines et al., in prep.; Euclid Collaboration: Holloway et al., in prep.). Candidate systems were selected during visual inspection of the highest-grade lens candidates. The single-plane lens-modelling results presented by Euclid Collaboration: Nightingale et al. (in prep.) further reveal several systems with additional arcs, which we included in our candidate list. Finally, we used the DSPL modelling described in Sect. 2 to measure and assess the DSPL nature of the modelled candidates. To our knowledge, this is the largest search for DSPLs conducted to date. In total, we identified 84 galaxy-scale DSPL candidates, of which 80 are newly identified; three are also reported in Q1 ([17]), and one, DES J0408−5354, was already known from DES ([28]). Of these, 11 form our Gold sample, for which the morphology and lens models strongly support a DSPL interpretation, although spectroscopic redshifts are still required to confirm it. The remaining 73 form our Silver sample, which additionally requires higher-resolution imaging and more detailed lens modelling to establish its DSPL nature.
The paper is organised as follows. § 2 introduces the lensing formalism and the modelling strategy used to interpret the candidates. § 3 presents the Gold and Silver DSPL samples and the lens-model results. § 4 discusses several ambiguous objects and the limitations of the current classification. § 5 compares the number of DSPLs found in DR1 with expectations from previous forecasts. Finally, § 6 summarises our conclusions. The appendices present the Gold lens-model reconstructions and list the official designations and sky coordinates of the Silver candidates.
Modelling double-source-plane lenses
Background theory
In this section, we present the theory, conventions, and notation used across the paper to describe the adopted lens-modelling techniques. For a lensed image at position , the scaled deflection angle of a lens galaxy is related to its deflection potential via
and the relation between the deflection potential and the lensing convergence is
where convergence is defined as
The convergence is the lens surface mass density normalised by the critical lensing surface density , which is
where is the angular-diameter distance to the deflector, is the angular-diameter distance to the source, is the angular-diameter distance from the deflector to the source, is the speed of light, and is Newton’s gravitational constant.
In a DSPL system, the lens equation of the first source plane can be written as
where is the scaled deflection angle of the primary deflector for source 1 at the image position , and is the unlensed angular position of source 1. The lens equation of the second source plane is then
where is the unlensed angular position of source 2, and is the deflection angle of the mass on the first source plane. Here, is the ratio of the angular-diameter distances between the different redshift planes,
where the subscripts s1 and s2 denote source 1 and source 2, respectively, and d denotes the deflector. In a simple case for which we assume that the lens galaxy has a singular isothermal sphere (SIS) profile, the parameter is just the ratio of the two Einstein radii.
We modelled the lenses using an elliptical power-law (EPL) mass model [40] with an external shear component. The convergence of the EPL model can be parametrised as
Here we choose a Cartesian coordinate system in which the - and -axes are aligned with the principal axes of the EPL mass ellipse. In this coordinate system, is the squared elliptical radius, is the axis ratio, is the power-law index of the density slope, and is the Einstein radius. We use to represent the Einstein radius of the primary lens with respect to source 1, which is the free Einstein-radius parameter of the EPL model. The Einstein radius of the same lens with respect to source 2 (with source 1 absent), , is not a free parameter: it is fixed by , , and , and serves only as a derived reference quantity. The mass of source 1 is not neglected: we model it as an SIS, whose deflection is in the lens equation of the second source plane, and we parametrise it by , the Einstein radius of source 1 for light from source 2. Both and are free parameters, with the priors listed in Table 2. The external shear approximates the tidal perturbation from mass outside the explicitly modelled deflector, such as nearby galaxies or line-of-sight structure. For the stochastic variational inference (SVI; Wingate & Weber [42]) models used in this work, we parametrised the external shear using the Cartesian components and and assigned each component an independent uniform prior. This choice is not rotationally invariant and therefore introduces a preference in the shear direction. We use it at the SVI stage because the Cartesian components vary continuously and allow us to find a good approximate lens model quickly. When the shear is sampled using its strength and position angle, the angle is periodic but the parameter interval has two boundaries. Two physically equivalent directions can then lie on opposite sides of the interval, so a gradient-based optimiser cannot move smoothly between them and may become stuck at a boundary. For rigorous posterior sampling with Markov chain Monte Carlo methods such as Hamiltonian Monte Carlo, we will instead sample the shear strength and position angle with an isotropic prior and treat the angular periodicity explicitly.
| Parameter | Prior |
Table 2. Default priors for the principal DSPL mass-model parameters. Here denotes a uniform distribution between and , and denotes a normal distribution with mean and standard deviation , truncated to .
Lens modelling strategy
We used the open-source lens-modelling code Herculens (Galan et al. [22]).1 Herculens is based on the automatic differentiation and compilation features of JAX and can run on graphics processing units (GPUs).2 Here, we briefly describe the DSPL modelling used in this work and refer readers to Enzi et al. [13] and Li et al. [17] for more detailed accounts of this approach with Herculens. The purpose of our lens modelling is to assess whether these systems are plausible DSPLs, rather than to infer their parameters precisely. The lack of redshift information for the lenses and sources prevents us from constructing robust mass models or exploiting the full statistical power of Markov chain Monte Carlo (MCMC) methods.
The data we used for the modelling came from the band because it has the highest pixel resolution (0\farcs1 pixel). We used the provided root mean square (RMS) maps and PSFs from the DR1 data products. The modelling was performed through the NumPyro implementation of SVI. We employed the AdaBelief optimiser (Zhuang et al. [43]) with a low-rank multivariate normal distribution as the guiding probability distribution. This guiding distribution is not complex enough to accurately recover the full posterior distribution, but it is computationally much cheaper than MCMC methods because it can use the automatic-differentiation capabilities of JAX (Bradbury et al. [6]). The inference was carried out by optimising the guiding distribution to minimise the Kullback–Leibler divergence to the target posterior rather than sampling from it directly. This approach is computationally fast and can recover parameter point estimates close to those obtained from MCMC. However, the posterior standard deviations returned by SVI are substantially smaller than their MCMC counterparts and are therefore overconfident; we do not interpret them as reliable parameter uncertainties. Nevertheless, the resulting best-fitting models are sufficient to assess whether the observed image configurations are consistent with a DSPL interpretation.
Since the lensed images of source 1, which form the inner Einstein ring, and any images of source 2 close to the lens centre overlap with the lens light, we modelled the lens light simultaneously with the arcs. In our model, the lens light was modelled with the sum of several elliptical Gaussian luminosity profiles with flexible centres (Shajib [34]; He et al. 2024; Enzi et al. [13]). The sources were first modelled with a single elliptical Gaussian luminosity profile to provide an approximate answer. Then, we used the parametrised lens model as the starting point and described the sources as fields that vary on regularly pixelated grids. The surface-brightness values on each pixel grid formed a vector . In Fourier space, its mode amplitudes were determined by a power spectrum (e.g., Galan et al. [21]; Rüstig et al. [31]):
where is the wave vector of a pixel on the Fourier grid, is the isotropic Matérn power spectrum (Stein [37]) evaluated at that pixel, denotes element-wise multiplication, is standard Gaussian white noise whose elements are independent samples from , and denotes the inverse Fourier transform. The default priors on the principal mass-model parameters are summarised in Table 2. The upper bound on excludes profiles steeper than , which do not give physical solutions for our systems. For some individual systems, we narrowed the prior range on or to obtain more stable lens models.
For each target, we ran between eight and ten SVI realisations, each with 20 000 iterations. Each SVI realisation had a different random initialisation, and we selected the result with the lowest average loss value as our lens model.
Lens-model results
We classified 84 DSPL candidates and divided them into two classes, Gold and Silver. The Gold sample consists of 11 systems in which the multiple images produced by each of the two sources are clearly identifiable and can be well reproduced by our lens models. Colour images of the Gold DSPL candidates are shown in Figure 2, an example lens model is shown in Figure 3, and the complete set of Gold lens-model reconstructions is shown in Appendix A. Integral-field-unit (IFU) spectroscopy of these targets would measure the redshifts of all components and is the most efficient route to confirming them as DSPLs.

Figure 2. Colour images of all 11 Gold DSPL candidates (see Sect. 3 for the Gold and Silver classification). The colour composites map the , , and bands to the red, green, and blue channels, respectively. The numbers in the upper-left corners correspond to the entries in Table 1, which are ordered by increasing right ascension. Each cutout is on a side; the scale bar and compass in the first panel indicate and the orientation (north up, east to the left), which is the same for all panels.

Figure 3. Example lens models of two Gold DSPL candidates. Each row shows a different lens system; from left to right, the panels show the observed cutout, lens-light-subtracted data, reconstructed model, normalised residuals, the separated lensed arcs from the two source planes, and the corresponding pixelated source reconstructions. The scale of all coordinate axes is arcseconds.
The Gold lens modelling results are presented in Table 1. The quoted magnitudes are measured from the -band images. For the deflector, we integrated the total flux of the best-fitting multi-Gaussian expansion (MGE) lens-light model and converted it to an AB magnitude using the -band zeropoint. For the two sources, we first subtracted the MGE lens-light model from the image and then summed the residual flux inside the corresponding arc masks used in the lens modelling. The lens photometric redshifts are Euclid-only template-fitting estimates computed with Prosphoros, the photometric-redshift code used by the Euclid photometric-redshift (PHZ) pipeline (Euclid Collaboration: Desprez et al. [15]; Euclid Collaboration: Tucci et al. [18]). We used aperture photometry in a 0\farcs5 radius aperture centred on the cutout, with the available , , , and bands. The quoted redshift uncertainties are the 16th–84th percentile interval of the one-dimensional redshift probability density function (PDF). Photometric redshifts for the full DR1 strong-lens sample, derived with a more complete treatment than the simple template fit used here, are presented by Euclid Collaboration: Nersesian et al. (in prep.). We did not attempt photometric redshifts for the sources. The arcs are resolved from the deflector only in ; in the NISP bands, whose PSF is about three times broader, they blend with the much brighter lens light.
| No. | Official name | RA (°) | Dec (°) | (″) | (″) | magnitude lens | magnitude s1 | magnitude s2 | lens | |||
| 1 | Euclid DR1 J035815.518−505648.90 | 59.5647 | −50.9469 | 0.48 | 2.15 | 2.78 | 0.66 | 0.003 | 21.14 | 22.35 | 23.13 | |
| 2 | Euclid DR1 J040450.854−580447.18 | 61.2119 | −58.0798 | 1.03 | 2.57 | 1.97 | 0.05 | 0.012 | 18.60 | 22.46 | 23.37 | |
| 3 | Euclid DR1 J040457.324−492221.54 | 61.2388 | −49.3727 | 1.16 | 2.21 | 1.18 | 0.21 | 0.082 | 19.72 | 22.05 | 23.67 | |
| 4 | Euclid DR1 J044453.064−380845.60 | 71.2211 | −38.1460 | 1.58 | 1.80 | 1.64 | 0.19 | 0.244 | 21.30 | 22.53 | 23.61 | |
| 5 | Euclid DR1 J045417.497−340950.95 | 73.5729 | −34.1642 | 0.79 | 1.63 | 1.51 | 0.16 | 0.100 | 19.80 | 21.89 | 23.79 | |
| 6 | Euclid DR1 J060447.736−364412.03 | 91.1989 | −36.7367 | 1.81 | 2.19 | 1.51 | 0.28 | 0.033 | 19.39 | 23.70 | 23.49 | |
| 7 | Euclid DR1 J115942.552+674249.68 | 179.9273 | 67.7138 | 1.43 | 1.36 | 1.13 | 0.04 | 0.105 | 16.73 | 22.29 | 21.73 | |
| 8 | Euclid DR1 J125718.567+640613.04 | 194.3274 | 64.1036 | 1.00 | 1.88 | 1.70 | 0.28 | 0.017 | 17.12 | 20.43 | 20.79 | |
| 9 | Euclid DR1 J172848.634+620851.61 | 262.2026 | 62.1477 | 0.71 | 1.65 | 1.74 | 0.16 | 0.034 | 20.79 | 22.88 | 23.27 | |
| 10 | Euclid DR1 J181422.716+670816.26 | 273.5947 | 67.1379 | 0.65 | 1.49 | 1.19 | 0.05 | 0.152 | 19.68 | 23.24 | 23.23 | |
| 11 | Euclid DR1 J181910.597+751312.57 | 274.7942 | 75.2202 | 0.93 | 2.19 | 1.21 | 0.40 | 0.120 | 20.05 | 22.52 | 21.62 |
Table 1. DR1 Gold DSPL candidates with official source designations, sky coordinates, lens-model parameters from the selected best-SVI realisation for each target, Euclid-band photometry, and available Euclid-only photometric redshifts measured in a arcsec aperture centred on the cutout. Candidates are numbered in order of increasing right ascension. The external-shear amplitude is .
Reliable arc photometry would therefore require a joint multiband model of the lens light and the sources, which is beyond the scope of this work, and photometry alone cannot constrain the source redshifts.
Some systems have a large value of (), which describes the lensing strength of source 1 on source 2. There are two reasons why this can occur. First, the values in Table 1 are the SVI medians from the selected run. SVI gives posterior standard deviations that are too small, so these values should not be treated as precise measurements. In some systems, the light
rays from source 2 pass far from source 1, so the data do not constrain the mass of source 1 well enough. The result can then depend on the SVI guide and random initialisation, and a much smaller value of may also fit the data. Second, source 1 may have a genuinely large lensing strength. Our search tends to select systems with two rings of very different radii. A large outer ring can therefore favour systems in which source 1 makes a strong contribution to the lensing of source 2.
We classify Silver DSPL candidates based on several forms of ambiguity. Nine additional Silver candidates were selected by
visual inspection and have not yet been lens-modelled. Some systems contain one complete Einstein ring and an additional lensed arc with no counter image, while others contain two lensed arcs but no clear multiple-image configuration. Robust lens modelling is more challenging in these cases because the image-plane constraints are incomplete. We also retain systems for which a single-plane lens model with multiple sources can reproduce the observed configuration, because the current data do not establish that the components lie at distinct source redshifts.
Euclid DR1 J023930.713−515748.30 is included in Silver because its best-fitting is very close to the single-source-plane limit. Three systems in our sample are also reported in Q1 ([17]), and Silver candidate 28, DES J0408−5354, was previously discovered in DES by Lin et al. [28]; the remaining 80 candidates are newly identified here. DES J0408−5354 is a DSPL very close to the single-plane limit: MUSE spectroscopy places its second source at , close to the quasar at ([35]), giving
Notes. The lens light is fitted jointly with the mass model and the sources, so where it blends with the arcs the split of flux between deflector and sources is degenerate, and the SVI posterior underestimates this uncertainty. The photometric redshifts are indicative only. Their quoted uncertainties are severely underestimated: they propagate only the background noise in the aperture and include no allowance for the PSF mismatch between and the NISP bands, for contamination of the aperture by the lensed arcs, or for template mismatch.
. The Silver colour images are shown together in Figure 4, examples of the incomplete-image and single-plane-compatible cases are shown in Figure 5 and Figure 6, respectively, and their coordinates are presented in Table A.1. Several lens candidates also include an arc that is difficult to model and difficult to explain as an astrophysical contaminant.

Figure 4. Colour images of all 73 Silver DSPL candidates, shown with the same setup as Figure 2. The colour composites map the , , and bands to the red, green, and blue channels, respectively. For panel 37, the , , and bands are instead mapped to the red, green, and blue channels, respectively. The numbers in the upper-left corners correspond to the Silver entries in Table A.1, which are ordered by increasing right ascension. All panels are approximately on a side, except panels 44 and 53, which use the native 7\farcs9 and 6\farcs7 modelling cutouts, respectively; the scale bar and compass in the first panel indicate and the orientation (north up, east to the left), which is the same for all panels.

Figure 5. Same panel layout as Figure 3, but for Silver DSPL candidate 11 (Euclid DR1 J022748.280−385955.46). At least one source does not produce a clear multiple-image configuration, so the object is kept in the Silver sample.

Figure 6. Alternative lens models for Silver candidate 47 (Euclid DR1 J053233.630−371545.77), which can be interpreted either as a single-plane lens with multiple sources or as a DSPL. The top panel shows the single-plane lens model, while the bottom panel shows the DSPL lens model. Both models reproduce the observed configuration with relatively clean residuals, so the object remains in the Silver sample.
Discussion
Several objects in our sample have ambiguous morphologies, and their physical nature remains difficult to determine from the current Euclid imaging and lens modelling alone.
Silver candidate 70, Euclid DR1 J184618.330+723826.77, is a particularly interesting system. It consists of an elliptical galaxy, a nearby satellite galaxy, and several blue arc-like features. These features have morphologies that are typical of gravitationally lensed arcs. However, the system contains nine apparent images in total, and their spatial distribution is difficult to explain with a standard strong-lensing configuration. Their colours are also very similar, making it difficult to assign the different images to distinct source-plane positions. We therefore keep this object in the Silver sample.
Silver candidate 32, Euclid DR1 J041020.748−441418.62, is another difficult lens system to model. Its inner ring can be well reproduced by a gravitational lens model, but the outer blue structure is much harder to interpret. This outer structure does not resemble a conventional spiral arm, nor does it look like a nearby galaxy undergoing tidal interaction. Because its nature remains unclear, we classify it as a Silver candidate.
Silver candidate 10, Euclid DR1 J022455.052−601235.97, is a promising disc-galaxy DSPL candidate. The image shows multiple ring-like structures, including symmetric arc-like features above and below the lens, as well as additional arc-like structures on the left and right sides of the disc core. These features are characteristic of strong gravitational lensing. However, the lens light contains a prominent dust lane that overlaps with the arcs, and the outer arc is very faint. As a result, our DSPL models for this system are not stable: different SVI chains converge to different configurations. Given this uncertainty, we classify this object as a Silver lens candidate. Deeper imaging at higher angular resolution, for example with HST or JWST, would be valuable for this system.
Silver candidate 47, Euclid DR1 J053233.630−371545.77, features a central galaxy with six different images. We performed both DSPL and single-plane lens models. The modelling results are shown in Figure 6. Both lens models can reproduce the configuration of the data with relatively clean residuals. The single-plane model has an Einstein radius of . The DSPL model has an Einstein radius of the first ring of and .
Limitations and comparison to expectations
Our Q1 study (Euclid Collaboration: Li et al. [17]) used the LensPop code (Collett [7]) to forecast that Euclid will discover approximately 1700 galaxy-scale DSPLs over the full survey area. Scaling this prediction to the 1900 deg covered by DR1 gives an expectation of about 200 DSPLs. Alternatively, extrapolating from the three robust DSPL candidates identified in Q1 leads to an expectation of about 90 DSPL systems in DR1. Both numbers are substantially higher than the number of Gold DSPL candidates found in this work. Even if all of the Silver DSPL candidates are real DSPLs, there remains a large shortfall relative to the forecasts. We attribute the difference to two effects: the limitations of our search strategy and the definition of a DSPL. Figure 7 compares the distribution of in our Gold lens sample with this LensPop forecast. The forecast sample is truncated at , the upper bound of our prior (Table 2). Only 13 of the 1729 forecast systems (0.8%) have . In these systems source 1 lies just behind the deflector (median ), so the inner Einstein radius is very small (median ), close to the VIS PSF and embedded in the lens light, and such systems are unlikely to be recognised in Euclid imaging. The median of the Gold sample is 1.51, compared with 1.18 for the LensPop forecast, and 8 of the 11 Gold systems (73%) have , whereas only 47% of the forecast systems do. Because our candidates were selected by visual inspection, our selection naturally favours large : when is close to unity, the two Einstein rings nearly coincide and blend together at the Euclid resolution, as in Silver candidates 9 and 60 (Figure 4). Such systems are either classified as Silver or missed by our visual inspection.
Figure 7. Distribution of inferred from the Gold DSPL lens models using the minimum-loss SVI realisation for each target, compared with the LensPop forecast of Euclid Collaboration: Li et al. [17] after removing systems with . The LensPop histogram is normalised to the total number of Gold systems, while the Gold histogram shows the observed counts from this work. The red ticks along the horizontal axis indicate the individual measurements for the Gold systems.
Our search is limited in that we did not systematically search for DSPLs: we did not train our machine-learning classifiers to identify DSPLs. Instead, we visually inspected the best-ranked strong-lens candidates identified by the DR1 strong-lensing finding networks (Euclid Collaboration: Vincken et al., in prep.; Euclid Collaboration: Fogliardi et al., in prep.) and prioritised through the ECLIPSE discovery pipeline (Euclid Collaboration: Holloway et al., in prep.). We do not know how many DSPLs are ranked below these visual-inspection thresholds. Additionally, we primarily searched for DSPLs in the same colour-composite images, so faint second sources are likely to have been missed. Whilst some DSPL candidates are identified in lens-light-subtracted residuals from the mass models of Euclid Collaboration: Nightingale et al. (in prep.), we did not systematically calibrate this. Another family of missing DSPL should be systems with source that lacks VIS image as they are at high redshift or dusty.
The DSPL definition adopted in LensPop is based on the length of the arc: a source galaxy is classified as lensed if the arc length is larger than . However, this does not necessarily imply the formation of a bright and obvious counter image. Several of our Silver DSPL candidates fall into this category: the second source is likely substantially magnified, but it does not form a clear multiple-image configuration. For these systems, it is difficult to use the second source plane to provide additional constraints on the lens model. As discussed by Li et al. [17], the key advantage of a DSPL is that the second Einstein ring provides an additional enclosed-mass measurement, which improves constraints on the radial mass profile. If the second source does not form multiple images, however, constraints on the deflection field are much weaker and almost all information about the lens mass profile comes from the first complete Einstein ring. Furthermore, because the first source plane may itself contain non-negligible mass, the intrinsic morphology of the second source cannot be robustly constrained without a clear multiple-image configuration.
Conclusions
We present 84 galaxy-scale DSPL candidates in DR1, consisting of 11 Gold and 73 Silver systems; 80 of these candidates are newly identified. This substantially enlarges the published sample of galaxy-scale DSPL candidates. The detailed lens modelling shows that the modelled candidates can be explained by a DSPL configuration, although some Silver systems are also compatible with a single-plane interpretation. We performed lens modelling and attempted to derive for the candidates with available models; nine additional Silver candidates remain to be modelled. For the Silver candidates, higher-resolution imaging and spectroscopic redshifts are needed for a definitive conclusion about their nature.
The shortfall relative to the forecasts shows that many DSPLs remain undetected in DR1, either because they were assigned low rankings by the single-source-plane lens-finding classifiers or because they were missed during visual inspection. It is clear that piggybacking on the efforts to find galaxy–galaxy lenses in Euclid can never provide a high-completeness DSPL search: if Euclid is to reach its full DSPL discovery potential, then a dedicated search is needed, including training DSPL-specific machine-learning finders, automated lens modelling of the candidates, and teaching citizen scientists to look out for them.
The 84 candidates form a valuable sample from which to select individual high-value systems for follow-up. In each such system, the additional radial constraints from a second source plane improve the separation of the stellar and dark matter components and allow assumptions about the stellar initial mass function (IMF) to be tested. For the Jackpot DSPL, Li et al. [17] constrain the IMF mismatch parameter from lensing alone to , sufficient to distinguish its Salpeter-like IMF from a Chabrier IMF at approximately . A sample of around 50 well-constrained DSPLs spanning a broad redshift range could therefore test whether the IMF normalisation is universal and whether it evolves with redshift. For comparison, Project Dinos II combines strong lensing and stellar dynamics for 56 lenses and constrains the redshift-dependence parameter of the dark matter inner slope to [36]. A similarly sized DSPL sample could test the evolution of dark matter density profiles using the additional lensing constraints from the second source plane. DSPLs can also probe the mass associated with source 1 through its lensing effect on source 2, providing information about lower-mass galaxies at higher redshift.
Acknowledgements. This project and WJRE have received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (LensEra: grant agreement No 945536). TEC is funded by the Royal Society through a University Research Fellowship. WJRE acknowledges the support by the Italian Ministry of University and Research (grant FIS2023-01611, CUP C53C25000300001). NL is supported through a graduate studentship from the UKRI STFC and the University of Portsmouth, and acknowledges travel support from STFC for UK participation in the Rubin LSST through grant ST/X001334/1. PH is funded by the Royal Society (RF\ERE\210033). JWN is supported by an STFC/UKRI Ernest Rutherford Fellowship, Project Reference: ST/X003086/1. NL, PH, and JWN acknowledge travel support provided by Euclid:UK through STFC grant APP91126. AG acknowledges support by the SNSF through PostDoc Mobility Return grant P5_2_235360. AS acknowledges support from the National Key R&D Program of China (No. 2023YFA1607800, 2023YFA1607802), and from the Yangyang Development Fund. Numerical computations were done on the Sciama High Performance Compute (HPC) cluster which is supported by the ICG, SEPNet and the University of Portsmouth. This work has made use of the Euclid Data Release 1 (DR1) from the European Space Agency (ESA) mission Euclid, available at https://www.cosmos.esa.int/web/euclid/data-release-dr1. The Euclid Consortium acknowledges the European Space Agency and a number of agencies and institutes that have supported the development of Euclid, in particular the Agenzia Spaziale Italiana, the Austrian Forschungsförderungsgesellschaft funded through BMIMI, the Belgian Science Policy, the Canadian Euclid Consortium, the Deutsches Zentrum für Luft- und Raumfahrt, the DTU Space and the Niels Bohr Institute in Denmark, the French Centre National d’Etudes Spatiales, the Fundação para a Ciência e a Tecnologia, the Hungarian Academy of Sciences, the Ministerio de Ciencia, Innovación y Universidades, the National Aeronautics and Space Administration, the National Astronomical Observatory of Japan, the Nederlandse Onderzoekschool Voor Astronomie, the Norwegian Space Agency, the Research Council of Finland, the Romanian Space Agency, the Swiss Space Office (SSO) at the State Secretariat for Education, Research, and Innovation (SERI), and the United Kingdom Space Agency. A complete and detailed list is available on the Euclid web site (www.euclid-ec.org/consortium/community/).
Authors and affiliations
Euclid Collaboration: T. Li⋆1, N. E. P. Lines1, P. Holloway1, T. E. Collett1, J. W. Nightingale2, S. H. Vincken1, 3, 4, D. J. Bowden5, S. Schuldt6, 7, 8, M. Fogliardi9, 10, M. Grespan5, 11, A. Verma5, C. Krawczyk1, L. R. Ecker12, 13, 14, W. J. R. Enzi1, 15, G. Despali16, 10, 17, A. Galan18, A. Sonnenfeld19, 20, 21, F. Courbin22, 23, 24, R. B. Metcalf16, 10, G. Mahler25, R. J. Massey26, B. Altieri27, S. Andreon28, N. Auricchio10, C. Baccigalupi29, 30, 31, 32, M. Baldi33, 10, 17, S. Bardelli10, P. Battaglia10, R. Bender13, 12, A. Biviano30, 29, E. Branchini34, 35, 28, M. Brescia36, 37, J. Brinchmann38, 39, 40, S. Camera41, 42, 43, G. Cañas-Herrera44, V. Capobianco43, C. Carbone8, J. Carretero45, 46, M. Castellano47, G. Castignani10, S. Cavuoti37, 48, A. Cimatti49, C. Colodro-Conde50, G. Congedo51, C. J. Conselice52, L. Conversi53, 27, Y. Copin54, H. M. Courtois55, M. Cropper56, A. Da Silva57, 58, H. Degaudenzi18, G. De Lucia30, C. Dolding56, H. Dole59, 60, F. Dubath18, X. Dupac27, S. Escoffier61, M. Fabricius13, 12, M. Farina62, R. Farinelli10, F. Faustini47, S. Ferriol54, M. Frailis30, E. Franceschi10, M. Fumana8, L. Gabarra5, S. Galeotta30, K. George63, C. Giocoli10, 17, J. Gracia-Carpio13, A. Grazian64, F. Grupp13, 12, L. Guzzo65, 28, 66, W. G. Hartley18, S. V. H. Haugan67, S. Hemmati68, H. Hoekstra44, W. Holmes69, I. M. Hook70, F. Hormuth71, A. Hornstrup72, 73, K. Jahnke74, M. Jhabvala75, A. Kiessling69, B. Kubik54, M. Kümmel12, M. Kunz76, H. Kurki-Suonio7, 77, A. M. C. Le Brun78, P. Liebing56, S. Ligori43, P. B. Lilje67, V. Lindholm7, 77, I. Lloro79, M. Magliocchetti62, G. Mainetti80, O. Marggraf81, M. Martinelli47, 82, N. Martinet83, F. Marulli16, 10, 17, N. Mauri49, 17, E. Medinaceli10, M. Melchior4, M. Meneghetti10, 17, G. Meylan84, P. Monaco85, 30, 31, 29, M. Moresco16, 10, C. Moretti30, 29, 31, L. Moscardini16, 10, 17, R. Nakajima81, C. Neissner86, 46, R. C. Nichol87, S.-M. Niemi88, C. Padilla86, S. Paltani18, F. Pasian30, W. J. Percival89, 90, 91, V. Pettorino88, S. Pires92, G. Polenta93, M. Poncet94, G. D. Racca44, 88, F. Raison13, A. Renzi95, 96, 10, J. Rhodes69, G. Riccio37, I. Risso34, 35, 28, F. Rizzo30, E. Romelli30, M. Roncarelli10, B. Rusholme68, R. Saglia12, 13, Z. Sakr97, 98, 99, D. Sapone100, P. Schneider81, A. Secroun61, E. Sihvola7, 77, P. Simon81, C. Sirignano95, 96, G. Sirri17, L. Stanco96, P. Tallada-Crespí45, 46, A. N. Taylor51, H. I. Teplitz101, I. Tereno57, 58, S. Toft102, 103, R. Toledo-Moreo104, 105, F. Torradeflot46, 45, I. Tutusaus106, 107, 98, J. Valivita7, 77, T. Vassallo30, 63, A. Veropalumbo28, 35, 34, Y. Wang68, J. Weller12, 13, A. Zacchei30, 29, G. Zamorani10, F. M. Zerbi28, E. Zucca8, I. T. Andika12, M. Ballardini9, 108, 10, P. Bergamini10, S. Borgani85, 29, 30, 31, E. Bozzo15, 110, C. Burigana111, 8, F. Caro47, D. Di Ferdinando17, J. A. Escartin Vigo13, J. García-Bellido97, T. Gasparetto47, C. M. Gutierrez50, 112, A. Hall51, Y. Kang113, A. Loureiro114, 115, R. Maoli116, 47, S. Nadathur1, M. Pöntinen7, E. Sarpa30, V. Scottez117, 118, M. Sereno10, 17, K. Tanidis119, M. Tenti17, A. Troja30, M. Tucci18, M. Viel29, 30, 32, 31, 109, M. Wiesmann67, Abdurro’uf120, J. A. Acevedo Barroso69, Y. Akrami97, 121, G. Angora37, 9, M. Archidiacono65, 66, G. Aricò17, F. Atrio-Barandela122, M. Aubert54, M. Baes123, N. Bartolo95, 96, 64, L. Bazzanini5, 9, J. R. Bermejo-Climent124, 125, D. Bertacca95, 64, M. Bethermin2, F. Beutler51, A. Blanchard98, F. Bouché1, M. L. Brown52, S. Bruton47, A. Calabrese47, C. S. Carvalho129, N. Chartab68, F. Cogato10, P. Colodro-Caballero130, M. Correntí47, 93, C. Brusi113, O. Cucciati10, T. de Boer131, F. De Paolis132, 133, 134, G. Desprez130, A. Díaz-Sánchez135, S. Di Domizio34, 35, J. M. Diego136, P.-A. Duc126, V. Duret137, Y. Fang12, A. Farina28, 35, A. Finoguenov7, A. Fontana47, A. Franco132, 133, 134, N. Frusciante138, 139, Y. Fu44, 130, K. Ganga140, R. Gavazzi83, 141, E. Gaztanaga106, 107, 1, Z. Ghaffari30, 29, F. Giacomini17, F. Gianotti10, G. Gozaliasl142, 7, A. Gruppuso10, 17, M. Guidi33, 10, C. Hernández-Monteagudo112, 50, T. Heuberger143, J. Hjorth144, M. Jauzac145, 26, 146, 147, 148, S. Joudaki45, J. J. E. Kajava148, 6, 149, D. Karagiannis150, 9, 151, J. Kim5, C. C. Kirkpatrick152, A. Kovács125, 124, I. Kovac̆ić123, K. Koyama1, S. Kruk1, M. C. Lam51, S. Lee153, 69, M. D. Lepine30, G. F. Lesci15, J. Lesgourgues154, T. I. Liaudat155, L. Linke113, S. J. Liu62, E. Magnier131, A. Manjón-García135, F. Mannucci156, F. R. Marleau113, C. J. A. P. Martins157, 39, M. Miluzio27, 158, G. Morgante10, C. Murray92, K. Naidoo1, 74, P. Natoli9, 108, A. Navarro-Alsina81, S. Nesseris97, L. Nicastro10, M. Oguri159, 160, F. Oppizzi35, 96, 95, F. Pace41, 42, 43, 58, D. Paoletti10, 110, K. Paterson74, C. Pattison1, L. Pizzuti161, D. Potter162, G. W. Pratt2, S. Quai16, 10, M. Radovich64, F. Ricci163, 47, K. Rojas143, W. Rosten13, M. Sahlén164, D. B. Sanders131, A. Schneider162, D. Sciotti47, 82, E. Sellentin165, S. Serjeant1, F. Shankar166, L. C. Smith131, I. Szapudi131, G. Testera35, R. Teyssier169, S. Tosi34, 28, 35, C. Uhlmann170, C. Valieri17, A. Venhola171, D. Vergani10, G. Vernardos172, 173, S. Vinciguerra83, M. von Wietersheim-Kramsta26, 145, N. A. Walton168, and A. H. Wright174
1 Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth PO1 3FX, UK
2 School of Mathematics, Statistics and Physics, Newcastle University, Herschel Building, Newcastle-upon-Tyne, NE1 7RU, UK
3 Fontys University of Applied Sciences, School of ICT, Rachelsmolen 1, 5612 MA Eindhoven, The Netherlands
4 University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland
5 Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, UK
6 Finnish Centre for Astronomy with ESO (FINCA), Quantum, Vesilinnantie 5, University of Turku, 20014 Turku, Finland
7 Department of Physics, P.O. Box 64, University of Helsinki, 00014 Helsinki, Finland
8 INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy
9 Dipartimento di Fisica e Scienze della Terra, Università degli Studi di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy
10 INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy
11 National Centre for Nuclear Research, ul. Pasteura 7, 02-093, Warsaw, Poland
12 Universitäts-Sternwarte München, Fakultät für Physik, Ludwig-Maximilians-Universität München, Scheinerstr. 1, 81679 München, Germany
13 Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany
14 Max-Planck-Institut für Astrophysik, Karl-Schwarzschild-Str. 1, 85748 Garching, Germany
15 INAF, Istituto di Radioastronomia, Via Piero Gobetti 101, 40129 Bologna, Italy
16 Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, via Piero Gobetti 93/2, 40129 Bologna, Italy
17 INFN-Sezione di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy
18 Department of Astronomy, University of Geneva, ch. d’Ecogia 16, 1290 Versoix, Switzerland
19 Department of Astronomy, School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China
20 State Key Laboratory of Dark Matter Physics, School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China
21 Shanghai Key Laboratory for Particle Physics and Cosmology, Shanghai 200240, China
22 Institut de Ciències del Cosmos (ICCUB), Universitat de Barcelona (IEEC-UB), Martí i Franquès 1, 08028 Barcelona, Spain
23 Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig de Lluís Companys 23, 08010 Barcelona, Spain
24 Institut de Ciencies de l’Espai (IEEC-CSIC), Campus UAB, Carrer de Can Magrans, s/n Cerdanyola del Vallés, 08193 Barcelona, Spain
25 STAR Institute, University of Liège, Quartier Agora, Allée du six Août 19c, 4000 Liège, Belgium
26 Department of Physics, Institute for Computational Cosmology, Durham University, South Road, Durham, DH1 3LE, UK
27 ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain
28 INAF-Osservatorio Astronomico di Brera, Via Brera 28, 20122 Milano, Italy
29 IFPU, Institute for Fundamental Physics of the Universe, via Beirut 2, 34151 Trieste, Italy
30 INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy
31 INFN, Sezione di Trieste, Via Valerio 2, 34127 Trieste TS, Italy
32 SISSA, International School for Advanced Studies, Via Bonomea 265, 34136 Trieste TS, Italy
33 Dipartimento di Fisica e Astronomia, Università di Bologna, Via Gobetti 93/2, 40129 Bologna, Italy
34 Dipartimento di Fisica, Università di Genova, Via Dodecaneso 33, 16146, Genova, Italy
35 INFN-Sezione di Genova, Via Dodecaneso 33, 16146, Genova, Italy
36 Department of Physics "E. Pancini", University Federico II, Via Cinthia 6, 80126, Napoli, Italy
37 INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy
38 European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany
39 Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal
40 Faculdade de Ciências da Universidade do Porto, Rua do Campo de Alegre, 4150-007 Porto, Portugal
41 Dipartimento di Fisica, Università degli Studi di Torino, Via P. Giuria 1, 10125 Torino, Italy
42 INFN-Sezione di Torino, Via P. Giuria 1, 10125 Torino, Italy
43 INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese (TO), Italy
44 Leiden Observatory, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands
45 Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Avenida Complutense 40, 28040 Madrid, Spain
46 Port d’Informació Científica, Campus UAB, C. Albareda s/n, 08193 Bellaterra (Barcelona), Spain
47 INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy
48 INFN – Sezione di Napoli, Via Cinthia 6, 80126, Napoli, Italy
49 Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy
50 Instituto de Astrofísica de Canarias, E-38205 La Laguna, Tenerife, Spain
51 Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK
52 Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK
53 European Space Agency/ESRIN, Largo Galileo Galilei 1, 00044 Frascati, Roma, Italy
54 Université Lyon 1, CNRS/IN2P3, IP2I Lyon, Villeurbanne, France
55 Université Lyon 1, CNRS/IN2P3, IUF, IP2I Lyon, Villeurbanne, France
56 Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK
57 Departamento de Física, Faculdade de Ciências, Universidade de Lisboa, Edifício C8, Campo Grande, PT1749-016 Lisboa, Portugal
58 Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal
59 Université Paris-Saclay, CNRS, Institut d’astrophysique spatiale, 91405, Orsay, France
60 Institut universitaire de France (IUF), 1 rue Descartes, 75231 PARIS CEDEX 05, France
61 Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France
62 INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy
63 University Observatory, LMU Faculty of Physics, Scheinerstr. 1, 81679 Munich, Germany
64 INAF-Osservatorio Astronomico di Padova, Via dell’Osservatorio 5, 35122 Padova, Italy
65 Dipartimento di Fisica "Aldo Pontremoli", Università degli Studi di Milano, Via Celoria 16, 20133 Milano, Italy
66 INFN-Sezione di Milano, Via Celoria 16, 20133 Milano, Italy
67 Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway
68 Caltech/IPAC, 1200 E. California Blvd., Pasadena, CA 91125, USA
69 Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA
70 School of Physics and Astronomy, Lancaster University, Lancaster, LA1 4YB, UK
71 Felix Hormuth Engineering, Goethestr. 17, 69181 Leimen, Germany
72 Technical University of Denmark, Elektrovej 327, 2800 Kgs. Lyngby, Denmark
73 Cosmic Dawn Center (DAWN), Denmark
74 Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany
75 NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA
76 Université de Genève, Département de Physique Théorique and Centre for Astroparticle Physics, 24 quai Ernest-Ansermet, CH-1211 Genève 4, Switzerland
77 Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, 00014 Helsinki, Finland
78 Laboratoire d’etude de l’Univers et des phenomenes eXtremes, Observatoire de Paris, Université PSL, Sorbonne Université, CNRS, 92190 Meudon, France
79 SKAO, Jodrell Bank, Lower Withington, Macclesfield SK11 9FT, UK
80 Centre de Calcul de l’IN2P3/CNRS, 21 avenue Pierre de Coubertin 69627 Villeurbanne Cedex, France
81 Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany
82 INFN-Sezione di Roma, Piazzale Aldo Moro, 2 - c/o Dipartimento di Fisica, Edificio G. Marconi, 00185 Roma, Italy
83 Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France
84 Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Observatoire de Sauverny, 1290 Versoix, Switzerland
85 Dipartimento di Fisica - Sezione di Astronomia, Università di Trieste, Via Tiepolo 11, 34131 Trieste, Italy
86 Institut de Física d’Altes Energies (IFAE), The Barcelona Institute of Science and Technology, Campus UAB, 08193 Bellaterra (Barcelona), Spain
87 School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK
88 European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands
89 Waterloo Centre for Astrophysics, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada
90 Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada
91 Perimeter Institute for Theoretical Physics, Waterloo, Ontario N2L 2Y5, Canada
92 Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France
93 Space Science Data Center, Italian Space Agency, via del Politecnico snc, 00133 Roma, Italy
94 Centre National d’Etudes Spatiales – Centre spatial de Toulouse, 18 avenue Edouard Belin, 31401 Toulouse Cedex 9, France
95 Dipartimento di Fisica e Astronomia "G. Galilei", Università di Padova, Via Marzolo 8, 35131 Padova, Italy
96 INFN-Padova, Via Marzolo 8, 35131 Padova, Italy
97 Instituto de Física Teórica UAM-CSIC, Campus de Cantoblanco, 28049 Madrid, Spain
98 Institut de Recherche en Astrophysique et Planétologie (IRAP), Université de Toulouse, CNRS, UPS, CNES, 14 Av. Edouard Belin, 31400 Toulouse, France
99 Université St Joseph; Faculty of Sciences, Beirut, Lebanon
100 Departamento de Física, FCFM, Universidad de Chile, Blanco Encalada 2008, Santiago, Chile
101 Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA
102 Cosmic Dawn Center (DAWN)
103 Niels Bohr Institute, University of Copenhagen, Jagtvej 128, 2200 Copenhagen, Denmark
104 Universidad Politécnica de Cartagena, Departamento de Electrónica y Tecnología de Computadoras, Plaza del Hospital 1, 30202 Cartagena, Spain
105 European University of Technology EUt+, European Union
106 Institute of Space Sciences (ICE, CSIC), Campus UAB, Carrer de Can Magrans, s/n, 08193 Barcelona, Spain
107 Institut d’Estudis Espacials de Catalunya (IEEC), Edifici RDIT, Campus UPC, 08860 Castelldefels, Barcelona, Spain
108 Istituto Nazionale di Fisica Nucleare, Sezione di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy
109 ICSC - Centro Nazionale di Ricerca in High Performance Computing, Big Data e Quantum Computing, Via Magnanelli 2, Bologna, Italy
110 INFN-Bologna, Via Irnerio 46, 40126 Bologna, Italy
111 Astronomical Observatory of the Autonomous Region of the Aosta Valley (OAVdA), Loc. Lignan 39, I-11020, Nus (Aosta Valley), Italy
112 Universidad de La Laguna, Dpto. Astrofísica, E-38206 La Laguna, Tenerife, Spain
113 Universität Innsbruck, Institut für Astro- und Teilchenphysik, Technikerstr. 25/8, 6020 Innsbruck, Austria
114 Oskar Klein Centre for Cosmoparticle Physics, Department of Physics, Stockholm University, Stockholm, SE-106 91, Sweden
115 Astrophysics Group, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK
116 Dipartimento di Fisica, Sapienza Università di Roma, Piazzale Aldo Moro 2, 00185 Roma, Italy
117 Institut d’Astrophysique de Paris, 98bis Boulevard Arago, 75014, Paris, France
118 ICL, Junia, Université Catholique de Lille, LITL, 59000 Lille, France
119 Center for Astrophysics and Cosmology, University of Nova Gorica, Nova Gorica, Slovenia
120 Department of Astronomy, Indiana University, 727 East Third Street, Bloomington, IN 47405, USA
121 CERCA/ISO, Department of Physics, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA
122 Departamento de Física Fundamental. Universidad de Salamanca. Plaza de la Merced s/n. 37008 Salamanca, Spain
123 Universiteit Gent, Department of Physics and Astronomy, Proeftuinstraat 86 N3, 9000 Ghent, Belgium
124 Konkoly Observatory, HUN-REN CSFK, MTA Centre of Excellence, Budapest, Konkoly Thege Miklós út 15-17, H-1121, Hungary
125 MTA-CSFK Lendület Large-Scale Structure Research Group, Konkoly-Thege Miklós út 15-17, H-1121 Budapest, Hungary
126 Université de Strasbourg, CNRS, Observatoire astronomique de Strasbourg, UMR 7550, 67000 Strasbourg, France
127 Scuola Superiore Meridionale, Via Mezzocannone 4, 80138, Napoli, Italy
128 California Institute of Technology, 1200 E California Blvd, Pasadena, CA 91125, USA
129 Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Tapada da Ajuda, 1349-018 Lisboa, Portugal
130 Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands
131 Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA
132 Department of Mathematics and Physics E. De Giorgi, University of Salento, Via per Arnesano, CP-193, 73100, Lecce, Italy
133 INFN, Sezione di Lecce, Via per Arnesano, CP-193, 73100, Lecce, Italy
134 INAF-Sezione di Lecce, c/o Dipartimento Matematica e Fisica, Via per Arnesano, 73100, Lecce, Italy
135 Departamento Física Aplicada, Universidad Politécnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain
136 Instituto de Física de Cantabria, Edificio Juan Jordá, Avenida de los Castros, 39005 Santander, Spain
137 Univ. Grenoble Alpes, CNRS, IPAG, 38000 Grenoble, France
138 Dipartimento di Fisica "E. Pancini", Universita degli Studi di Napoli Federico II, Via Cintia 6, 80126, Napoli, Italy {sup}139` INFN Sezione di Napoli, Universita degli Studi di Napoli "Federico II", Compl. Univ. di Monte S. Angelo, Edificio G, Via Cinthia, I-80126, Napoli, Italy
{sup}140` Université Paris Cité, CNRS, Astroparticule et Cosmologie, 75013 Paris, France
{sup}141` Institut d’Astrophysique de Paris, UMR 7095, CNRS, and Sorbonne Université, 98 bis boulevard Arago, 75014 Paris, France
{sup}142` Department of Computer Science, Aalto University, PO Box 15400, Espoo, FI-00 076, Finland
{sup}143` University of Applied Sciences and Arts of Northwestern Switzerland, School of Computer Science, 5210 Windisch, Switzerland
{sup}144` DARK, Niels Bohr Institute, University of Copenhagen, Jagtvej 155, 2200 Copenhagen, Denmark
{sup}145` Department of Physics, Centre for Extragalactic Astronomy, Durham University, South Road, Durham, DH1 3LE, UK
{sup}146` Astrophysics Research Centre, University of KwaZulu-Natal, Westville Campus, Durban 4041, South Africa
{sup}147` School of Mathematics, Statistics & Computer Science, University of KwaZulu-Natal, Westville Campus, Durban 4041, South Africa
{sup}148` Department of Physics and Astronomy, Vesilinnantie 5, University of Turku, 20014 Turku, Finland
{sup}149` Serco for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain
{sup}150` Van Swinderen Institute, University of Groningen, Nijenborgh 3, 9747 AG Groningen, The Netherlands
{sup}151` Department of Physics and Astronomy, University of the Western Cape, Bellville, Cape Town, 7535, South Africa
{sup}152` Department of Physics and Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, 00014 Helsinki, Finland
{sup}153` Ohio University, Physics & Astronomy Department,1 Ohio University, Athens, OH 45701, USA
{sup}154` Institute for Theoretical Particle Physics and Cosmology (TTK), RWTH Aachen University, 52056 Aachen, Germany
{sup}155` IRFU, CEA, Université Paris-Saclay 91191 Gif-sur-Yvette Cedex, France
{sup}156` INAF-Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125, Firenze, Italy
{sup}157` Centro de Astrofísica da Universidade do Porto, Rua das Estrelas, 4150-762 Porto, Portugal
{sup}158` HE Space for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain
{sup}159` Center for Frontier Science, Chiba University, 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan
{sup}160` Department of Physics, Graduate School of Science, Chiba University, 1-33 Yayoi-Cho, Inage-Ku, Chiba 263-8522, Japan
{sup}161` Dipartimento di Fisica “G. Occhialini”, Università degli Studi di Milano Bicocca, Piazza della Scienza 3, 20126 Milano, Italy
{sup}162` Department of Astrophysics, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland
{sup}163` Department of Mathematics and Physics, Roma Tre University, Via della Vasca Navale 84, 00146 Rome, Italy
{sup}164` Theoretical astrophysics, Department of Physics and Astronomy, Uppsala University, Box 524, 751 20 Uppsala, Sweden
{sup}165` Mathematical Institute, University of Leiden, Einsteinweg 55, 2333 CA Leiden, The Netherlands
{sup}166` School of Physical Sciences, The Open University, Milton Keynes, MK7 6AA, UK
{sup}167` School of Physics & Astronomy, University of Southampton, Highfield Campus, Southampton SO17 1BJ, UK
{sup}168` Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK
{sup}169` Department of Astrophysical Sciences, Peyton Hall, Princeton University, Princeton, NJ 08544, USA
{sup}170` Fakultät für Physik, Universität Bielefeld, Postfach 100131, 33501 Bielefeld, Germany
{sup}171` Space physics and astronomy research unit, University of Oulu, Pentti Kaiteran katu 1, FI-90014 Oulu, Finland
{sup}172` Department of Physics and Astronomy, Lehman College of the CUNY, Bronx, NY 10468, USA
{sup}173` American Museum of Natural History, Department of Astrophysics, New York, NY 10024, USA
{sup}174` Ruhr University Bochum, Faculty of Physics and Astronomy, Astronomical Institute (AIRUB), German Centre for Cosmological Lensing (GCCL), 44780 Bochum, Germany
Gold lens-model reconstructions

Figure A.1. Lens-model reconstructions for Gold candidates 1–6. From left to right, the panels show the observed data, lens-light-subtracted data, reconstructed model, normalised residuals, and the pixelated reconstructions of source 1 and source 2.

Figure A.2. Lens-model reconstructions for Gold candidates 7–11. The panels follow the layout described in Fig. A.1.
Table A.1. DR1 Silver DSPL candidates with official source designations and sky coordinates. Candidates are numbered in order of increasing right ascension, and the numbering follows the labels printed in Figure 4.
| Silver candidates 1–37 | Silver candidates 38–73 | ||||||
| No. | Official name | RA (°) | Dec (°) | No. | Official name | RA (°) | Dec (°) |
| 1 | Euclid DR1 J004938.260−533600.41 | 12.4094 | −53.6001 | 38 | Euclid DR1 J042829.614−352625.48 | 67.1234 | −35.4404 |
| 2 | Euclid DR1 J005136.896−541119.61 | 12.9037 | −54.1888 | 39 | Euclid DR1 J043010.921−241017.92 | 67.5455 | −24.1716 |
| 3 | Euclid DR1 J015016.504−475016.16 | 27.5688 | −47.8378 | 40 | Euclid DR1 J043302.910−495131.42 | 68.2621 | −49.8587 |
| 4 | Euclid DR1 J015306.152−432312.45 | 28.2756 | −43.3868 | 41 | Euclid DR1 J044353.266−225204.10 | 70.9719 | −22.8678 |
| 5 | Euclid DR1 J015716.307−415630.25 | 29.3179 | −41.9417 | 42 | Euclid DR1 J044448.827−381852.90 | 71.2034 | −38.3147 |
| 6 | Euclid DR1 J015749.436−453740.42 | 29.4560 | −45.6279 | 43 | Euclid DR1 J050145.976−411800.54 | 75.4416 | −41.3002 |
| 7 | Euclid DR1 J020203.314−561048.81 | 30.5138 | −56.1802 | 44 | Euclid DR1 J050312.533−301125.81 | 75.8022 | −30.1905 |
| 8 | Euclid DR1 J020636.832−475440.27 | 31.6535 | −47.9112 | 45 | Euclid DR1 J050456.363−340032.88 | 76.2348 | −34.0091 |
| 9 | Euclid DR1 J020712.749−314638.30 | 31.8031 | −31.7773 | 46 | Euclid DR1 J051447.016−405357.27 | 78.6959 | −40.8992 |
| 10 | Euclid DR1 J022455.052−601235.97 | 36.2294 | −60.2100 | 47 | Euclid DR1 J053233.630−371545.77 | 83.1401 | −37.2627 |
| 11 | Euclid DR1 J022748.280−385955.46 | 36.9512 | −38.9987 | 48 | Euclid DR1 J055925.749−520844.09 | 89.8573 | −52.1456 |
| 12 | Euclid DR1 J023930.713−515748.30 | 39.8780 | −51.9634 | 49 | Euclid DR1 J062002.124−491417.05 | 95.0089 | −49.2381 |
| 13 | Euclid DR1 J024734.522−515450.94 | 41.8938 | −51.9141 | 50 | Euclid DR1 J100050.272+013251.75 | 150.2095 | 1.5477 |
| 14 | Euclid DR1 J025109.558−612310.57 | 42.7898 | −61.3863 | 51 | Euclid DR1 J103048.472+755521.17 | 157.7020 | 75.9225 |
| 15 | Euclid DR1 J030425.787−473743.72 | 46.1074 | −47.6288 | 52 | Euclid DR1 J103633.039+683649.44 | 159.1377 | 68.6137 |
| 16 | Euclid DR1 J030718.758−422841.88 | 46.8282 | −42.4783 | 53 | Euclid DR1 J104003.271+742251.11 | 160.0136 | 74.3809 |
| 17 | Euclid DR1 J031158.590−463445.16 | 47.9941 | −46.5792 | 54 | Euclid DR1 J105143.454+724215.93 | 162.9311 | 72.7044 |
| 18 | Euclid DR1 J031543.262−622035.56 | 48.9303 | −62.3432 | 55 | Euclid DR1 J122728.143+702548.39 | 186.8673 | 70.4301 |
| 19 | Euclid DR1 J032408.363−452614.20 | 51.0348 | −45.4373 | 56 | Euclid DR1 J124443.089+644001.66 | 191.1795 | 64.6671 |
| 20 | Euclid DR1 J032732.056−364503.45 | 51.8836 | −36.7510 | 57 | Euclid DR1 J124942.621+661313.49 | 192.4276 | 66.2204 |
| 21 | Euclid DR1 J032801.253−402957.18 | 52.0052 | −40.4992 | 58 | Euclid DR1 J142643.320+441110.32 | 216.6805 | 44.1862 |
| 22 | Euclid DR1 J033258.327−571551.80 | 53.2430 | −57.2644 | 59 | Euclid DR1 J142652.199+433113.80 | 216.7175 | 43.5205 |
| 23 | Euclid DR1 J033611.349−574906.11 | 54.0473 | −57.8184 | 60 | Euclid DR1 J150758.546+724738.33 | 226.9939 | 72.7940 |
| 24 | Euclid DR1 J033709.402−275811.26 | 54.2892 | −27.9698 | 61 | Euclid DR1 J151918.638+721127.70 | 229.8277 | 72.1910 |
| 25 | Euclid DR1 J034234.580−412024.11 | 55.6441 | −41.3400 | 62 | Euclid DR1 J171314.881+602207.86 | 258.3120 | 60.3688 |
| 26 | Euclid DR1 J034704.604−451603.69 | 56.7692 | −45.2677 | 63 | Euclid DR1 J173609.281+605546.61 | 264.0387 | 60.9296 |
| 27 | Euclid DR1 J040105.630−354724.50 | 60.2734 | −35.7901 | 64 | Euclid DR1 J173839.271+610400.88 | 264.6636 | 61.0669 |
| 28 | Euclid DR1 J040821.624−535359.10 | 62.0901 | −53.8998 | 65 | Euclid DR1 J174745.539+640250.16 | 266.9397 | 64.0473 |
| 29 | Euclid DR1 J040920.805−470401.02 | 62.3367 | −47.0670 | 66 | Euclid DR1 J175229.144+610937.80 | 268.1214 | 61.1605 |
| 30 | Euclid DR1 J041000.944−510607.51 | 62.5039 | −51.1021 | 67 | Euclid DR1 J175446.694+713921.20 | 268.6946 | 71.6559 |
| 31 | Euclid DR1 J041013.488−552227.26 | 62.5562 | −55.3742 | 68 | Euclid DR1 J175746.354+651109.29 | 269.4431 | 65.1859 |
| 32 | Euclid DR1 J041020.748−441418.62 | 62.5864 | −44.2385 | 69 | Euclid DR1 J181935.805+570832.64 | 274.8992 | 57.1424 |
| 33 | Euclid DR1 J041739.371−301716.04 | 64.4140 | −30.2878 | 70 | Euclid DR1 J184618.330+723826.77 | 281.5764 | 72.6408 |
| 34 | Euclid DR1 J041906.519−295937.60 | 64.7772 | −29.9938 | 71 | Euclid DR1 J233054.472−330559.85 | 352.7270 | −33.1000 |
| 35 | Euclid DR1 J042037.009−341447.85 | 65.1542 | −34.2466 | 72 | Euclid DR1 J234338.579−335641.58 | 355.9107 | −33.9449 |
| 36 | Euclid DR1 J042532.089−482052.54 | 66.3837 | −48.3479 | 73 | Euclid DR1 J235319.516−602119.86 | 358.3313 | −60.3555 |
| 37 | Euclid DR1 J042724.367−233942.21 | 66.8515 | −23.6617 |
Table A.1.
References
- [1]Ballard, D. J., Enzi, W. J. R., Collett, T. E., Turner, H. C., & Smith, R. J. 2024, MNRAS, 528, 7564
- [2]Barone, T. M., Keerthi Vasan G. C., Tran, K.-V., et al. 2026, AJ, 171, 57
- [3]Belokurov, V., Evans, N. W., Moiseev, A., et al. 2007, ApJ, 671, L9
- [4]Birrer, S. & Amara, A. 2018, Phys. Dark Universe, 22, 189
- [5]Bowden, D. J., Sahu, N., Shajib, A. J., et al. 2025, ApJ, 993, 124
- [6]Bradbury, J., Frostig, R., Hawkins, P., et al. 2018, JAX: composable transformations of Python+NumPy programs, https://github.com/jax-ml/jax, software
- [7]Collett, T. E. 2015, ApJ, 811, 20
- [8]Collett, T. E. & Auger, M. W. 2014, MNRAS, 443, 969
- [9]Collett, T. E., Auger, M. W., Belokurov, V., Marshall, P. J., & Hall, A. C. 2012, MNRAS, 424, 2864
- [10]Despali, G., Heinze, F. M., Fassnacht, C. D., et al. 2025, A&A, 699, A222
- [11]Dux, F., Millon, M., Lemon, C., et al. 2025, A&A, 694, A300
- [12]Einstein, A. 1936, Science, 84, 506
- [13]Enzi, W. J. R., Krawczyk, C. M., Ballard, D. J., & Collett, T. E. 2025, MNRAS, 540, 247
- [14]Euclid Collaboration: Cropper, M., Al-Bahlawan, A., Amiaux, J., et al. 2025, A&A, 697, A2
- [15]Euclid Collaboration: Desprez, G., Paltani, S., Coupon, J., et al. 2020, A&A, 644, A31
- [16]Euclid Collaboration: Jahnke, K., Gillard, W., Schirmer, M., et al. 2025, A&A, 697, A3
- [17]Euclid Collaboration: Li, T., Collett, T. E., Walmsley, M., et al. 2026, A&A, 711, A29
- [18]Euclid Collaboration: Tucci, M., Paltani, S., Hartley, W. G., et al. 2026, A&A, 711, A5
- [19]Euclid Data Release 1. 2026, https://doi.org/10.57780/esa-d079c51
- [20]Falco, E. E., Gorenstein, M. V., & Shapiro, I. I. 1985, ApJ, 289, L1
- [21]Galan, A., Caminha, G. B., Knollmüller, J., Roth, J., & Suyu, S. H. 2024, A&A, 689, A304
- [22]Galan, A., Vernardos, G., Peel, A., Courbin, F., & Starck, J. L. 2022, A&A, 668, A155
- [23]Gavazzi, R., Treu, T., Koopmans, L. V. E., et al. 2008, ApJ, 677, 1046
- [26]Johnson, D., Collett, T., Li, T., & Fleury, P. 2025, JCAP, 08, 067
- [28]Lin, H., Buckley-Geer, E., Agnello, A., et al. 2017, ApJ, 838, L15
- [29]Minor, Q. E., Gad-Nasr, S., Kaplinghat, M., & Vegetti, S. 2021, MNRAS, 507, 1662
- [30]Nightingale, J., Hayes, R., Kelly, A., et al. 2021, J. Open Source Softw., 6, 2825
- [31]Rüstig, J., Guardiani, M., Roth, J., Frank, P., & Enßlin, T. A. 2024, A&A, 682, A146
- [32]Schneider, P. 2014, A&A, 568, L2
- [33]Schneider, P. & Sluse, D. 2013, A&A, 559, A37
- [34]Shajib, A. J. 2019, MNRAS, 488, 1387
- [35]Shajib, A. J., Birrer, S., Treu, T., et al. 2020, MNRAS, 494, 6072
- [36]Sheu, W., Shajib, A. J., Treu, T., et al. 2025, MNRAS, 541, 1
- [37]Stein, M. 2012, Interpolation of Spatial Data: Some Theory for Kriging, Springer Series in Statistics (Springer New York)
- [38]Suyu, S. H. & Halkola, A. 2010, A&A, 524, A94
- [39]Tanaka, M., Wong, K. C., More, A., et al. 2016, ApJ, 826, L19
- [40]Tessore, N. & Metcalf, R. B. 2015, A&A, 580, A79
- [41]Vegetti, S., Koopmans, L. V. E., Bolton, A., Treu, T., & Gavazzi, R. 2010, MNRAS, 408, 1969
- [42]Wingate, D. & Weber, T. 2013, arXiv e-prints, arXiv:1301.1299arxiv.org/abs/1301.1299
- [43]Zhuang, J., Tang, T., Ding, Y., et al. 2020, in Advances in Neural Information Processing Systems, Vol. 33, 18795–18806
- [44]Zwicky, F. 1937a, Phys. Rev., 51, 290
