Euclid Data Release 1 (DR1): Calibration strategy and observations

Introduction

Euclid is a fundamental physics experiment designed measure cosmological parameters with unprecedented accuracy [33]. Every design choice was optimised within the mission’s cost, mass, and engineering constraints. As a result, Euclid was built with minimal movable mechanisms, and equipped with exceptionally well-characterised instruments and telescope, together forming the scientific payload module (PLM). Achieving the required dark energy figure of merit, FoM $\ge400$, drives all specifications for the hardware, the survey, the calibration plan, and the Science Ground Segment (SGS) that processes the data.

DR1 ([48]) is released in two stages: DR1-Foundation, in November 2026, and the complete DR1 in mid-2027, the latter including higher-level science products for galaxy clustering and weak lensing, whose complexity requires the additional development time. This paper equally applies to the full DR1, because the latter builds on derived data products.

This paper introduces the calibration strategy and the observations that support it. Both have been shaped by the anticipated and actual behaviour of Euclid in space and its evolution there, which are presented in a companion paper ([43]). The requirement flow and the calibration concept are described in Sect. 2, together with a performance assessment against the top-level calibration requirements. Details of the DR1 processing pipelines and calibration tasks are given in the separate papers accompanying this release (see [16]). The factors that govern the development and maintenance of the calibration plan, including the sensitivity of Euclid to solar stray light and to X-rays from solar flares, are explained in Sects. 3 and 4. The self-calibration concept, which forms the central pillar of the calibration plan, is introduced in Sect. 5, and the plan itself is laid out in Sect. 6. The limitations of the DR1 data and the improvements planned for DR2 are discussed in Sect. 7. Synthetic transformations between the Euclid, Roman, James Webb Space Telescope (JWST), and CSST photometric systems are presented in Sect. 8; they are given in the AB magnitude system ([73]). A summary is given in Sect. 9.

Requirement flow and DR1 performance

Mission Requirements Document

The Euclid Mission Requirements Document (MRD) scopes and permeates the design of the space and ground segments, and is immutable; details about the MRD verification approach are given in [90]. On its scientific side, the MRD has 22 requirements on weak lensing, nine on galaxy clustering, seven on the Euclid Wide Survey (EWS), and five on the Euclid Deep Survey (EDS), listed in Table A.1. They set the imaging and spectroscopic capabilities, as well as the survey area and depth ([41]). These, in turn, determined the mirror size, field of view (FoV), instrumentation ([19, 25]), survey strategy, calibration plan, and processing pipelines.

As the implementation progressed, an increasing number of properties became fixed within the boundaries of the MRD: first the high-level architecture (for example optics, instrument mass and volume, bandpasses and dispersers, and detector layouts), followed by the lower-level hardware and software details. At the same time, performance trade-offs were managed and the calibration plan evolved in step. The MRD scientific requirements flow down to a two-tier calibration concept (Figure 1) as follows.

Vector (PDF)

Calibration concept document – part A

The CalCD-A contains 15 top-level requirements on point-spread function (PSF) and shear calibration, spectrophotometry, photometric redshifts, and spectroscopy (Table A.2). Like the MRD, the CalCD-A is under ownership of the European Space Agency (ESA) and immutable.

Formally, CalCD-A requirements must be met by DR3 at the end of the nominal mission. Tables A.1 and A.2 show that this is already the case for many of them, with more details in the companion papers and with considerable possibilities for further improvements in DR2 (Sect. 7). The CalCD-A requirements that are still unmet – purity and completeness of the spectroscopic redshift sample, and the VIS PSF model – are discussed in Sect. 2.5.

Calibration concept document part B

The CalCD-B unfolds the CalCD-A requirements into individual error budgets and dependence trees, comprising approximately 350 entries. It is under ownership of the SGS and was frozen in 2018 after it reached sufficient maturity to guide the – parallel – development of the calibration plan (Sect. 3).

The large envelope of the CalCD-B demands simplicity in order to remain tractable and purposeful as a static reference. It treats all effects contributing to an error budget as statistically independent, and uses quadratic error propagation whenever possible. This is often unrealistic, since several effects can be entangled at the physical level, such as nonlinearity, gain, charge-transfer inefficiency (CTI, [68]), and the ‘brighter-fatter effect’ (BFE, [2, 78, 55]). The permissible knowledge error of such complexes can be evaluated with simulations. This, however, requires realistic models for these effects and of their target – in this example, the PSF – in the simulator well before launch. Thus, the CalCD-B does not capture BFE or CTI, but both effects are accounted for in the calibration plan as well as the VIS and shear data-processing chains.

The CalCD-B also uses ground-characterisation data whenever possible ([92, 54, 42, 50, 45, 19, 25, 22, 27]). However, some of these tests need to be conducted in a thermal vacuum and the facilities that offer these conditions are a limited resource. Furthermore, tests cannot necessarily replicate space conditions accurately, such as a zodiacal background, in-flight temperatures, zero gravity, non-standard instrument operations, and the irradiation by the Sun; some characteristics can be determined accurately only in flight ([43]).

When a given error budget remained otherwise unconstrained, its noise contributors were thus assumed to contribute equally. The CalCD-B merely serves as a guideline for the SGS, which may rebalance the error budgets accounting for pipeline performance and characteristics of real data, as long as the CalCD-A requirements are met. When constructing the in-flight calibration plan (Sect. 3.2), a conservative approach was followed to ensure sufficient signal-to-noise ratio (S/N) and margins, in particular where the CalCD-B was uncertain.

Validation and verification process

Calibration products undergo a multi-stage quality assessment before use. An initial check is performed after the pipeline has produced the calibration product, confirming successful execution and matching expectations. In the next step, the detector and instrument scientists review the data for correct instrument-specific behaviour, and check for evolutionary trends. The final inspection by the calibration scientist closes the loop, and may include a check against the CalCD-B. Formally, the first two stages constitute validation – assessing whether the product looks correct and behaves as expected – while the final comparison against the CalCD-B is verification, confirming that the product meets its requirements. These stages largely overlap in practice and run on different timescales as the data become available; the flat hierarchy and redundancy are intentional, so that issues can be caught that a single reviewer might miss. Outcomes are reviewed on a weekly basis.

For every major data release (DR), once the data have been reprocessed with a frozen pipeline configuration, the SGS does a full verification against CalCD-A requirements. Verification may also occur between releases, for example when major development milestones have been achieved and are regression-tested on several hundred square degrees of data (§ 7 in Euclid Collaboration: Ruffenach et al. [40]). ESA may additionally call a results review to verify MRD conformance.

Requirements not yet met by DR1

Most mission-level requirements have been met already by DR1-Foundation ahead of DR3 (Sect. 2.2). The two areas where requirements are not yet met have developed more slowly, because not all necessary data have been taken for spectroscopy, and because the PSF model has a large dependency tree that requires careful calibration, characterisation, and control of systematic errors.

Redshift requirements

MRD–GC–006 (Table A.1) requires a purity of $\mathcal{P} > 99\%$ for the EDS reference redshift sample. This cannot yet be formally assessed, because both blue- and red-grism data at full depth are needed, as well as an external redshift catalogue that is complete to the selection limits of $H_{\mathrm{E}} < 24$ and $f_{\mathrm{H}\alpha} > 2 \times10^{-16}\,\mathrm{erg\,cm^{-2}\,s^{-1}}$. For a first assessment, the centre of the COSMOS field (Scoville et al. [84]), observed to full EDS depth during performance verification (PV), was used together with around 8000 reference redshifts from Khostovan et al. [61], which are incomplete at this depth. A purity of $\mathcal{P} = 85\%$ (93–100%) is found for a redshift-reliability threshold of $\mathrm{SPE\_Z\_PROB} > 0.99$ ($> 0.999$). The purity in the Euclid Deep Field North (EDF-N) is expected to be lower than in COSMOS, because DR1 has not yet reached full depth in this area, which lowers the fraction of correctly assigned redshifts for fainter targets. This is discussed in sections 4 and 5 of Euclid Collaboration: Le Brun et al. [30]. Meeting MRD–GC–006 is a prerequisite for assessing the following two requirements.

MRD–GC–007 requires $\mathcal{P} > 80\%$ for the EWS. The current best estimate, obtained from the COSMOS and EDF-N data, is $\mathcal{P} \approx75\%$ at a number density of galaxies with correct redshifts of $\mathcal{D} > 1700\,\mathrm{deg}^{-2}$. To enable a better purity assessment, the survey plan was modified so that the EDF-N is completed for DR2 instead of DR3 (see also section 2.3 in Euclid Collaboration: Tereno et al. [46]).

MRD–GC–002 requires a redshift completeness of $\mathcal{C} > 45\%$ for the EWS, with a goal of $\mathcal{C} > 65\%$. This cannot yet be assessed directly, because the available deep catalogues are themselves incomplete and impure. Instead, the completeness is estimated from the number density $\mathcal{D}$ of galaxies with correct redshifts in the COSMOS data. For $\mathrm{SPE\_Z\_PROB} > 0.99$, a value of $[\mathcal{D}, \mathcal{P}] \approx[750\,\mathrm{deg}^{-2}, 62\%]$ is found, below the target of $[1700\,\mathrm{deg}^{-2}, 80\%]$. The higher zodiacal and Galactic backgrounds in the COSMOS field were compensated for by approximately twice the exposure time. Preliminary estimates of $[1600\,\mathrm{deg}^{-2}, 63\%]$ and $[1700\,\mathrm{deg}^{-2}, 68\%]$ are obtained from the EDF-N and EDF South (EDF-S), respectively, which suggests that the number density in COSMOS is low because of cosmic variance. A more robust assessment remains contingent on the full-depth EDF-N data.

PSF requirements

Shear calibration requires a highly complicated, chromatic model of the VIS PSF that is still under development at the time of writing. Its aspects and performance are discussed in current and upcoming DR1 publications: shear overview, requirement, method, and validation (Taylor et al., in prep.); shear calibration (Martinet et al., in prep.); the PSF model (Miller et al., in prep.); PSF calibration (Duncan et al., in prep.); PSF calibration fields (Kim et al., in prep.); and Euclid’s thermo-optical properties and guiding performance (Euclid Collaboration: Anselmi et al. [14]; Euclid Collaboration: Schirmer et al. [43], and Whittam et al., in prep.). An overview of the observational programme that supports the PSF model is presented in Sects. 6.3 and 6.4.

Implementation of the calibration plan

At the system-engineering level

Lorenzo Alvarez et al. [64] describe the model-based system-engineering approach chosen for the development of the Euclid mission. The implementation was done in Enterprise Architect (EA)¹, a comprehensive tool that provides, among many other capabilities, automated flow checks and compliance matrices for the development of the CalCD-B. However, the large volume of the CalCD-B (Sect. 2.3) and the complexity of EA made this approach intractable for the SGS. Furthermore, as Euclid matured, ESA’s system-engineering support shifted to other missions in earlier development stages.

We considered exporting the calibration- and processing-related components from this model for maintenance in simpler tools. This, however, would have required considerable time and training across a broad group of core personnel to ensure consistent use and correct operation – changes of this were untenable in the run up to launch. We then descoped the calibration framework (Sect. 3.2) from the model, while CalCD-A and -B persist in EA as static references for our operations (Figure 1).

Calibration Framework (CalF)

The calibration plan has been developed and is maintained on a lighter and more pragmatic system, based on python-sphinx and originally developed by the Kilo-Degree Survey (KiDS; de Jong et al. [9]) team for the same purpose. This calibration framework (hereafter CalF) provides an html information base as the central reference for all practical calibration matters. It is flexible enough to accommodate differences between expected and actual performance, as well as unforeseen events, especially during early operations. Compliance matrices linking calibration observations (CALBLOCKs) and calibration products (CALPRODUCTs) are generated automatically from the sphinx internal hyperlink targets whenever the CalF is updated. The proper flow-down of all CalCD-B requirements to CALPRODUCTs was verified in ESA’s critical design reviews before launch.

In 2022, around 1.5 years before launch, the CalF was placed under configuration control as the survey design and PV planning became increasingly stable and rigid. After major updates, frozen releases of the CalF were issued, establishing it as the new reference for all downstream stakeholders, such as the survey computations, ESA’s mission operations, the instrument operation teams, and the SGS processing functions (Euclid Collaboration: Aussel et al. [16]; Euclid Collaboration: Ruffenach et al. [40]). The CalF does not automatically issue update notifications, nor does it have an interface to receive change requests. These are all interactive processes between team members, ultimately signed off by configuration-control boards to ensure full compliance with hardware, software, operations, the survey, and science.

Updates since launch

The CalF absorbs all calibration-related dynamics from operations, respecting the frozen nature of the CalCD-B. Three updates were made after commissioning and PV, to suppress parasitic stray-light paths from the Sun, and to become more robust against X-rays from solar flares. During routine operations, additional updates occur whenever instrument monitoring, product validation, space weather, or survey constraints require changes; four such updates were made over 2.7 years of science operations, for example adding more calibration-lamp configurations for flat-fields and nonlinearity measurements. No adjustments were required because of molecular contamination, which was anticipated and folded into the plan from the start, yet has posed considerable challenges at times (sections 5 and 6 in Euclid Collaboration: Schirmer et al. [43]). Overall, the launch and DR1 configurations of the CalF are around 95% identical, and only minor changes have occurred since toward DR2. Most of the current development concentrates on how the calibration data are processed and applied to the science data, rather than on how they are obtained in space.

Limitations on the calibration plan

The spacecraft, telescope, instruments, orbit, and mission lifetime pose numerous constraints on the PV and calibration plans. As an additional layer of complication, every on-orbit operation causes a reaction in the spacecraft that must be evaluated for subtle impacts on calibration and performance. There is no ground to absorb torque or force, and any excess heat – or lack thereof – caused by non-standard operations eventually changes the thermo-optical equilibrium (Euclid Collaboration: Schirmer et al. [43]; Euclid Collaboration: Anselmi et al. [14], and Whittam et al., in prep.).

Reference observing sequence

Onboard activities are orchestrated by the MIL-STD-1553B data bus protocol standard², implemented in a deterministic serial data bus used in spacecraft to connect sub-systems. It operates with a dual-redundant, command-response architecture, prioritising reliability and fault tolerance over data rate (1 Mbit s$^{-1}$); science data are transferred using a SpaceWire³ interface with higher data rates (several MByte s$^{-1}$). At bus level, simultaneous on-sky observations with NISP and VIS are substantially more complicated than a single-instrument observation, because both instruments’ timelines become interwoven.⁴

Euclid utilises a restricted set of observing modes cast into the reference observing sequence (ROS) that exploits simultaneous observations. The ROS was extensively tested in this regard, and is therefore used whenever possible for both science observations and on-sky calibration purposes. It comprises four dither positions, at each of which the same series of VIS and NISP science observations is repeated (section 4 in Euclid Collaboration: Scaramella et al. [41]). Aside from small sky offsets, the dithers differ only in the VIS calibrations (Table 1) that are taken while NISP acquires its imaging exposures. At the end of the ROS, during the slew to the next field, NISP takes a slew dark with the same exposure duration as a NISP photometric image. The slew dark is not pure, because it also contains the charge-persistence signal (Euclid Collaboration: Kubik et al. [28]) from the detectors’ previous illumination history; its purpose is to deliver high-cadence snapshots to assess detector health. The total duration of the ROS is on average 73.9 min, including the variable length of approximately 230 s for a typical field slew.

VIS
4 $I_E$-band nominal exposures	566 s each
2 $I_E$-band short exposures	95 s each
Bias	2 per day
Dark	4 per day
Flat	6 per day
Trap pumping	6 per day
Charge injection	8 per day
NISP
4 red-grism spectro exposures	574 s each
4 $Y_E$-band exposures	112 s each
4 $J_E$-band exposures	112 s each
4 $H_E$-band exposures	112 s each
1 slew dark	112 s

Although the ROS often generates data beyond the immediate calibration requirements, its use ensures high reliability, minimises development and maintenance costs, takes the calibration data in the same way as the nominal survey observations, maintains the thermal stability of the spacecraft, and increases the astrophysical value of the acquired data. Special, simultaneous on-sky NISP and VIS calibrations were implemented only for a few recurring, long-duration tasks to maximise efficiency.

Some on-sky calibrations need only the observations performed at the first dither position, which we refer to as ROS_D1, repeated at each node of a non-standard dither pattern. Concatenating many ROS_D1 observations does not disturb the thermal equilibrium.

Spacecraft restrictions

Numerous restrictions apply to the design of the calibration plan.

• For self-calibration purposes, a field with perennial visibility must be observed on a monthly basis to detect and absorb evolution of the PLM. The only such field usable is near the North Ecliptic Pole (Sect. 5).

• Any calibration observation must minimise the thermal disturbance of the PLM, maximising PSF stability for the survey. To this end, the spacecraft must stay within the allowed ranges of the azimuth angle (AA) and solar aspect angle (SAA), and the instruments must operate with a survey-like cadence. If one instrument stops taking data, for example because it is not needed for some calibration, or it observes at a much higher cadence, then the temperatures in the PLM change and, subsequently, also the PSF. Even changes as small as a few decimal kelvin of the baseplate carrying the primary mirror on one side and the instruments on the other will be seen as a change in the PSF after a few hours (see Euclid Collaboration: Schirmer et al. [43]; Euclid Collaboration: Anselmi et al. [14]).

• The typical thermo-optical e-folding timescale was estimated to be 21 h before launch (see also section 4 in Euclid Collaboration: Anselmi et al. [14]), meaning the telescope may need 2–10 days to reach thermo-optical stability after a major excursion (Euclid Collaboration: Schirmer et al. [43]; Euclid Collaboration: Anselmi et al. [14]).

• The data volume created by calibration activities may exceed the downlink telemetry limit of 820 Gbit for a few days only.

• The high-gain antenna must repoint at least every 50 min to track the designated ground station. To avoid spacecraft jitter during the exposures, antenna movements are synchronised with a dither or field slew. Hence, Euclid cannot stare at the same point on the sky for longer than 50 min without making an offset.

• The absolute pointing accuracy is around 3″ or 30 VIS pixels, limiting system characterisations where higher pointing accuracy is needed; for example, placing a bright star exactly on a detector edge to check for unwanted reflections and glints.

• The minimum eigenslew length (dither offset) is 73″. Offsets smaller than this threshold must be synthesised via larger, partially compensating slews. Precise sub-pixel offsets cannot be made.

• The NISP filter and grism-wheel movements are not perfectly compensated by the spacecraft. Wheel movements are forbidden while VIS is taking flats or on-sky exposures.

• The minimum replanning time for Euclid is five weeks, meaning failed calibrations cannot simply be repeated back-to-back. When an outage occurs, the spacecraft rejoins the survey at the next possible point. Lost calibrations of great importance must be rescheduled, and might require an expensive recomputation of the survey. Therefore, the calibration plan must be robust enough to tolerate the loss of single instances of most calibration types without replacement.

The only permissible exceptions are those that do not alter the precomputed survey timeline, such as a change of calibration-lamp flux or type.

• Orbital maintenance and other service module (SVM)-related operations occur during 6 h-long fixed time slots every four weeks on a Monday during working hours, the spacecraft operations (SOPS) window; the cadence was decreased in January 2026 to once every eight weeks due to excellent spacecraft performance (section 2.3.3 in Euclid Collaboration: Schirmer et al. [43]). Instruments do not take data during this time; long calibration observations that cannot be interrupted must not fall into these windows.

• The secondary-mirror focusing and tilt mechanism has a limited lifetime budget of total distance travelled. It must not be exceeded by the defocused observations taken during the phase-diversity campaign (CALBLOCK-F-013, see Appendix § 14).

Instrument restrictions: VIS

• The stray-light sensitivity of VIS (Euclid Collaboration: Mellier et al. [33]) must be respected. The spacecraft attitude is limited to $\mathrm{AA} \in[-8^\circ, -3^\circ]$ and $\mathrm{SAA} \in[87^\circ, 112^\circ]$, narrowing the windows during the year when specific calibration fields other than the self-calibration field become visible.

• While the stray-light patterns are present in both open- and closed-shutter configurations, they are not identical.

• Because the VIS shutter must remain open to receive light from the calibration lamp, astrophysical sources are inevitably superimposed on every flat field. Each flat requires a corresponding 2 s science exposure within a few pixels of the same dither position, to identify affected pixels and exclude them during data processing.

• The VIS onboard processing time depends on the number of sources in the image, and on the brightness of a flat-field exposure.

Instrument restrictions: NISP

• The NISP wheel positioning is not perfectly deterministic, meaning the optical distortions are slightly different when restoring a previous configuration.

• The onboard image processing of a single dither must be completed before the end of the first exposure of the next dither (Medinaceli et al. [71]). Up to 46 co-added groups, telemetry, and ancillary data can be acquired per dither (Bonoli et al. [6]).

• NISP uses the multiple accumulated (MACC) mode for exposures, specified as $\operatorname{MACC}(g,r,d)$, as described in section 3.7.2 in Euclid Collaboration: Jahnke et al. [25]. The parameters describe, respectively, the number of groups ($g \le15$), the number of frames per group ($f \le16$), and the number of dropped frames between groups, with a frame exposure time of approximately 1.45 s.

• NISP can hold at most seven exposures at a single dither.

Options and design choices

• VIS and NISP can take darks and flats regardless of time and current spacecraft pointing, outside SOPS windows reserved for orbital maintenance.

• On-sky calibrations and filler observations during waiting times (for example, thermal stabilisation after decontamination; sections 5 and 6 in Euclid Collaboration: Schirmer et al. [43]) should always be taken at the self-calibration field to have a well-known reference; as a spin-off, this increases the depth of the Euclid Ultra Deep Field (EUDF) and supports time-domain astrophysics.

• The fluence in the NISP flats is controlled by the lamp flux, not the exposure time, so that the flats can be taken with the same MACC(4,16,4) mode as the survey imaging observations.

• Slew darks (Table E.1) are nominal darks that must use the MACC(4,16,4) and MACC(15,16,11) modes from imaging and spectroscopy, respectively. Slew darks are taken at the end of every ROS for detector-monitoring and high-energy space-weather effects (Gabarra et al. 2026).

• The ROS must be used as much as possible, in order to have calibrations taken in a manner as close as possible to the science observations, and to maximise thermal stability of the PLM.

Within these constraints, the calibration plan was designed for maximum synergy, extracting multiple calibration products from a single CALBLOCK while simultaneously protecting science time (see also section 4 in Euclid Collaboration: Mellier et al. [33]). Concise descriptions of the CALBLOCKs are given in Appendix § 14.

Self-calibration

Illumination non-uniformities

Causes

Despite Euclid’s vignetting-free design, the telescope pupil is seen at different angles from across the focal-plane array (FPA), producing a spatial variation in illumination (geometric apodisation, a passive byproduct of the chosen optical design; see Reiss [80]). Mirror-coating variations introduce additional non-uniformities, through spatial variations and its dependence on angle of incidence. Although the coating reflectances were measured on the ground as a function of wavelength and angle of incidence across optical surfaces, no ray-tracing model including apodisation was made. The combined effects are expected to produce static illumination variations of 1–2% across the FPA, on angular scales $\gtrsim1'$.

VIS has a shutter-induced exposure-time gradient (Sect. 7.4), whilst for NISP there are contributions from uncorrected persistence. Considerable vignetting of 10% occurs in the NISP dispersed images over the outermost column of four detectors, when the $\pm4^\circ$ rotated grism positions are used (section 2.4.3 in Euclid Collaboration: Copin et al. [18]).

A time-dependence is caused by space weathering that slowly erodes optical surfaces (Engelhart et al. 2017; Rodmann et al. [83]; McElwain et al. [69]) on timescales of years, and by molecular contamination (Euclid Collaboration: Schirmer et al. [44]) that can act on much shorter timescales of weeks and months, see Sect. 7.8 here, and section 6 in Euclid Collaboration: Schirmer et al. [43]. Both factors introduce chromaticity, because of the wavelength dependence of scattering and interference effects (Sect. 7.3.2).

Mitigation

We account for these effects jointly, using repeated and widely dithered observations of the self-calibration field (Sect. 5.2). Each visit of the field lasts 23.5 h and is scheduled every 25 to 35 days to sufficiently sample the expected evolutionary processes (with exception of the heavy contamination period between March and June 2023, see section 6.2 in Euclid Collaboration: Schirmer et al. [43]). This frequency is high enough to tolerate the loss of a single visit, for example due to solar activity.⁵ The correcting photometric flat, or large-scale flat (LSF), is computed from the differential photometry of sources as dither offsets move them across the FPA (respective sections 4.4 and 3.2.2 in Euclid Collaboration: McCracken et al. [32]; Euclid Collaboration: Polenta et al. [37]).

Differential photometry can also be obtained from overlapping regions of adjacent survey fields. In the EWS, however, these overlaps amount to only around 2.2% of the field area, reflecting the tiling strategy that maximises unique sky coverage within the mission lifetime. This fraction is too small to support an uber-calibration strategy across all detectors as in Padmanabhan et al. [74]; ); Wolf et al. [93]; ); Magnier et al. [66]; ); Xiao & Yuan [94]; we consider it as an independent sanity check for DR3.

Location of the self-calibration field

The self-calibration field must have perennial visibility. Euclid’s continuous viewing zone (CVZ) is set by thermal considerations: the requirement SAA ≥ 87° – applicable to any time in the year – restricts the maximum angular distance from an ecliptic pole to 3°. Accounting for the dither-pattern size (Sect. 5.3), Euclid’s FoV, and perspective changes during the wide L2 halo orbit (section 2.2 in Euclid Collaboration: Schirmer et al. [43]), the centre of the self-calibration field must fall within 2.3° of a pole.

Exclusion of the southern continuous viewing zone

To avoid method-dependent biases, calibration fields are processed and analysed with the same tools as the nominal, sparse survey fields. This rules out the area around the SEP near the LMC, where heavy crowding would make calibration products unrepresentative of the science observations. In addition, the LMC’s stellar population is considerably bluer than the Galactic halo population seen in the EWS (Figure 2), which would introduce undesired biases in astrometric and photometric calibration products (see Sect. 7.3.2, and sections 4.2.3 Euclid Collaboration: McCracken et al. [31]).

Vector (PDF)

There is a risk that the LSF and other calibration products obtained from this line of sight are not representative of EWS conditions, for example due to the chromaticity of the VIS PSF (see sections 4.2.3 and 4.6 in Euclid Collaboration: McCracken et al. [31]).

Placement in the northern continuous viewing zone

The stellar density near the NEP is several times higher than the average across the EWS, providing good statistical S/N, yet it is low enough for standard pipeline tools. The halo stars are also redder and spectrally closer to both typical survey stars and field galaxies (Figure 2). The self-calibration dither pattern (Sect. 5.3) requires 2.0 deg², which fit comfortably within the 16.5 deg² area accessible within 2.3° of the pole.

The final centring at $(\alpha,\delta) = (268.8127,+65.286)$ was determined pre-launch, accounting for several constraints. Increased reddening toward the Galactic plane to the north-east, together with the presence of NGC 6543 – a bright planetary nebula (PN) of 6′ diameter – motivated a shift toward the south-west. There, we established three faint hot white dwarfs (WDs) as absolute flux calibrators with the Hubble Space Telescope (HST),⁶ and at least one was required to fall within the self-calibration field. A final step ensured no stars brighter than 7.0 AB mag in any of the $g$, $r$, $i$, $J$, $H$, and $K_{\mathrm{s}}$ bands appeared in the field, based on the Asteroid Terrestrial-impact Last Alert System (ATLAS) catalogue of Tonry et al. [89].

Dither pattern

The dither pattern used to build the LSF must:

• sample all relevant spatial scales, from the diagonal diameter of the FPA down to the smallest scale of 10″ imposed by the scientific requirements;

• be fully contained within the CVZ;

• be fixed on the sky rather than rotating with Euclid’s orbit around the Sun, for operational reasons;

• fragment into clusters of four pseudo-random dithers with separations ≤ 5″, to support early versions of the

spectra-order decontamination pipeline (Euclid Collaboration: Romelli et al. [39]).

After correction by the LSF, the relative photometric accuracies for VIS, NISP-P, and NISP-S must be 1.0%, 1.5%, and 0.7%, respectively. Holmes et al. [57] find that random dither patterns outperform regular grids in terms of accuracy, and our simulations show they also converge faster. Accordingly, 60 dither points are required to ensure the relative photometric requirements for the VIS and NISP-P imaging channels are met, and by extrapolation also that of the NISP-S dispersed images.

To find the optimal dither pattern, we generated 1000 pseudo-random and compliant realisations of 60 dithers. Then, for each pattern, we computed the projected lengths onto the horizontal and vertical axes of all 1770 possible line segments between two dither points (regardless of the instrument). The mean length frequency decreases with scale (Figure 4), with an excess below 5′ arising from the sub-grouping imposed by the spectroscopic decontamination pipeline. The distribution shape naturally supports LSF reconstruction: the many short-separation segments sample the high-frequency modes, while the fewer low-frequency modes that fit in the FoV require correspondingly fewer low-frequency segments. To identify the best pattern among the 1000 realisations, we minimised the summed absolute deviations from both mean distributions simultaneously. This ensures that no angular scale is over- or under-represented for the computation of the LSF; this is verified by the theoretical length distribution for line segments randomly placed in a circle (black line in Figure 4), which the chosen dither pattern follows closely.

Vector (PDF)

The self-calibration dither pattern is shown in Figure 3. It has performed well for all three instrument modes, delivering LSFs that absorb residual systematic zero-point variations on all relevant angular scales after the preceding calibration steps (Euclid Collaboration: McCracken et al. [31]; Euclid Collaboration: Polenta et al. [36]; Euclid Collaboration: Fumana et al. [21]). For VIS, the minimum S/N threshold of 80 on individual stellar measurements – imposed to reduce sensitivity to faint cosmic rays – proved unproblematic, despite yielding a lower usable source density than anticipated. The stable VIS throughput allowed us to combine the data of five self-calibration visits for DR1, benefiting both from the larger statistical sample and from the additional sampling baselines resulting from the approximately 30° field rotation between monthly visits. For NISP, we use the individual monthly LSFs for DR1 because of the ongoing detector contamination (§ 6 in Euclid Collaboration: Schirmer et al. [43]).

Vector (PDF)

The chosen dither pattern will remain unchanged for the rest of the mission for several reasons: (i) the pattern results in excellent calibration products; (ii) the resources of finding a more performant pattern are considerable and not available in the SGS; (iii) the gain would be unclear; (iv) maintaining the same dither pattern ensures that any statistically significant changes in calibration products are due to actual changes in the PLM and not due to a discontinuity in our calibration scheme.

Observations at the dither positions

At each of the 60 dithers we execute the ROS_D1 that provides a nominal and short VIS exposure, as well as one image for each NISP filter, and therefore the imaging LSFs can be built after every visit. The spectroscopic LSFs for all five grism configurations – four red, one blue – require more visits because the ROS_D1 includes only one spectroscopic exposure. We cycle through four self-calibration variants, in which the imaging configuration is identical; the spectroscopic configuration interleaves all red-grism configurations for all dither positions, thus requiring four visits to complete the spectroscopic data set.

The blue grism was integrated differently, because its single dispersion direction does not provide sufficient angles for order decontamination. In each of the four self-calibration variants, the 60 dithers are therefore bracketed by eight blue-grism ROS_D1 at each end, repeating 16 of the 60 dithers. The AA is offset by $\pm1.4^\circ$ to $\pm2.0^\circ$ for the two blue-grism brackets to provide the necessary angles for order decontamination; the grism wheel itself must be kept fixed, since rotating it would introduce vignetting in the blue-grism data. These small AA rotations do not disturb the thermal equilibrium, which is more susceptible to potentially large SAA changes due to the return trip from a survey field to the NEP near SAA = $90^\circ$ (§ 4 in Euclid Collaboration: Schirmer et al. [43]). After four self-calibration visits, the blue grism has stepped through all 60 dither positions at least once, and the 16 additional ROS_D1 runs provide extra data for the monthly imaging LSFs.

Calibration plan

Performance-verification phase

Following spacecraft commissioning (see e.g. Euclid Collaboration: Cogato et al. [17], for NISP), 60 days were allocated for performance verification (PV) to characterise the PLM under inflight conditions, and to update ground-calibration data (see also Euclid Collaboration: Mellier et al. [33]; Euclid Collaboration:

Notes. The ROS is executed around 20 times per day, cycling through different configurations to provide a range of calibrations (marked with bold face characters). ROS configurations using the blue grism are used only for the EDS and Euclid Auxiliary Fields (EAFs). The NISP slew darks offer a great resource to monitor detector health, effects of space weather, and charge persistence (Gabarra et al. 2026). Most NISP calibrations are taken outside the ROS with dedicated calibration blocks.

Jahnke et al. [25]). A Gantt chart of the PV schedule is shown in Fig. B.1, and technical details are presented in § 5 of Euclid Collaboration: Tereno et al. [46].

Whenever feasible, we used the ROS for calibration observations (Sect. 4.1). Its use has the added benefit of providing astrophysical data that would normally not be taken, such as deep exposures toward the Magellanic Clouds for wavelength and distortion calibration (Euclid Collaboration: Paterson et al. [35]; Euclid Collaboration: Fumana et al. [21]), and observations of the Orion star-forming regions (Euclid Collaboration: Aussel et al. [15]). Approximately 40% of PV activities utilised the ROS. Apart from numerous technical characterisations and calibrations, we validated the zodiacal background and Galactic stray-light models used for survey S/N estimates, and obtained survey-like data from passes over the EDS areas and the COSMOS field (Table C.1).

Determination of the PLM’s thermal timescale was a central task in PV. Changes in spacecraft attitude with respect to the Sun result in a thermo-optical response as the PLM equilibrates. The duration of this response must be known to ensure that VIS data are suitable for PSF calibration and shear measurements. We determined an e-folding timescale of 21 h by first allowing the spacecraft to equilibrate at AA = $+3.5^\circ$ and then reorienting it by a deliberately large amount, to AA = $-7.5^\circ$, to create a considerable thermal disturbance not encountered in nominal survey operations. We then monitored the PSF continuously for several days until stability was restored. In Euclid Collaboration: Schirmer et al. [43] we analyse the principal thermal characteristics from two years of routine survey operations, showing that changes in SAA have a much more profound effect than those in AA (see also Euclid Collaboration: Anselmi et al. [14]).

Survey phase

The 18 different CALBLOCKs scheduled in the survey are summarised in Appendix § 14. They span the full range from hardware calibrations to target characterisations. Data products derived from them – and from PV – are summarised in Appendices § 15

Notes. Location, area, and planned DR3 depth compared to the EWS. Fields printed in bold-face are included in DR1, with their completeness in terms of exposure time shown in the last column. The COSMOS field consists of a $2 \times2$ mosaic, of which in DR1 50% of the area has 30% depth, and a central pointing from PV that has 100% depth.

to H; processing details are given in the companion papers (see Euclid Collaboration: Aussel et al. [16]).

Among the two with the most added astrophysical value are: first, the observations of several hundred thousand galaxies to $6\times$ the S/N of the EWS, to calibrate the weak lensing noise bias ([56] (Refregier et al. 2012)) and to characterise the morphologies of galaxies; and second, for the galaxy clustering science a large spectroscopic sample of galaxies is needed with the aim of 99% completeness and purity (Euclid Collaboration: Monaco et al. [34]), taken with the red and blue grisms. These two cases were merged with the separate scientific demand for a deep survey of at least 40 deg$^2$, into the EDS, covering 53 deg$^2$ (Euclid Collaboration: Tereno et al. [46]). These deep data facilitate the fine-tuning of pipeline parameters, such as S/N-dependent thresholds for masking the VIS dichroic and crosstalk ghosts. The EUDF is approximately 1.5 mag deeper still over 2 deg$^2$, serving numerous technical, scientific, and cross-calibration purposes with other surveys (Sect. 8).

Additional deep fields (Table 2) are used to calibrate photometric redshifts for Euclid’s passbands (Euclid Collaboration: Hartley et al. [23]), and the effect of colour gradients in galaxies on shape measurement in the presence of a chromatic PSF ([85, 26] (Er et al. 2018)). The Euclid Auxiliary Fields (EAFs) were selected to have a large number of galaxies with known spectroscopic redshifts, as well as deep coverage with HST and ground-based observatories. We observe them repeatedly with the ROS, with every visit offset along a Vogel spiral (Vogel [91]) for homogeneous total exposure-time maps.

Field	RA	Dec	Area	DR3 Depth	DR1
AEGIS	$214.\!^{\circ}827$	$+52.\!^{\circ}82$	1.0	$4\times$	100%
CDFS	$53.\!^{\circ}117$	$-27.\!^{\circ}81$	0.5	$5\times$	100%
EUDF	$268.\!^{\circ}813$	$+65.\!^{\circ}29$	2.0	$8\times$	50%
COSMOS	$150.\!^{\circ}119$	$+2.\!^{\circ}21$	2.0	$5\times$	61%
GOODS-N	$189.\!^{\circ}250$	$+62.\!^{\circ}25$	0.5	$4\times$	0%
VVDS	$36.\!^{\circ}500$	$-4.\!^{\circ}50$	0.5	$5\times$	0%
SXDS	$34.\!^{\circ}500$	$-5.\!^{\circ}00$	2.0	$5\times$	0%

Phase-diversity campaign

Our strategy to build and apply a PSF model consists of two complementary components. The first component is the phase-diversity campaign (PDC), used to construct a physics-based wavefront model that relates spacecraft attitude and thermal state to the PSF across the FPA. It was executed once, before the start of survey operations, and is presumed to remain valid throughout Euclid’s lifetime because of the SiC design. A Gantt chart of the PDC is shown in Figure 5. The second set of components are the visits of PSF calibration fields (Sect. 6.4) that act as frequent empirical checkpoints, validating this model against observed stellar fields at the actual attitudes and epochs of the survey. Implementation details are given in section 6 in Euclid Collaboration: Tereno et al. [46].

Vector (PDF)

Phase diversity is a wavefront-sensing technique in which the same stellar field is imaged at the nominal focus and at two defocused positions, one intra- and one extra-focal. To defocus, the secondary-mirror mechanism (M2M) is moved by $\pm18\,\mu\mathrm{m}$ along the optical axis, shifting the optical focal plane by $113\,\mu\mathrm{m}$ with respect to the detector plane. This broadens the nominal PSF – with a typical full width at half maximum (FWHM) of 1.4 pixels (Euclid Collaboration: McCracken et al. [31]) – into a non-uniform disc of 8 pixels diameter. An excellent S/N is required so that photon noise is not mistaken for true surface-brightness variations in the discs. The differences between the intra- and extra-focal disc images encode the wavefront aberrations, from which the full pupil-plane wavefront error (WFE) is reconstructed. The WFE is chromatic because the 200 dielectric coatings layers on the dichroic beamsplitter are non-uniform; their thickness variations were determined on the ground (Baron et al. [3]) and are included in the PSF model. The model is discussed in detail in Miller et al. (in prep.).

Utmost thermal stability is required, achieved by maintaining a fixed spacecraft attitude for seven days (see Euclid Collaboration: Schirmer et al. [43], and Figure 5) before the PDC observations at that attitude. This condition could not be met by any subset of the EWS, due to frequent attitude changes. Separate filler surveys were designed for each PDC attitude, using the ROS to maintain the typical heat signatures of the instruments injected into the PLM (section 4 in Euclid Collaboration: Schirmer et al. 2026). These surveys have unique tiling patterns and comprise an open-cluster field, the Euclid Medium Deep Survey, and the Ecliptic Survey that already produced spin periods for nearly 900 asteroids [59]; more details are given in sections 6.2 to 6.4 in Euclid Collaboration: Tereno et al. [46].

In total, ten distinct [SAA, AA] configurations were chosen to sample the range of thermal states encountered during the survey. Three configurations included both defocused and in-focus observations, the remaining seven comprised only in-focus observations. This mixed approach reduced the risk of vacuum-induced failures when actuating a mechanism [60]. It also limited the fraction of the M2M’s lifetime travel-distance budget consumed by PDC activities early in the mission. The entire PDC was scheduled as a single uninterrupted block of 2.5 months duration (Figure 5) after PV and before the start of survey operations, to have low and similar levels of molecular contamination across the data set (sections 5 and 6 in Euclid Collaboration: Schirmer et al. [43]).

Each of the three focus positions required 150 exposures of 289 s, dithered over a 100$^{\prime\prime}$-diameter circle. While 100 exposures would have delivered sufficient S/N, we were near the maximum of the 25th solar cycle, with powerful flares rendering the data from several VIS detectors unusable for PDC for several hours at a time. Given Euclid’s planning timescale of five weeks, a large concurrent flare could result in data loss, requiring a full repeat including thermal stabilisation. Based on the 2023 flare statistics and the associated data losses since launch, we increased the number of exposures per focus position to 150, lasting 18 h. This lowered the statistical risk of unrecoverable data loss to a few per cent, with the benefit of higher S/N in the absence of a flare.

PSF calibration fields

The PSF model predicts the PSF for every survey field, and it must be validated and monitored as the survey progresses. This is done via 23 dedicated PSF calibration fields, distributed across a range of ecliptic longitudes around both ecliptic poles, for maximum visibility at typical spacecraft attitudes [46]. Ground-based spectral energy distributions and Gaia photometry are required to characterise the stellar population in the magnitude range relevant for the EWS ($17.5 < I_{\mathrm{E}} < 21$).

Each field visit consists of 100 exposures of 289 s each, dithered along a Vogel spiral [91] to avoid clustering of dithers while at the same time achieving uniform sampling. The required S/N is five times higher than that of the EWS. The spacecraft’s attitude during a visit is matched to the mean attitude of the contemporary survey patch, so that the obtained PSF is representative. Observations are scheduled every 2–6 weeks, inserted into the observations of a contemporary patch at least three days after that patch began to allow for thermal equilibrium. Not every patch has an associated PSF-calibration visit (see also section 7.6 in Euclid Collaboration: Scaramella et al. [41]). If a visit fails, it cannot simply be rescheduled, since the patch and its associated thermal state will have moved on. In that case, a PSF calibration field observed at a similar thermal state is used as a substitute, or the PSF model itself is deemed sufficiently accurate for that state.

Limitations of DR1 and DR2 improvements

Continuous zero-point modelling

For DR1, we have implemented a piecewise linear fit of the zero point for VIS before the first decontamination, and a constant zero point following the second decontamination (section 2.2 Euclid Collaboration: McCracken et al. [31]). For NISP-P, we adopted piece-wise constant zero points over the entire period (Figure 7 in Euclid Collaboration: Polenta et al. 2026a). For DR2, we have sufficiently long baselines and a large volume of data processed with the same pipeline configuration. This will enable us to establish a finer and continuous interpolation of the zero points at any epoch in the survey.

Vector (PDF)

Network of absolute flux standards

The DR1 absolute flux calibration of VIS is indirectly tied to HST CALSPEC by calibration against Gaia DR3 $G$-band magnitudes. The $I_{\mathrm{E}}$ and $G$ passbands largely overlap, resulting in a well-defined transformation for stellar photometry as a function of Gaia $G_{\mathrm{BP}} - G_{\mathrm{RP}}$ colour (see figure 11 in Euclid Collaboration: Kümmel et al. [29]). For NISP-P, the absolute flux scale is tied to a hot WD secondary HST standard (J175318+644502, Sect. 5.2.2) in the self-calibration field, and three other CALSPEC standards.

Figure 6 shows that within the DR2 area, a considerable number of principal flux-calibration standards will become available. Among them are suitably faint CALSPEC standards [5, 4], the faint WDs of Narayan et al. [72], and blackbody-like helium WDs [87, 7]. Anchored to this network of standards, we expect Euclid’s absolute flux calibration to become traceable via an unbroken calibration chain linked to standards maintained by the National Institute of Standards and Technology (NIST; [76, 10, 77]), at the sub-per-cent level, and without requiring dedicated observations on our side.

Vector (PDF)

On the spectroscopic side, the situation is more difficult due to overlap of first-order spectrograms. Only a few HST CALSPEC standards are available with low enough order contamination from field crowding, and only at certain times of the year when the combination of position angle and dispersion directions are favourable for observing with a specific prism. Absolute spectrophotometric calibration of all prisms was obtained during PV using CALSPEC standard GRW+70 5824, and presumed to be temporally stable (AFX solution, section 4.3 in Euclid Collaboration: Fumana et al. [21]). For DR2 we need to account for time variability from detector contamination, in particular for the blue grism (Sect. 7.8). This can be obtained from secondary calibrators in the self-calibration data; spot-checks for the red-grism data can be done with brighter CALSPEC and blackbody standards [87, 7] in the DR2 and DR3 area.

Large-scale flat

Time dependence

The large-scale flat (LSF) is time-dependent due to variable molecular contamination of optical surfaces. For DR1 and VIS, a high-confidence, single LSF was used, constructed from five visits of the self-calibration field. For NISP, a monthly-built LSF was applied piece-wise without interpolation in time. This approach achieves relative photometric accuracies of 0.08%– 0.14% at the VIS quadrant level, and 0.9% globally for NISP (Table A.2), better than required.

For DR2, we will have a large data set processed with the same pipeline configuration, enabling the detection of more subtle effects. The VIS LSF exhibits spatially variable transmission changes on a monthly basis, with low spatial frequency and an amplitude of around 0.4%, presumably from slight contamination activity on folding mirror 3 (FoM3). Using monthly LSF builds, or an interpolating function to any epoch in the survey, will improve the relative photometric accuracy in DR2.

NISP is located upstream of FoM3 and is thus unaffected by its contamination. However, ice contamination is active on the NISP detectors, see Sect. 7.8 and section 5.4.2 in Euclid Collaboration: Schirmer et al. [43], and thus affects the LSF, together with charge-persistence effects (Sect. 7.5). Whether both effects can be isolated and a purely optical LSF can be constructed before the DR2 software freeze is currently unclear.

Chromaticity

The LSF is chromatic due to oblique angles of incidence on FoM1 and FoM2 with their hybrid coatings, and in particular on the dichroic beam splitter. The VIS LSF is built from Gaia stars, restricted in Gaia colour to $G_{\mathrm{BP}} - G_{\mathrm{RP}} = 0.2$–$2.5$ and to $I_{\mathrm{E}} = 18.5$–$19.8$. Since typical EWS galaxies are redder than Galactic stars (Figure 2), the ‘stellar’ LSF introduces a field-dependent bias in photometry. For a preliminary analysis, we divided the stellar sample with Gaia colours $G_{\mathrm{BP}} - G_{\mathrm{RP}} = 0.44$–$2.84$ into a blue, green, and red bin with equal number of sources, and recomputed the corresponding LSF. There is a maximum deviation of approximately 1.0% to 1.5% between the blue and red LSFs near the right edge of the VIS FPA, where we also observe a higher chromatic shift in the astrometric residuals between blue and red sources (see Figure 7 in Euclid Collaboration: McCracken et al. [31]). Since the LSF is an achromatic correction at the pixel level, we consider a fine-tuning of the allowed $G_{\mathrm{BP}} - G_{\mathrm{RP}}$ range that minimises the photometric bias for galaxies.

The NISP LSF is also subject to optical chromaticity. However, due to the well-defined bandpass filtering (Euclid Collaboration: Schirmer et al. [42]) and the longer wavelengths, we presume that this effect is considerably smaller than for VIS, and a correction might not be warranted for DR2.

VIS shutter correction

VIS employs a blade-type shutter mounted in the converging beam path ahead of the FPA (Euclid Collaboration: Cropper et al. [19]). The blade rotates about a hinge positioned near the FPA, with its leading edge sweeping through the convergent beam during closure. Due to the beam geometry, the effective exposure time exhibits a considerable nonlinear gradient along the FPA’s horizontal axis (see Figure 7).

The shutter operation introduces an exposure-time gradient of approximately 4.0 s across the FPA (Figure 7), causing a zero-point change of 0.8% for nominal exposures (560.5 s) and 4.7% for short exposures (89.5 s). The optical component of the LSF (Sect. 7.3) varies by around 1%; this is similar to the shutter effect in the nominal exposures, and dominated by the shutter in the short exposures.

The shutter effect is purely geometric, achromatic, very stable in time, and unaffected by ice contamination. In DR2 it will be removed prior to the computation of the LSF, thus separating mechanical from optical components and providing consistent treatment of exposures with different exposure times.

NISP persistence correction

The persistence behaviour from charge traps in the NISP detectors depends on the trap population in a pixel, its illumination and saturation history, the brightness of the local stimulus, and the ionisation mechanism. The cumulative persistence signal from the various traps in a pixel has an approximate power-law decay in time, with a nonlinear dependence on the factors mentioned above (Euclid Collaboration: Kubik et al. [27]). We find that overall the persistence response has decreased since launch, possibly because of repeated (super-)saturation by bright sources, which most pixels experience approximately every 2–3 months of survey operations. It appears that this passivates some charge traps on time-scales of months, an inoculation effect that we already saw in ground tests before launch; it has also been described for other detectors of similar architecture (McLeod & Smith [70]).

Before launch, we did consider to super-saturate the detectors with the calibration lamp (Euclid Collaboration: Hormuth et al. [24]). Done on a monthly basis in flight, this could have resulted in a reduced, controlled, and approximately stable persistence state, while also guarding against the spotty and unpredictable natural inoculation caused by bright sources. However, it is currently unclear what this inoculation actually does at the diode level in the detector. We decided to not use inoculation, because mitigating a little understood effect (persistence) with a procedure that is even less understood carried too many risks for the detectors and their overall calibration.

Because of this complexity, while a persistence correction is planned in the long run, DR1 relies on masking pixels affected by persistence. This masking uses an empirical power-law model obtained from PV data, accounting for up to five hours of illumination history. For NISP photometry images all spectroscopic observations from this past time interval are used for masking, with an additional Hough transform-based step to further detect and improve spectroscopy persistence streaks. For spectroscopic data both past photometric and spectroscopic observations are used for masking. The masking is conservative, resulting in a

fairly high masked pixel fraction of approximately 5% across the FPA. Details about persistence characterisation and masking are given in Euclid Collaboration: Kubik et al. [27].

For DR2, we are investigating several improvements, such as parameter optimisation to lower the masked pixel fraction to possibly 2%. For actual persistence correction, a more accurate model is needed that includes signal dependence, time evolution, and the different response to cosmic-ray hits.

NISP distortion models

The NISP filters have optical power (Euclid Collaboration: Jahnke et al. [25]) and thus contribute to the total optical distortion seen by the NISP detectors. The filter-wheel assembly does not reproduce filter positions perfectly, resulting in slight but measurable changes in the optical distortion between exposures. To account for this, the NIR-PF (Euclid Collaboration: Polenta et al. [37]) determines the camera distortion for every exposure individually.

In survey areas with very low source density, aggressive persistence masking can result in a critically low source density in two detectors that have higher persistence than the others. As a result, area fractions of 0.70%, 0.23%, and 0.14% of DR1 have an astrometric residual exceeding 0′′.05, 0′′.08, and 0′′.1, respectively. Formally, our requirement is that after all calibrations the total $1\sigma$-error in the measured position of an object must be smaller than 0.3 pixels or 0′′.09. The affected areas are typically 1′ to 3′ in size. Complementing the nominal SGS data products, we provide HEALPix maps (Górski et al. [53]) with the astrometric error as a function of sky position. A zoom-in into one affected tile is shown in Figure 8. Details are discussed in sections 3.6 and 4.1 of Euclid Collaboration: Radovich et al. [38].

This limitation does not affect the accuracy of photometric redshifts (Euclid Collaboration: Tucci et al. [47]), because the astrometric residuals are small compared to the merge data set (MER) aperture radii (Euclid Collaboration: Romelli et al. [39]) used for multi-wavelength photometry. In addition, the photo-$z$ pipeline allows for a few per cent of systematic flux uncertainty to accommodate any small patch-to-patch zero-point shifts that may exist, and therefore we expect a negligible impact on Euclid core science. However, photometry of individual compact sources that fall into affected areas – identifiable in the HEALpix maps – need to account for this effect in DR1. We have verified that an adjustment to the persistence-masking threshold for these two detectors resolves the issue, and therefore it will not be present in DR2.

NISP-S wavelength-calibration systematic errors

The DR1 spectra carry two small systematic wavelength errors of similar magnitude. They are at or below our requirements, but detectable in comparisons with reference spectra. Corrections are planned for DR2.

The first systematic effect is an uncorrected plate-scale change caused by differential velocity aberration. The bulk aberration offsets are absorbed when registering the dispersed images on the sky, by comparing the aberrated positions of the 0th-order spectrograms against aberration-free coordinates in the barycentric International Celestial Reference System (ICRS), accounting for shift and rotation but not for scale. Therefore, the plate-scale change remained, introducing a field-dependent error in the predicted position of the first order along the dispersion axis; this error varies with epoch and ecliptic coordinates. The same systematic entered the dispersion calibration, based on PV observations of a PN in October 2023, and was propagated to all DR1 fields. The combined effect is a zero-point error that shifts all wavelengths in the first-order spectrum of a source by the same amount $\Delta\lambda$. The shift varies linearly with the source position along the dispersion direction: it vanishes on the line through the field centre perpendicular to the dispersion direction and reaches its maximum at the field edges, around 0.5 pixel, that is 0.7 nm for the red grisms and 0.6 nm for the blue grism. In velocity, this corresponds to 110–175 km s$^{-1}$ for the red grisms and 140–205 km s$^{-1}$ for the blue grism, decreasing with wavelength across the first order. The maximum shift per field in DR1 is shown in the upper row of Figure 9. Because the shift is antisymmetric about the field centre, sources at opposite edges of the same field and exposure differ by up to twice this value. Grisms with opposite dispersion directions have shifts of opposite sign; in combined spectrograms the shifts therefore cancel only partially, owing to the dither offsets and masking, and spectral features are broadened.

Vector (PDF)

The second systematic effect arises because the PN emission-line wavelengths determined by Euclid Collaboration: Paterson et al. [35] were not Doppler-corrected from the barycentric frame to the spacecraft frame at the epoch of the PN observations, and because DR1 spectra are projected onto a wavelength grid in the spacecraft frame instead of the barycentric frame. Together, these scale all wavelengths by a common factor, so measured wavelengths must be corrected as $\lambda_{\mathrm{corr}} = (1+\delta)\lambda_{\mathrm{meas}}$, with $|\delta| \le1.5 \times10^{-4}$. The factor is constant per field and the same for all grisms. The resulting velocity offset, $-c\ \delta$, where $c$ is the speed of light, is shown in the lower row of Figure 9.

NISP contamination effects and correction

Growing water ice on the detectors’ anti-reflection (AR) coating stacks introduces a time-variable component to the detector quantum efficiency (QE) by means of optical interference. This affects predominantly the $Y_{\mathrm{E}}$ band and blue-grism data.

In DR1, photometric effects in the imaging data were corrected for – implicitly to first order – by the monthly relative flux calibration (section 3.2.2 in Euclid Collaboration: Polenta et al. [37]). The residual colour terms remained small enough to not violate the relative photometry calibration requirements (see Table A.2, and section 4.3 in Euclid Collaboration: Polenta et al. [36]). On the spectroscopic side, the relative and absolute flux calibration of the blue-grism data are based on PV observations and were not updated monthly, and thus the slow transmission modulation is fully present in the DR1 data. The red-grism data are susceptible to ice mostly at the blue end (1250 to 1350 nm) of their sensitivity range that extends to 1850 nm (see also section 4.3 in Euclid Collaboration: Fumana et al. [21]); we consider them weakly affected. The completeness and purity estimates for the spectroscopic reference sample (Sect. 2.5.1) are based on the detectability of emission lines, which degrades or improves depending on the phase of the interference fringe. This could cause notable differences when comparing samples selected in the EDF-N with those from the EDF-S, as the blue-grism observations of these fields are not scheduled at the same epochs.

The DR2 period will have maximum effects from interference. We estimate the following temporal and spatial flux variations in the uncalibrated LE1 data, that is before any processing:

• blue-grism data, 15 to 20% in the 930–1350 nm range,

• red-grism data, 3 to 7% in the 1250–1850 nm range,

• $Y_{\mathrm{E}}$ band, 0.08 to 0.12 mag, dependent on spectral energy distribution (SED),

• $J_{\mathrm{E}}$ band, 0.03 to 0.06 mag,

• $H_{\mathrm{E}}$ band, 0.01 to 0.03 mag.

These effects are too large to be corrected to within CalCD-A requirements (Table A.2) with the DR1 pipeline. The mitigation scheme that is currently developed is as follows: Using the complex refractive index of the H2RG substrate, we are building four-layer AR coating-stack models that approximate the measured QE for the uncontaminated detectors. Using the known refractive index of amorphous ice, we then infer the ice thickness and contamination rate using the monthly flat measurements at five wavelengths. Together, these yield models for the detectors’ effective QEs at any time in the survey. These can be used to recompute effective central wavelengths for the imaging passbands needed for photo-z estimates (Euclid Collaboration: Hartley et al. [23]), and explicit wavelength-dependent corrections for all grism data. This is work in progress, with a DR2 or DR3 perspective.

Photometric transformations between Euclid and each of Roman, JWST, and CSST

The large overlap between the Euclid and Roman surveys (Figure 6) will enable photometric consistency checks over an area of approximately 1000 deg$^{2}$. Using JWST data over smaller areas, the relative photometric accuracy of the nonlinearity corrections implemented by Euclid and JWST can be assessed over a wider dynamic range, because sources that appear faint for NISP are bright for JWST. To prepare for these tests, we need accurate reference transformations between these photometric systems and high-S/N data, that is, either brighter stars selected from a wide area, or data from the deep fields. Ideally suited are warm F-, G-, and K-type dwarfs, which form quasi-linear and well-defined sequences in broadband colour–colour space.

Cooler stars have strong molecular absorption bands causing photometric scatter, and in hotter stars the increasing ionisation of hydrogen changes the atmospheric opacity, making colour–colour relations nonlinear.

For these cross-comparisons we need to address two problems: deriving reference transformations from well-defined sequences of warm stars in synthetic colour–colour space; and identifying their stellar counterparts in the survey data to enable a direct photometric comparison across surveys. To quantify the effect of extinction, we considered two lines of sight: the EDF-S [Galactic coordinates $\ell= 256^\circ$, $b = -47^\circ$, $E(B - V) = 0.01$–0.02], which will be fully covered by both the medium and deep layers of the Roman High Latitude Wide Area Survey (HLWAS); and the EUDF [$\ell= 95^\circ$, $b = +30^\circ$, $E(B - V) = 0.04$ to 0.05], which contains the smaller JWST North Ecliptic Pole Extragalactic Unified Survey (NEXUS) field (Shen et al. [86]).

Stellar population synthesis

We simulated the Galactic stellar populations along the two lines of sight using the Besançon Galactic model⁷ (Robin et al. [82]), which describes the spatial density profiles of the thin and thick stellar discs, the bulge, and the halo. For each star the model provides the fundamental atmospheric parameters – effective temperature, surface gravity, metallicity, and stellar radius ($T_{\mathrm{eff}}$, $\log g$, [M/H], $R_*$) – the heliocentric distance $d$, and the extinction $A_V$ integrated along the line of sight to the stellar distance using the Marshall et al. [67] three-dimensional dust map. For each line of sight we performed a cone search with circular area of 1 deg$^{2}$ to a distance of 100 kpc, obtaining statistically representative stellar samples.

Our tests show that the line-of-sight extinction dependence introduces effects on the order of 1 mmag only. Therefore, we concatenated the two catalogues and conclude that the transformations shown further below are valid for the entire EWS area. The joint catalogue was filtered with $16.5 \le I_{\mathrm{E}} \le28.0$ to match the approximate brightness limits of deep Euclid data, retaining 30 700 of 30 367 stars.

Broadband photometry from synthetic spectra

We compute synthetic photometry by matching each simulated star to the nearest node of the Phoenix synthetic stellar atmosphere grid⁸ (Allard et al. [1]; Husser et al. [58]) in the 3-dimensional parameter space ($T_{\mathrm{eff}}$/1000 K, $\log g$, [M/H]/0.5). The Phoenix models provide the flux density $F_\lambda$ emitted by a stellar surface element with unit area 1 cm$^{2}$ into the hemisphere above, and the star’s luminosity density is then

At a distance $d$, the observer receives the extincted flux density

where $k(\lambda)$ is the extinction curve evaluated with dust_extinction (Gordon [51]) using the Gordon et al. [52] law with standard extinction parameter $R_V = 3.1$.

The apparent AB magnitude in a passband with total system throughput $T(\lambda)$ is

For $T(\lambda)$ we use Euclid Collaboration: Schirmer et al. [42] for NISP, the mission data base for VIS, and sedpy⁹ for Roman and JWST; updated throughput curves for CSST were kindly provided by Liu [63], including the wide WVCSST and WICSST filters that replaced yCSST. For respective details about JWST, Roman, and CSST passbands we refer to Rieke et al. [81], Switzer et al. [88], and Luo et al. [65].

Rejection of cool stars

There is a great abundance of cool red dwarfs with molecular absorption bands from TiO and H$_2$O, producing considerable intrinsic photometric scatter, which is problematic for photometric comparison purposes. We used the spectral signature of these molecules to select against cool stars in real survey data. TiO forms in stellar photospheres below 4000 K (spectral type K7V and later, see Pecaut & Mamajek [75]) and strongly absorbs across the $I_E$ band. Absorption by H$_2$O becomes prominent below 3000 K (spectral type M5V and later), with its strongest features appearing above 1.3 µm and below that wavelength only in much colder L-type dwarfs. The $Y_E$-band flux is therefore mostly unaffected by both molecules, and the TiO absorption in the $I_E$ band drives the $I_E-Y_E$ colour to large positive values for cool red dwarfs.

Using the synthetic photometry from Eq. (3), we find that a single threshold $I_E-Y_E < 0.69$ provides a clean selection against stars cooler than 4000 K: applied to our noise-free, synthetic catalogue it rejects 92 % of stars with $T_{\rm eff} < 4000$ K while retaining 97 % of warmer FGK-types. Adding the redder $J_E$ and $H_E$ bands to form a multi-colour cut does not improve on this. A more conservative threshold of $I_E-Y_E < 0.50$ rejects 97.4% of stars colder than 4500 K, and retains 91.5% of stars warmer than that; we did not use this setting because of its lower completeness.

Computing the reference transformations

We fit second-order polynomials to the FGK-type sequence, unless a first-order fit provides an equally good description. Only stars that fall within $T_{\rm eff}=4500$ K to $T_{\max}$ are included in the fit. The upper limit, $T_{\max}$, was fine-tuned for every transformation because – depending on the curvature of the FGK sequence – including hotter stars resulted in a better fit, or hotter stars simply extended the linear FGK sequence and there was no reason to exclude them. The fit itself was performed iteratively using a $2.5\sigma$ outlier rejection until convergence was achieved.

An example plot is shown in Figure 10, all other plots in Figs. B.2 to B.7, and Table B.1 has all transformation equations. We validated our code by comparing selected Euclid-JWST transformations against those published by Durbin et al. [11] in the Vega system. The slopes of the fits agree to within a few per cent. The absolute values could not be directly compared, because Durbin et al. [11] used a non-standard AB renormalisation for Vega and JWST/NIRCam; after recomputing the standard AB conversion, the agreement on the absolute scale to 8% or better. Recomputing the transformations for 2MASS (Cutri et al. [8]), we do not find considerable qualitative or quantitative differences with our earlier work (appendix C in Euclid Collaboration: Schirmer et al. [42]); there, we computed linear transformations for spectral types O to K to have an all-sky reference system for early commissioning work. Transformations to 2MASS have limited value due to the very small dynamic-range overlap with Euclid, and therefore we do not reproduce them here.

Summary and conclusions

We have presented the development, implementation, observations, and performance of the Euclid mission calibration plan, starting out with the top-level calibration requirements derived from the scientific requirements. The calibration plan is centred on the self-calibration principle, is fully adequate, and required only a few minor modifications since launch.

Most top-level spectrophotometric requirements are already met by DR1. The assembly of a 99% pure and complete spectroscopic redshift reference sample requires more observations, and building of the VIS PSF model is still ongoing because of the considerable complexity of the optical system. Formally, all requirements must be met by DR3.

We have summarised the most important upcoming changes for DR2 that will lead to considerable improvements, beginning with VIS shutter correction, more accurate absolute flux calibration, better handling of persistence and contamination effects in NISP, and improved wavelength calibration. Lastly, we have computed photometric transformation equations between Euclid on one side, and Roman, JWST, and CSST on the other side.

Acknowledgements. 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 BMMI, 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 Netherlandse 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/). 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. Mischa Schirmer thanks the many people in the EC and at ESA who tirelessly contributed to developing and implementing the Euclid calibration strategy over more than a decade. Working with you has been a privilege, shaped by your kindness, expertise, dedication, and rigorous scrutiny, all in pursuit of a single common goal.

Authors and affiliations

Euclid Collaboration: M. Schirmer$^{\star1}$, K. Paterson$^{1}$, A. Alavi$^{2}$, B. Altieri$^{3}$, P. N. Appleton$^{2}$, N. Auricchio$^{4}$, H. Aussel$^{5}$, R. Azzollini$^{6}$, P. Battaglia$^{4}$, R. Barbier$^{7}$, L. Betancourt$^{8}$, D. R. Boxhoorn$^{9}$, J. Brinchmann$^{10, 11, 12}$, G. Buenadicha$^{3}$, G. P. Candini$^{6}$, J. M. Carrasco$^{13, 14, 15}$, P. Casenove$^{16}$, S. Conseil$^{17, 18}$, Y. Copin$^{18}$, J.-C. Cuillandre$^{5}$, J. E. Davies$^{1}$, S. Davini$^{19}$, S. E. Deustua$^{20}$, J. Dinis$^{21, 22}$, C. Dolding$^{6}$, X. Dupac$^{3}$, S. Dusini$^{23}$, J. Martín-Fleitas$^{24}$, M. Frailis$^{25}$, E. Franceschi$^{4}$, P.-A. Frugier$^{5}$, M. Fumana$^{26}$, L. Gabarra$^{27}$, W. Gillard$^{28}$, P. Gómez-Alvarez$^{29, 3}$, C. Grenet$^{30}$, J. Gracia-Carpio$^{31}$, A. Gregorio$^{32, 25, 33}$, F. Grupp$^{31, 34}$, L. Guzzo$^{35, 36, 37}$, W. G. Hartley$^{38}$, S. Hemmati$^{2}$, O. Herent$^{30}$, J. Hoar$^{3}$, H. Hoekstra$^{39}$, W. Holmes$^{40}$, K. Jahnke$^{1}$, J. J. E. Kajava$^{41, 42, 43}$, F. Keller$^{44}$, R. Kohley$^{3}$, B. Kubik$^{18}$, M. Kümmel$^{34}$, K. Kuijken$^{39}$, C. Laigle$^{30}$, R. Laureijs$^{9}$, V. Le Brun$^{45}$, G. Libet$^{16}$, P. Liebing$^{6}$, T. Maciaszek$^{16}$, R. J. Massey$^{46}$, D. C. Masters$^{47}$, H. J. McCracken$^{30}$, E. Medinaceli$^{4}$, J. Mendes$^{48}$, L. Miller$^{27}$, A. Mora$^{49}$, P. W. Morris$^{8}$, S. Mottet$^{30}$, R. Nakajima$^{50}$, D. Neumann$^{39}$, H. N. Nguyen-Kim$^{30}$, J. W. Nightingale$^{51}$, K. Okumura$^{52}$, S. Paltani$^{38}$, W. J. Percival$^{53, 54, 55}$, G. Polenta$^{56}$, M. Portaluppi$^{44}$, G. D. Racca$^{39, 44}$, M. Radovich$^{57}$, I. Risso$^{58, 19, 36}$, E. Romelli$^{25}$, M. Ruffenach$^{16}$, S. Awan$^{6}$, R. Scaramella$^{59}$, C. Scarlata$^{60}$, M. Schmidt$^{48}$, M. Scodeggio$^{26}$, A. Secroun$^{28}$, G. Seidel$^{1}$, A. D. Short$^{44}$, C. Sirignano$^{61, 23}$, J. Skottfelt$^{62}$, L. C. Smith$^{63}$, F. Soldano$^{30}$, S. Szmolka$^{44}$, M. Talia$^{64, 4}$, A. N. Taylor$^{65}$, I. Tereno$^{21, 22}$, E. A. Valentijn$^{9}$, R. Vavrek$^{3}$, L. M. Gaspar Venancio$^{44}$, G. Verdoes Kleijn$^{9}$, A. Zacchei$^{25, 66}$, S. Andreon$^{36}$, C. Baccigalupi$^{66, 25, 33, 67}$, M. Baldi$^{68, 4, 69}$, A. Balestra$^{57}$, S. Bardelli$^{4}$, A. N. Belikov$^{9, 70}$, R. Bender$^{31, 34}$, A. Biviano$^{25, 66}$, A. Bonchi$^{56}$, E. Branchini$^{58, 19, 36}$, M. Brescia$^{71, 72}$, S. Camera$^{73, 74, 75}$, G. Cañas-Herrera$^{39}$, V. Capobianco$^{75}$, C. Carbone$^{26}$, J. Carretero$^{76, 77}$, F. J. Castander$^{78, 15}$, M. Castellano$^{59}$, G. Castignani$^{4}$, S. Cavuoti$^{72, 79}$, A. Cimatti$^{80}$, C. Colodro-Conde$^{81}$, G. Congedo$^{65}$, C. J. Conselice$^{82}$, L. Conversi$^{83, 3}$, F. Courbin$^{14, 84, 85}$, H. M. Courtois$^{7}$, M. Cropper$^{6}$, A. Da Silva$^{21, 22}$, R. da Silva$^{59, 56}$, H. Degaudenzi$^{38}$, G. De Lucia$^{25}$, H. Dole$^{86}$, F. Dubath$^{38}$, C. A. J. Duncan$^{65}$, S. Escoffier$^{28}$, M. Farina$^{87}$, R. Farinelli$^{4}$, S. Farrens$^{5}$, F. Faustini$^{59}$, S. Ferriol$^{18}$, N. Fourmanoit$^{28}$, S. Galeotta$^{25}$, K. George$^{88}$, B. Gillis$^{65}$, C. Giocoli$^{4, 69}$, B. R. Granett$^{36}$, A. Grazian$^{57}$, S. V. H. Haugan$^{89}$, I. M. Hook$^{9}$, F. Hormuth$^{90}$, A. Hornstrup$^{91}$, M. Huertas-Company$^{81, 94, 95}$, M. Jhabvala$^{96}$, S. Kermiche$^{28}$, A. Kiessling$^{40}$, M. Kunz$^{97}$, H. Kurki-Suonio$^{98, 99}$, A. M. C. Le Brun$^{100}$, D. Le Mignant$^{45}$, S. Lise$^{75}$, P. B. Lilje$^{89}$, V. Lindholm$^{98, 99}$, I. Lloro$^{101}$, M. Magliocchetti$^{87}$, G. Mainetti$^{102}$, O. Mansutti$^{25}$, S. Marcin$^{103}$, O. Marggraf$^{50}$, K. Markovic$^{40}$, M. Martinelli$^{59, 104}$, N. Martinet$^{45}$, F. Marulli$^{64, 4, 69}$, N. Mauri$^{80, 69}$, S. Mei$^{105, 106}$, M. Melchior$^{107}$, M. Meneghetti$^{4, 69}$, E. Merlin$^{57}$, G. Meylan$^{108}$, P. Monaco$^{32, 25, 33, 66}$, M. Moresco$^{64, 4}$, C. Moretti$^{25, 66, 33}$, L. Moscardini$^{64, 4, 69}$, C. Neissner$^{109, 77}$, R. C. Nichol$^{110}$, S.-M. Niemi$^{44}$, C. Padilla$^{109}$, F. Pasian$^{25}$, V. Pettorino$^{44}$, A. Pezzotta$^{36}$, S. Pires$^{5}$, M. Poncet$^{16}$, L. A. Popa$^{111}$, F. Raison$^{31}$, A. Refregier$^{61, 23, 4}$, A. Renzi$^{1}$, J. Rhodes$^{40}$, G. Riccio$^{72}$, F. Rizzo$^{25}$, M. Roncarelli$^{5}$, C. Rosset$^{112}$, B. Rusholme$^{2}$, R. Saglia$^{113, 114, 115}$, D. Sakr$^{113, 114, 115}$, A. G. Sánchez$^{31}$, D. Sapone$^{116}$, B. Sartoris$^{117}$, P. Schneider$^{50}$, T. Schrabback$^{118}$, E. Sefusatti$^{25, 66, 33}$, M. Seiffert$^{40}$, E. Sihvola$^{119}$, P. Simon$^{50}$, G. Sirri$^{69}$, L. Stanco$^{23}$, P. Tallada-Crespí$^{76, 77}$, H. I. Teplitz$^{47}$, N. Tessore$^{6}$, S. Toft$^{120, 121}$, R. Toledo-Moreo$^{122, 123}$, F. Torradeflot$^{77, 76}$, A. Tsyganov$^{124}$, I. Tutusaus$^{78, 15, 114}$, L. Valenziano$^{4, 125}$, J. Valivita$^{98, 99}$, T. Vassallo$^{25, 88}$, A. Veropalumbo$^{36, 19, 58}$, Y. Wang$^{2}$, J. Weller$^{34, 31}$, O. R. Williams$^{124}$, G. Zamorani$^{4}$, F. M. Zerbi$^{36}$, E. Zucca$^{4}$, I. T. Andika$^{34}$, M. Ballardini$^{126, 127, 4}$, P. Bergamini$^{4}$, S. Borgani$^{32, 66, 25, 33, 128}$, E. Bozzo$^{38}$, C. Burigana$^{129}$, R. Cabanac$^{114}$, M. Calabrese$^{130, 26}$, A. Cappi$^{117}$, F. Caro$^{59}$, T. Castro$^{121, 25, 66}$, D. Di Ferdinando$^{69}$, J. A. Escartin$^{70}$, G. Fabbian$^{86}$, J. García-Bellido$^{113}$, T. Gasparotto$^{59}$, V. Gautard$^{52}$, C. M. Gutierrez$^{81, 132}$, A. Hall$^{65}$, Y. Kang$^{118}$, A. Loureiro$^{133, 134}$, J. Macias-Perez$^{135}$, R. Maoli$^{136, 59}$, R. B. Metcalf$^{64, 4}$, M. Pöntinen$^{98}$, E. Sarpa$^{25}$, V. Scottez$^{137, 138}$, M. Sereno$^{4, 69}$, K. Tanidis$^{139}$, M. Tenti$^{69}$, M. Tucci$^{38}$, M. Viel$^{66, 25, 67, 33, 128}$, M. Wiesmann$^{89}$, J. A. Acevedo Barroso$^{40}$, Y. Akrami$^{113, 140}$, G. Angora$^{72, 126}$, S. Anselmi$^{23, 61, 141}$, M. Archidiacono$^{35, 37}$, G. Aricò$^{69}$, F. Atrio-Barandela$^{142}$, M. Aubert$^{18}$, E. Aubourg$^{105, 143}$, M. Baes$^{144}$, L. Bazzanini$^{126, 4}$, J. R. Bermejo-Climent$^{145, 146}$, D. Bertacca$^{147, 57, 23}$, M. Bethermin$^{147}$, F. Beutler$^{65}$, A. Blanchard$^{114}$, F. Bouchet$^{148, 79}$, M. L. Brown$^{82}$, S. Bruton$^{8}$, A. Calabaro$^{59}$, B. Camacho Quevedo$^{66, 67, 25}$, P. Carrilho$^{149}$, C. S. Carvalho$^{150}$, G. Charles$^{45}$, F. Cogato$^{4}$, T. E. Collett$^{151}$, A. R. Cooray$^{152}$, P. Corcho-Caballero$^{9}$, M. Correnti$^{59, 56}$, B. Csizi$^{118}$, O. Cucciati$^{4}$, T. de Boer$^{153}$, F. De Paolis$^{154, 155, 156}$, G. Desprez$^{9}$, A. Díaz-Sánchez$^{157}$, S. Di Domizio$^{58, 19}$, J. M. Diego$^{158}$, V. Duret$^{28}$, Y. Fang$^{34}$, A. Farina$^{36, 19}$, A. M. N. Ferguson$^{65}$, A. Finoguenov$^{98}$, F. Fontanot$^{25, 66}$, A. Franco$^{154, 155, 156}$, A. Fumagalli$^{25}$, Y. Fu$^{39, 9}$, K. Ganga$^{105}$, R. Gavazzi$^{45, 30}$, E. Gaztanaga$^{78, 15, 151}$, Z. Ghaffari$^{25, 66}$, F. Giacomini$^{69}$, F. Gianotti$^{4}$, G. Gozaliasl$^{159, 98}$, A. Gruppuso$^{4, 69}$, M. Guidi$^{68, 4}$, N. A. Hatch$^{160}$, C. Hernández-Monteagudo$^{132, 81}$, T. Heuberger$^{103}$, H. Hildebrandt$^{161}$, J. Hjorth$^{162}$, M. Jauzac$^{163, 46, 164, 165, 114}$, V. Kansal$^{166, 167}$, D. Karagiannis$^{168, 126, 169}$, J. Kim$^{27}$, C. Kitching$^{119}$, K. Koyama$^{146, 145}$, L. Kovač$^{144}$, K. Koyama$^{151}$, S. Kruk$^{3}$, M. C. Lam$^{65}$, M. Lattanzi$^{127}$, F. Leclercq$^{30}$, S. Lee$^{170, 40}$, J. Le Graet$^{28}$, L. Legrand$^{171, 86}$, M. Lembo$^{30}$, M. D. Lepinzan$^{25}$, G. Leroy$^{163, 46}$, J. Lesgourgues$^{172}$, T. I. Liaudat$^{143}$, L. Linke$^{118}$, S. J. Liu$^{87}$, G. Maggio$^{25}$, A. Manjón-García$^{157}$, F. Mannucci$^{173}$, F. R. Marleau$^{118}$, C. J. A. P. Martins$^{174, 10}$, M. Miluzio$^{3, 175}$.

Article number, page 18 G. Morgante⁴, S. Nadathur¹⁵¹, K. Naidoo^151,1, P. Natoli^126,127, A. Navarro-Alsina⁵⁰, A. Nersesian^176,144, S. Nesseris¹¹³, L. Nicastro⁴, F. Oppizzi^19,23,61, F. Pace^73,74,75,22, D. Paoletti^4,125, F. Passalacqua²³, L. Patrizii⁶⁹, C. Pattison¹⁵¹, A. Pisani²⁸, L. Pizzuti¹⁷⁷, D. Potter¹⁷⁸, M. Poulain¹⁷⁹, G. W. Pratt⁵, S. Quai^64,4, G. Rodighiero^61,57, W. Roster³¹, M. Sahlén¹⁸⁰, D. B. Sanders¹⁵³, A. Schneider¹⁷⁸, D. Sciotti^59,104, E. Sellentin^181,39, S. Serjeant¹⁸², F. Shankar¹⁸³, A. Shulevski^184,9,185, R. E. Smith¹⁸⁶, E. Soubrie⁸⁶, I. Szapudi¹⁵³, G. Testera¹⁹, R. Teyssier¹⁸⁷, S. Tosi^58,36,19, A. Troja²⁵, C. Uhlemann^188,51, M. Urbano^160,147, C. Valieri⁶⁹, A. Venhola¹⁷⁹, D. Vergani⁴, E. Vilenius⁶, S. Vinciguerra⁴⁵, M. von Wietersheim-Kramsta^46,163, N. A. Walton⁶³, L. Wang^189,9, A. H. Wright¹⁶¹, H. W. Yeung⁶⁵, L. Zalesky¹⁵³, and D. Scott¹⁹⁰

¹ Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany

² Caltech/IPAC, 1200 E. California Blvd., Pasadena, CA 91125, USA

³ ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain

⁴ INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy

⁵ Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France

⁶ Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK

⁷ Université Lyon 1, CNRS/IN2P3, IUF, IP2I Lyon, Villeurbanne, France

⁸ California Institute of Technology, 1200 E California Blvd, Pasadena, CA 91125, USA

⁹ Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands

¹⁰ Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal

¹¹ Faculdade de Ciências da Universidade do Porto, Rua do Campo de Alegre, 4150-007 Porto, Portugal

¹² European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany

¹³ Departament de Física Quàntica i Astrofísica (FQA), Universitat de Barcelona (UB), Martí i Franquès 1, E-08028 Barcelona, Spain

¹⁴ Institut de Ciències del Cosmos (ICCUB), Universitat de Barcelona (IEEC-UB), Martí i Franquès 1, 08028 Barcelona, Spain

¹⁵ Institut d’Estudis Espacials de Catalunya (IEEC), Edifici RDIT, Campus UPC, 08860 Castelldefels, Barcelona, Spain

¹⁶ Centre National d’Etudes Spatiales – Centre spatial de Toulouse, 18 avenue Edouard Belin, 31401 Toulouse Cedex 9, France

¹⁷ Centre de Recherche Astrophysique de Lyon, UMR5574, CNRS, Université Claude Bernard Lyon 1, ENS de Lyon, 69230, Saint-Genis-Laval, France

¹⁸ Université Lyon 1, CNRS/IN2P3, IP2I Lyon, Villeurbanne, France

¹⁹ INFN-Sezione di Genova, Via Dodecaneso 33, 16146, Genova, Italy

²⁰ National Institute of Standards and Technology, 100 Bureau Dr., Gaithersburg, MD 20899, USA

²¹ Departamento de Física, Faculdade de Ciências, Universidade de Lisboa, Edifício C8, Campo Grande, PT1749-016 Lisboa, Portugal

²² Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal

²³ INFN-Padova, Via Marzolo 8, 35131 Padova, Italy

²⁴ Aurora Technology for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

²⁵ INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy

²⁶ INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy

²⁷ Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, UK

²⁸ Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France

²⁹ FRACTAL S.L.N.E., calle Tulipán 2, Portal 13 1A, 28231, Las Rozas de Madrid, Spain

³⁰ Institut d’Astrophysique de Paris, UMR 7095, CNRS, and Sorbonne Université, 98 bis boulevard Arago, 75014 Paris, France

³¹ Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany

³² Dipartimento di Fisica - Sezione di Astronomia, Università di Trieste, Via Tiepolo 11, 34131 Trieste, Italy

³³ INFN, Sezione di Trieste, Via Valerio 2, 34127 Trieste TS, Italy

³⁴ Universitäts-Sternwarte München, Fakultät für Physik, Ludwig-Maximilians-Universität München, Scheinerstr. 1, 81679 München, Germany

³⁵ Dipartimento di Fisica "Aldo Pontremoli", Università degli Studi di Milano, Via Celoria 16, 20133 Milano, Italy

³⁶ INAF-Osservatorio Astronomico di Brera, Via Brera 28, 20122 Milano, Italy

³⁷ INFN-Sezione di Milano, Via Celoria 16, 20133 Milano, Italy

³⁸ Department of Astronomy, University of Geneva, ch. d’Ecogia 16, 1290 Versoix, Switzerland

³⁹ Leiden Observatory, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands

⁴⁰ Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA

⁴¹ Department of Physics and Astronomy, Vesilinnantie 5, University of Turku, 20014 Turku, Finland

⁴² Finnish Centre for Astronomy with ESO (FINCA), Quantum, Vesilinnantie 5, University of Turku, 20014 Turku, Finland

⁴³ Serco for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

⁴⁴ European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands

⁴⁵ Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France

⁴⁶ Department of Physics, Institute for Computational Cosmology, Durham University, South Road, Durham, DH1 3LE, UK

⁴⁷ Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA

⁴⁸ European Space Agency/ESOC, Robert-Bosch-Str. 5, 64293 Darmstadt, Germany

⁴⁹ Telespazio UK S.L. for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

⁵⁰ Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany

⁵¹ School of Mathematics, Statistics and Physics, Newcastle University, Herschel Building, Newcastle-upon-Tyne, NE1 7RU, UK

⁵² CEA Saclay, DFR/IRFU, Service d’Astrophysique, Bat. 709, 91191 Gif-sur-Yvette, France

⁵³ Waterloo Centre for Astrophysics, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada

⁵⁴ Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada 55 Perimeter Institute for Theoretical Physics, Waterloo, Ontario N2L 2Y5, Canada

56 Space Science Data Center, Italian Space Agency, via del Politecnico snc, 00133 Roma, Italy

57 INAF-Osservatorio Astronomico di Padova, Via dell’Osservatorio 5, 35122 Padova, Italy

58 Dipartimento di Fisica, Università di Genova, Via Dodecaneso 33, 16146, Genova, Italy

59 INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy

60 Minnesota Institute for Astrophysics, University of Minnesota, 116 Church St SE, Minneapolis, MN 55455, USA

61 Dipartimento di Fisica e Astronomia "G. Galilei", Università di Padova, Via Marzolo 8, 35131 Padova, Italy

62 Centre for Electronic Imaging, Open University, Walton Hall, Milton Keynes, MK7 6AA, UK

63 Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK

64 Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, via Piero Gobetti 93/2, 40129 Bologna, Italy

65 Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK

66 IFPU, Institute for Fundamental Physics of the Universe, via Beirut 2, 34151 Trieste, Italy

67 SISSA, International School for Advanced Studies, Via Bonomea 265, 34136 Trieste TS, Italy

68 Dipartimento di Fisica e Astronomia, Università di Bologna, Via Gobetti 93/2, 40129 Bologna, Italy

69 INFN-Sezione di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy

70 ATG Europe BV, Huygensstraat 34, 2201 DK Noordwijk, The Netherlands

71 Department of Physics "E. Pancini", University Federico II, Via Cinthia 6, 80126, Napoli, Italy

72 INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy

73 Dipartimento di Fisica, Università degli Studi di Torino, Via P. Giuria 1, 10125 Torino, Italy

74 INFN-Sezione di Torino, Via P. Giuria 1, 10125 Torino, Italy

75 INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese (TO), Italy

76 Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Avenida Complutense 40, 28040 Madrid, Spain

77 Port d’Informació Científica, Campus UAB, C. Albareda s/n, 08193 Bellaterra (Barcelona), Spain

78 Institute of Space Sciences (ICE, CSIC), Campus UAB, Carrer de Can Magrans, s/n, 08193 Barcelona, Spain

79 INFN – Sezione di Napoli, Via Cinthia 6, 80126, Napoli, Italy

80 Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Università di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy

81 Instituto de Astrofísica de Canarias, E-38205 La Laguna, Tenerife, Spain

82 Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK

83 European Space Agency/ESRIN, Largo Galileo Galilei 1, 00044 Frascati, Roma, Italy

84 Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig de Lluís Companys 23, 08010 Barcelona, Spain

85 Institut de Ciencies de l’Espai (IEEC-CSIC), Campus UAB, Carrer de Can Magrans, s/n Cerdanyola del Vallès, 08193 Barcelona, Spain

86 Université Paris-Saclay, CNRS, Institut d’astrophysique spatiale, 91405, Orsay, France

87 INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy

88 University Observatory, LMU Faculty of Physics, Scheinerstr. 1, 81679 Munich, Germany

89 Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway

90 School of Physics and Astronomy, Lancaster University, Lancaster, LA1 4YB, UK

91 Felix Horthuth Engineering, Goethestr. 17, 69181 Leimen, Germany

92 Technical University of Denmark, Elektrovej 327, 2800 Kgs. Lyngby, Denmark

93 Cosmic Dawn Center (DAWN), Denmark

94 Université PSL, Observatoire de Paris, Sorbonne Université, CNRS, LERMA, 75014, Paris, France

95 Université Paris-Cité, 5 Rue Thomas Mann, 75013, Paris, France

96 NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

97 Université de Genève, Département de Physique Théorique and Centre for Astroparticle Physics, 24 quai Ernest-Ansermet, CH-1211 Genève 4, Switzerland

98 Department of Physics, P.O. Box 64, University of Helsinki, 00014 Helsinki, Finland

99 Helsinki Institute of Physics, Gustaf Hällström katu 2, University of Helsinki, 00014 Helsinki, Finland

100 Laboratoire d’étude de l’Univers et des phénomènes eXtrêmes, Observatoire de Paris, Université PSL, Sorbonne Université, CNRS, 92190 Meudon, France

101 SKAO, Jodrell Bank, Lower Withington, Macclesfield SK11 9FT, UK

102 Centre de Calcul de l’IN2P3/CNRS, 21 avenue Pierre de Coubertin 69627 Villeurbanne Cedex, France

103 University of Applied Sciences and Arts of Northwestern Switzerland, School of Computer Science, 5210 Windisch, Switzerland

104 INFN-Sezione di Roma, Piazzale Aldo Moro, 2 - c/o Dipartimento di Fisica, Edificio G. Marconi, 00185 Roma, Italy

105 Université Paris Cité, CNRS, Astroparticule et Cosmologie, 75013 Paris, France

106 CNRS-UCB International Research Laboratory, Centre Pierre Binétruy, IRL2007, CPB-IN2P3, Berkeley, USA

107 University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland

108 Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Observatoire de Sauverny, 1290 Versoix, Switzerland

109 Institut de Física d’Altes Energies (IFAE), The Barcelona Institute of Science and Technology, Campus UAB, 08193 Bellaterra (Barcelona), Spain

110 School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK

111 Institute of Space Science, Str. Atomistilor, nr. 409 Măgurele, Ilfov, 077125, Romania

112 Institute for Particle Physics and Astrophysics, Dept. of Physics, ETH Zurich, Wolfgang-Pauli-Strasse 27, 8093 Zurich, Switzerland

113 Instituto de Física Teórica UAM-CSIC, Campus de Cantoblanco, 28049 Madrid, Spain

114 Institut de Recherche en Astrophysique et Planétologie (IRAP), Université de Toulouse, CNRS, UPS, CNES, 14 Av. Edouard Belin, 31400 Toulouse, France

115 Université St Joseph, Faculty of Sciences, Beirut, Lebanon

116 Departamento de Física, FCFM, Universidad de Chile, Blanco Encalada 2008, Santiago, Chile 117 Université Côte d’Azur, Observatoire de la Côte d’Azur, CNRS, Laboratoire Lagrange, Bd de l’Observatoire, CS 34229, 06304 Nice cedex 4, France

118 Universität Innsbruck, Institut für Astro- und Teilchenphysik, Technikerstr. 25/8, 6020 Innsbruck, Austria

119 Department of Physics and Helsinki Institute of Physics, Gustaf Hällströmin katu 2, University of Helsinki, 00014 Helsinki, Finland

120 Cosmic Dawn Center (DAWN)

121 Niels Bohr Institute, University of Copenhagen, Jagtvej 128, 2200 Copenhagen, Denmark

122 Universidad Politécnica de Cartagena, Departamento de Electrónica y Tecnología de Computadoras, Plaza del Hospital 1, 30202 Cartagena, Spain

123 European University of Technology EUt+, European Union

124 Centre for Information Technology, University of Groningen, P.O. Box 11044, 9700 CA Groningen, The Netherlands

125 INFN-Bologna, Via Irnerio 46, 40126 Bologna, Italy

126 Dipartimento di Fisica e Scienze della Terra, Università degli Studi di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy

127 Istituto Nazionale di Fisica Nucleare, Sezione di Ferrara, Via Giuseppe Saragat 1, 44122 Ferrara, Italy

128 ICSC - Centro Nazionale di Ricerca in High Performance Computing, Big Data e Quantum Computing, Via Magnanelli 2, Bologna, Italy

129 INAF, Istituto di Radioastronomia, Via Piero Gobetti 101, 40129 Bologna, Italy

130 Astronomical Observatory of the Autonomous Region of the Aosta Valley (OAVdA), Loc. Lignan 39, I-11020, Nus (Aosta Valley), Italy

131 Department of Mathematical Physics, Institute of Physics, University of São Paulo, R. do Matão 1371, 05508-090, São Paulo, SP, Brazil

132 Universidad de La Laguna, Dpto. Astrofísica, E-38206 La Laguna, Tenerife, Spain

133 Oskar Klein Centre for Cosmoparticle Physics, Department of Physics, Stockholm University, Stockholm, SE-106 91, Sweden

134 Astrophysics Group, Blackett Laboratory, Imperial College London, London SW7 2AZ, UK

135 Univ. Grenoble Alpes, CNRS, Grenoble INP, LPSC-IN2P3, 53, Avenue des Martyrs, 38000 Grenoble, France

136 Dipartimento di Fisica, Sapienza Università di Roma, Piazzale Aldo Moro 2, 00185 Roma, Italy

137 Institut d’Astrophysique de Paris, 98bis Boulevard Arago, 75014, Paris, France

138 ICL, Junia, Université Catholique de Lille, LITL, 59000 Lille, France

139 Center for Astrophysics and Cosmology, University of Nova Gorica, Nova Gorica, Slovenia

140 CERCA/ISO, Department of Physics, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA

141 Laboratoire Univers et Théorie, Observatoire de Paris, Université PSL, Université Paris Cité, CNRS, 92190 Meudon, France

142 Departamento de Física Fundamental. Universidad de Salamanca. Plaza de Merced s/n. 37008 Salamanca, Spain

143 IRFU, CEA, Université Paris-Saclay 91191 Gif-sur-Yvette Cedex, France

144 Universiteit Gent, Department of Physics and Astronomy, Proeftuinstraat 86 N3, 9000 Ghent, Belgium

145 Konkoly Observatory, HUN-REN CSFK, MTA Centre of Excellence, Budapest, Konkoly Thege Miklós út 15-17. H-1121, Hungary

146 MTA-CSFK Lendület Large-Scale Structure Research Group, Konkoly-Thege Miklós út 15-17, H-1121 Budapest, Hungary

147 Université de Strasbourg, CNRS, Observatoire astronomique de Strasbourg, UMR 7550, 67000 Strasbourg, France

148 Scuola Superiore Meridionale, Via Mezzocannone 4, 80138, Napoli, Italy

149 Department of Physics, Astronomy and Mathematics, University of Hertfordshire, College Lane, Hatfield AL10 9AB, UK

150 Instituto de Astrofísica e Ciências do Espaço, Faculdade de Ciências, Universidade de Lisboa, Tapada da Ajuda, 1349-018 Lisboa, Portugal

151 Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth PO1 3FX, UK

152 Department of Physics & Astronomy, University of California Irvine, Irvine CA 92697, USA

153 Institute for Astronomy, University of Hawai'i, 2680 Woodlawn Drive, Honolulu, HI 96822, USA

154 Department of Mathematics and Physics E. De Giorgi, University of Salento, Via per Arnesano, CP-193, 73100, Lecce, Italy

155 INFN, Sezione di Lecce, Via per Arnesano, CP-193, 73100, Lecce, Italy

156 INAF-Sezione di Lecce, c/o Dipartimento Matematica e Fisica, Via per Arnesano, 73100, Lecce, Italy

157 Departamento Física Aplicada, Universidad Politécnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain

158 Instituto de Física de Cantabria, Edificio Juan Jordá, Avenida de los Castros, 39005 Santander, Spain

159 Department of Computer Science, Aalto University, PO Box 15400, Espoo, FI-00 76, Finland

160 School of Physics and Astronomy, University of Nottingham, University Park, Nottingham NG7 2RD, UK

161 Ruhr University Bochum, Faculty of Physics and Astronomy, Astronomical Institute (AIRUB), German Centre for Cosmological Lensing (GCCL), 44780 Bochum, Germany

162 DARK, Niels Bohr Institute, University of Copenhagen, Jagtvej 155, 2200 Copenhagen, Denmark

163 Department of Physics, Centre for Extragalactic Astronomy, Durham University, South Road, Durham, DH1 3LE, UK

164 Astrophysics Research Centre, University of KwaZulu-Natal, Westville Campus, Durban 4041, South Africa

165 School of Mathematics, Statistics & Computer Science, University of KwaZulu-Natal, Westville Campus, Durban 4041, South Africa

166 ARC Centre of Excellence for Dark Matter Particle Physics, Melbourne, Australia

167 Centre for Astrophysics & Supercomputing, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia

168 Van Swinderen Institute, University of Groningen, Nijenborgh 3, 9747 AG Groningen, The Netherlands

169 Department of Physics and Astronomy, University of the Western Cape, Bellville, Cape Town, 7535, South Africa

170 Ohio University, Physics & Astronomy Department, 1 Ohio University, Athens, OH 45701, USA

171 Brazilian Center for Research in Physics (CBPF), Dr. Xavier Sigaud st. 150, zip 22290-180, Rio de Janeiro, RJ, Brazil

172 Institute for Theoretical Particle Physics and Cosmology (TTK), RWTH Aachen University, 52056 Aachen, Germany

173 INAF-Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125, Firenze, Italy

174 Centro de Astrofísica da Universidade do Porto, Rua das Estrelas, 4150-762 Porto, Portugal

175 HE Space for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Cañada, 28692 Madrid, Spain

176 STAR Institute, University of Liège, Quartier Agora, Allée du six Août 19c, 4000 Liège, Belgium

177 Dipartimento di Fisica “G. Occhialini”, Università degli Studi di Milano Bicocca, Piazza della Scienza 3, 20126 Milano, Italy

178 Department of Astrophysics, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland

179 Space physics and astronomy research unit, University of Oulu, Pentti Kaiteran katu 1, FI-90014 Oulu, Finland

180 Theoretical astrophysics, Department of Physics and Astronomy, Uppsala University, Box 524, 751 20 Uppsala, Sweden

181 Mathematical Institute, University of Leiden, Einsteinweg 55, 2333 CA Leiden, The Netherlands

182 School of Physical Sciences, The Open University, Milton Keynes, MK7 6AA, UK

183 School of Physics & Astronomy, University of Southampton, Highfield Campus, Southampton SO17 1BJ, UK

184 ASTRON, the Netherlands Institute for Radio Astronomy, Postbus 2, 7990 AA, Dwingeloo, The Netherlands

185 Center for Advanced Interdisciplinary Research, Ss. Cyril and Methodius University in Skopje, Macedonia

186 Department of Physics & Astronomy, University of Sussex, Brighton BN1 9QH, UK

187 Department of Astrophysical Sciences, Peyton Hall, Princeton University, Princeton, NJ 08544, USA

188 Fakultät für Physik, Universität Bielefeld, Postfach 100131, 33501 Bielefeld, Germany

189 SRON Netherlands Institute for Space Research, Landleven 12, 9747 AD, Groningen, The Netherlands

190 Department of Physics and Astronomy, University of British Columbia, Vancouver, BC V6T 1Z1, Canada

MRD and CalCD-A requirements

ID	MRD requirement	DR1 status
Weak lensing
MRD-WL-001	VIS off-band throughput The VIS channel off-band throughput shall be $<10^{-3}$ on average in the range 300–525 nm with the transition between on- and off-band $<25$ nm.	Mean throughput $\approx2 \times10^{-3}$, dielectric coating on FoM3 was replaced with silver coating. Transition width $\approx10$ nm.
MRD-WL-002	VIS PSF FWHM The FWHM of the VIS channel PSF at 800 nm shall be $\leq0.18^{\prime\prime}$ over one exposure. This requirement does not include the effects of pixelisation.	Median FWHM $\approx0.140^{\prime\prime}$, figure 9 in Euclid Collaboration: McCracken et al. [31].
MRD-WL-003	VIS pixel scale The pixel scale of the VIS channel shall be $0.10^{\prime\prime} \pm0.01^{\prime\prime}$.	$0.0991$–$0.1007^{\prime\prime}$ across the FoV.
MRD-WL-004	VIS PSF ellipticity The ellipticity of the VIS channel PSF, measured from an image that has been exposed for the nominal integration time, shall be $<0.15$.	$\lesssim0.025$; Miller et al. (in prep.), and Whittam et al. (in prep.).
MRD-WL-005	VIS PSF profile The profile of the VIS channel PSF over one exposure at 800 nm shall be such that $(R_{\mathrm{PSF}}/R_{\mathrm{ref}})^2 < 4$. $R_{\mathrm{ref}}$ is the value for a Gaussian profile with an FWHM of $0.2^{\prime\prime}$. As a goal, a value of $(R_{\mathrm{PSF}}/R_{\mathrm{ref}})^2 < 3$ should be pursued.	Approximately 1.9 (Miller et al., in prep.).
MRD-WL-006	VIS PSF model (ellipticity) The spatio-temporal variation of the VIS channel PSF shall allow a model of the PSF as a function of spatial position and time to be reconstructed at any point with a residual model uncertainty of $\sigma(e_i) < 2 \times10^{-4}$ per exposure and per ellipticity component with respect to the system PSF. The model should be constructed from stellar observations of as many exposures as required.	See Sect. 2.5.2.
MRD-WL-007	VIS PSF model (profile) The spatio-temporal variation of the VIS PSF shall allow a model of the PSF as a function of spatial position and time to be reconstructed at any point with a residual model uncertainty of $\sigma(R_{\mathrm{res}}/\langle R\rangle) < 10^{-3}$ with respect to the system PSF. The model should be constructed from stellar observations of as many exposures as required.	See Sect. 2.5.2.
MRD-WL-009	VIS geometric distortion The root mean square (RMS) post-calibration geometric distortion in the VIS channel averaged within a circular aperture of radius $4^{\prime}$ shall be $<0.003\%$. As a goal, a value $<0.001\%$ should be pursued.	RMS distortion gradient over $4^{\prime}$ of $7 \times10^{-5}\%$, maximum $2 \times10^{-3}\%$ in any single exposure; edge ellipticity exceeds $3 \times10^{-5}$ below $22^{\prime\prime}$; section 4.2.1, figure 4 in Euclid Collaboration: McCracken et al. [31].
MRD-WL-010	NISP-P off-band throughput The NISP-P off-band throughput shall be $<10^{-3}$.	$<10^{-4}$, figure 13 in Euclid Collaboration: Schirmer et al. [42].
MRD-WL-011	NISP-P photometric bands The NISP-P channel shall have three or more wavelength bands.	$Y_{\mathrm{E}}$}, $J_{\mathrm{E}}$, $H_{\mathrm{E}}$ bands.
MRD-WL-012	NISP-P PSF profile and compactness (EE50) The NISP-P radii encircling 50% of energy shall be $<0.4^{\prime\prime}$ for all wavelengths below 1486 nm.	$0.28^{\prime\prime}$ or better in all bands, table 3 in Euclid Collaboration: da Silva et al. [20].
MRD-WL-013	NISP-P PSF profile and compactness (EE80) The NISP-P radii encircling 80% of energy shall be $<0.7^{\prime\prime}$ for all wavelengths below 1486 nm.	$0.56^{\prime\prime}$ or better in all bands, table 3 in Euclid Collaboration: da Silva et al. [20].
MRD-WL-014	NISP-P pixel scale The NISP-P pixel scale shall be $0.30^{\prime\prime} \pm0.03^{\prime\prime}$.	$0.296$–$0.299^{\prime\prime}$ across the FoV.
MRD-WL-015	NISP-P relative photometric error The NISP-P post-calibration relative photometric error shall be $<1.5\%$.	$0.9\%$ from self-calibration visits (section 4.3 and figure 8 in Euclid Collaboration: Polenta et al. [36]), and $0.3\%$ from propagating the absolute calibration level through the DR1 period.

ID	MRD requirement	DR1 status
MRD-WL-016	Shape-measurement multiplicative bias The multiplicative bias in shape measurement shall be known to an accuracy $\sigma(\mu) \le2 \times10^{-3}$ under all observing conditions and for all types of objects used for weak lensing. As a goal, the multiplicative bias should be $|\mu| < 10^{-3}$ for monochromatic galaxies.	See Sect. 2.5.2.
MRD-WL-017	Shape-measurement additive bias The additive bias $c$ in shape measurement shall be known to an accuracy $\sigma(c) \le5 \times10^{-4}$ under all observing conditions and for all types of objects used for weak lensing.	See Sect. 2.5.2.
MRD-WL-018	VIS PSF model (ellipticity, non-convolutive contribution) The non-convolutive contribution to the residual model uncertainty of the ellipticity shall be $\sigma(e_{\mathrm{NC}}) < 1.1 \times10^{-4}$ per exposure per ellipticity component with respect to the system PSF.	See Sect. 2.5.2.
MRD-WL-019	VIS off-band throughput The VIS channel off-band throughput shall be $< 10^{-3}$ on average in the range 935–1600 nm with the transition between on- and off-band $< 35$ nm.	Mean throughput $\lesssim8 \times10^{-4}$, width $\approx23$ nm.
MRD-WL-020	VIS relative photometric error The VIS channel post-calibration relative photometric error shall be $< 1.0\%$.	0.76 mmag (internal) to 1.4 mmag (external, with respect to Gaia) at quadrant level (see section 4.3 in Euclid Collaboration: McCracken et al. [31]).
MRD-WL-021	VIS absolute photometric error As a goal, the VIS channel post-calibration absolute photometric error should be $< 5.0\%$.	3.8 mmag mean offset w.r.t. Gaia DR3; section 4.3 and table 1 in Euclid Collaboration: McCracken et al. [31].
MRD-WL-022	NISP-P absolute photometric error As a goal, the NISP-P channel post-calibration absolute photometric error should be $< 5.0\%$.	Tied to HST standard to 0.5% or better, figure 7 in Euclid Collaboration: Polenta et al. [36]; limited by the standards’ intrinsic variability at the 1% level.
MRD-WL-023	VIS passband transition shape The VIS channel transition shape between on-band and off-band has to be known with an accuracy better than 1%.	Known at 1 nm resolution with high S/N. FoV-dependence due to angle of incidence needs to be computed.
	Galaxy clustering
MRD-GC-001	NISP-S line-flux limit at 1600 nm The average effective H$\alpha$-line flux limit from a 0′′5 diameter source in the NISP spectroscopic channel shall be $\le2 \times10^{-16}$ erg cm$^{-2}$ s$^{-1}$ at 1600 nm. The flux limit is defined as the flux for which the S/N is $> 3.5$.	$\approx2 \times10^{-16}$ erg cm$^{-2}$ s$^{-1}$, section 5.2 in Euclid Collaboration: Jahnke et al. [25].
MRD-GC-002	NISP-S redshift completeness The completeness of the redshift measurements in the NISP spectroscopic channel shall be $> 45\%$. The completeness is defined as the number of galaxies for which a redshift is measured at 50% confidence level, divided by the total number of galaxies at the flux limit specified by MRD-GC-001. As a goal, a value $> 65\%$ should be pursued.	See Sect. 2.5.1.
MRD-GC-003	NISP-S average flux limit The average flux limit for observations in the NISP spectroscopic channel at all wavelengths shall be $< 1.2$ times the flux limit at 1600 nm specified in MRD-GC-001.
MRD-GC-004	NISP-S red-grism spectral resolution The NISP spectroscopic channel spectral resolution considering a reference 0′′5 diameter source shall be $\mathcal{R} = \lambda/\Delta\lambda> 380$ over the 1250–1850 nm spectral range. Here, $\Delta\lambda$ is defined as the minimum wavelength separation at which two spectral lines produced by a 0′′5 object and with the same equivalent width can still be separated.	$\mathcal{R} > 480$, table 2 in Euclid Collaboration: Jahnke et al. [25].
MRD-GC-005	NISP-S position error of spectral features After calibration, the maximum error in the measured position of a spectral feature in the NISP-S red channel (1250–1850 nm) shall be $< 38\%$ of one resolution element.	0.35 pixel, or 18% total error (0th and 1st order), averaged over all grisms (table 1 in Euclid Collaboration: Fumana et al. [21]).
MRD-GC-006	Spectroscopic purity subsample A purity $> 99\%$ shall be achieved for a near-infrared spectroscopic subsample containing more than 120 000 galaxies. The same selection criteria as for the spectroscopic galaxy sample chosen from the EWS is to be applied.	See Sect. 2.5.1.
ID	MRD requirement	DR1 status
MRD-GC-007	Spectroscopic redshift purity The purity of the redshift measurements from NISP spectra shall be larger than 80%. Purity is defined as the fraction of galaxies satisfying MRD-GC-001 and MRD-GC-002 for which the best redshift estimate has a standard deviation around their true value of $\sigma(z) < 0.001\ (1+z)$.	See Sect. 2.5.1 and Euclid Collaboration: Le Brun et al. (2026).
MRD-GC-008	NISP-S relative photometric error Within circles of diameter $0.8^\circ$ distributed over the whole survey, fluctuations in the zero point of the flux limit in the NISP-S channel shall be smaller than 0.7% (RMS).	$\lesssim0.7\%$ across DR1, section 5.5 in Euclid Collaboration: Fumana et al. [21].
MRD-GC-009	NISP-S absolute photometric error As a goal, the NISP spectroscopic channel post-calibration absolute photometric error should be $< 5.0\%$ in the red band (1250–1850 nm).	Red grism: $\approx1\%$. Blue grism: up to 2% at the short-wavelength end (figure 5 in Euclid Collaboration: Fumana et al. [21]) only during PV. Detector-contamination effects are still uncorrected, and the goal is missed for the blue-grism (Sect. 7.8).
Wide survey
MRD-WS-001	EWS VIS depth Observations in the VIS channel in the EWS shall achieve a depth of AB magnitude 24.5 for $10\sigma$ extended objects. Extended objects are defined as those with a S/N greater than 10 and a size larger than 1.25 times the PSF size.	Turn-around point at $\approx25.5$, figure 22 in Euclid Collaboration: McCracken et al. [32].
MRD-WS-002	EWS NISP-P depth Observations in each band of the NISP-P channel in the EWS shall achieve a depth of AB magnitude 24 (in $Y_E$, $J_E$, and $H_E$) for $5\sigma$ point sources.	25.0 in $2\times$ FWHM aperture.
MRD-WS-003	VIS number of dithers Observations in the VIS channel in the EWS shall be performed with at least four dithers.	Four dithers, section 4.1 in Euclid Collaboration: Mellier et al. [33].
MRD-WS-004	NISP number of dithers Observations in the NISP-P and NISP-S channels in the EWS shall be performed with at least four dithers in each band.	Four dithers, section 4.1 in Euclid Collaboration: Mellier et al. [33].
MRD-WS-005	NISP-S dispersion directions NISP-S observations in the EWS shall be performed such that the spectra obtained have $\geq3$ different orientations on the sky.	Four angles, section 3.5.3 in Euclid Collaboration: Mellier et al. [33].
MRD-WS-006	NISP-P images must complement NISP-S exposures NISP-P imaging of the fields covered by the NISP-S channel in the EWS shall be acquired with depth fainter than 24 AB mag for $5\sigma$ point sources. The required depth can be achieved through combination of the different NISP-P bands bands.	25.0 in $2\times$ FWHM aperture per band.
MRD-WS-007	EWS fraction of area useful for cosmological analysis At least 85% of the EWS covered area shall be useful for the core Euclid science analysis.	Area loss estimated at 11.7% (88.3% useful).
Deep survey
MRD-DS-001	EDS imaging depth The stack of images obtained in the EDS shall achieve a depth at least 2 mag deeper than the EWS depth in both the VIS and NISP-P imaging channels.	The EDS does not yet cover all visits. Current depth-scaling measurements with the DR1 pipeline fall within 0.1 mag of the requirement.
MRD-DS-002	EDS spectroscopic depth The stacked line flux obtained in the EDS in the NISP-S channel shall achieve a depth of at least $5\times10^{-17}$ erg s$^{-1}$ cm$^{-2}$ ($3.5\sigma$) in the wavelength range common to the EWS.	The EDS does not yet cover all visits.
MRD-DS-003	EDS blue-grism coverage The EDS shall include observations with the blue band (935–1300 nm) of the NISP-S channel over the same area as the red band.	Implemented, with a blue:red exposure-time ratio of 5:3, section 4.2 in Euclid Collaboration: Mellier et al. [33].
MRD-DS-004	EDS blue-grism depth As a goal, the depth in the NISP-S blue channel (935–1300 nm) for the stacked line flux should be better than $6\times10^{-17}$ erg s$^{-1}$ cm$^{-2}$ ($3.5\sigma$) at 1100 nm.	The EDS does not yet cover all visits. The blue:red exposure-time ratio of 5:3 was chosen to achieve comparable depth across both spectroscopic passbands.
ID	MRD requirement	DR1 status
MRD-DS-005	NISP-S blue-grism spectral resolution The NISP-S blue channel spectral resolution considering a reference $0.\!''5$ diameter source shall be $R = \lambda/\Delta\lambda> 260$ in the 935–1300 nm spectral range. The resolution element ($\Delta\lambda$) is defined as the minimum wavelength separation at which two spectral lines produced by a $0.\!''5$ object and with the same equivalent width can still be separated.	$R > 400$, table 2 in Euclid Collaboration: Jahnke et al. [25].

ID	CalCD-A requirement	DR1 status
R-CAL-A-VS-1000	VIS PSF ellipticity post-calibration error Assuming a perfect extraction, the data should be calibrated such that the intrinsic contribution to the error in the determination of the ellipticity components of the PSF model is $\sigma(\epsilon) \le7.0 \times10^{-5}$.	See Sect. 2.5.2.
R-CAL-A-VS-2000	VIS PSF $R^2$ post-calibration error Assuming a perfect extraction, the data shall be calibrated such that the intrinsic contribution to the error in the determination of the relative $R^2$ of the PSF model is $\sigma(R^2)/R^2 \le2.5 \times10^{-4}$.	See Sect. 2.5.2.
R-CAL-A-VS-3000	VIS distortion correction The RMS of the post-calibration geometric distortion in VIS measured on a scale of $4'$ shall be $< 0.003\%$.	0.08 mas RMS, or $8 \times10^{-5}\%$, section 4.2 in Euclid Collaboration: McCracken et al. [31].
R-CAL-A-VS-4000	VIS relative photometric error After all calibrations, the relative photometric error of the VIS instrument shall be $< 1.0\%$.	0.76 mmag (internal), 1.4 mmag (external) at quadrant level; section 4.3 and table 1 Euclid Collaboration: McCracken et al. [31].
R-CAL-A-VS-5000	VIS absolute photometric error After all calibrations, the absolute photometric error of the VIS instrument shall be $< 5.0\%$.	3.8 mmag mean offset w.r.t. Gaia DR3; section 4.3 and table 1 in Euclid Collaboration: McCracken et al. [31].
R-CAL-A-VS-6000	VIS colour gradient The following fields shall be observed with VIS to an S/N 4 times better than the EWS: AEGIS, SXDS, COSMOS, CDFS, GOODS-NORTH.	AEGIS and CDFS completed, observations ongoing.
R-CAL-A-VS-7000	VIS noise bias fields 40 deg$^2$ shall be observed with VIS to an S/N $> 6$ times that of the EWS. The fields must also be observed by EXT, at least as deep as the EWS.	Observations ongoing until completed in DR3.
R-CAL-A-NP-1000	NISP-P relative photometric error After all calibrations, the relative photometric error of NISP-P shall be $< 1.5\%$.	0.9% from self-calibration visits (section 4.3 and figure 8 in Euclid Collaboration: Polenta et al. 2026a), and 0.3% from propagating the absolute calibration level through the DR1 period.
R-CAL-A-NP-3000	NISP-P absolute photometric error After all calibrations, the absolute photometric error of NISP-P shall be $< 5.0\%$.	Tied to HST standard to 0.5% or better (figure 7 in Euclid Collaboration: Polenta et al. 2026a); limited by the standards’ intrinsic variability at the 1% level.
R-CAL-A-NP-4000	NISP-P colour-gradient fields The following fields shall be observed in all NISP-P bands to an S/N of four times better than the EWS: AEGIS, SXDS, COSMOS, CDFS, GOODS-NORTH.	AEGIS and CDFS completed, observations ongoing.
R-CAL-A-NP-5000	NISP-P photo-z calibration A sample of 20 000 galaxies representative of the EWS, which have adequate external visible and near-infrared spectroscopic observations by launch, shall be observed in all NISP-P bands and VIS to an S/N 5 times better than the EWS. At least four different lines-of-sight are required to account for cosmic variance, and the fields must be observed by EXT.	COSMOS, VVDS, CDFS, and SXDS fields; observations ongoing.
R-CAL-A-NS-1000	NISP-S wavelength error The error in the determination of the wavelength of a detected emission line on one spectral element – assumed to be four pixels – in the NISP-S images shall be $< 0.8$ pixel.	$\lesssim0.5$ pixel combined uncertainty of the full calibration chain (section 5 in Euclid Collaboration: Fumana et al. [21]), with residual systematics (see Sect. 7.7).
ID	CalCD-A requirement	DR1 status
R-CAL-A-NS-3000	NISP-S absolute spectrophotometric error After all calibrations, the absolute spectrophotometric error of NISP-S shall be $< 5.0\%$.	Met with caveats (see MRD-GC-009 in Table A.1).
R-CAL-A-NS-4000	Purity spectroscopic sub-sample A NISP spectroscopic subsample, with the same selection criteria as that of the spectroscopic galaxy sample chosen from the EWS, containing more than 120 000 galaxies shall be provided with purity $> 99\%$.	Observations ongoing until completed in DR3.
R-CAL-A-NS-5000	NISP-S spatial relative spectrophotometric error Within the EWS, a spatially varying solution for the relative flux calibration shall be provided. The solution shall be such that the relative flux at the flux limit of the sample, averaged over circles of diameter $0.^\circ8$, has a distribution with width less than $0.7\%$ (RMS).	$\lesssim0.7\%$ across DR1 for the red grism (section 5.5 in Euclid Collaboration: Fumana et al. [21]). Blue grism still affected by contamination (see Sect. 7.8).

Tables and figures

Vector (PDF)

Transformation	$x$	RMS	Required preselection to reject hot and cool stars
Euclid→Roman
$F062 = I_{\mathrm{E}} + 0.0451 + 0.3712\,x + 0.0612\,x^{2}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0090	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.20, 0.50]$
$F087 = I_{\mathrm{E}} - 0.0471 - 0.5742\,x + 0.0462\,x^{2}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0035	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.20, 0.50]$
$\mathbf{F106 = Y_{\mathrm{E}} - 0.0062 + 0.0313\,x - 0.0046\,x^{2}}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0006	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.25, 0.60]$
$F106 = Y_{\mathrm{E}} + 0.0001 + 0.0588\,x - 0.0226\,x^{2}$	$x = Y_{\mathrm{E}} - J_{\mathrm{E}}$	0.0007	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.33, 0.12]$
$\mathbf{F129 = J_{\mathrm{E}} - 0.0021 + 0.2839\,x + 0.0360\,x^{2}}$	$x = Y_{\mathrm{E}} - J_{\mathrm{E}}$	0.0011	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.30, 0.12]$
$F129 = J_{\mathrm{E}} + 0.0333 + 0.2528\,x - 0.0450\,x^{2}$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0013	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, -0.05]$
$F158 = H_{\mathrm{E}} - 0.0586 + 0.3104\,x$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0013	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, -0.02]$
$F184 = H_{\mathrm{E}} + 0.0565 - 0.0842\,x - 0.0518\,x^{2}$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0021	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, -0.03]$
Roman→Euclid
$I_{\mathrm{E}} = F062 - 0.0042 - 0.3940\,x - 0.0697\,x^{2}$	$x = F062 - F087$	0.0049	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.23, 0.60]$
$\mathbf{I_{\mathrm{E}} = F087 - 0.0042 + 0.6060\,x - 0.0697\,x^{2}}$	$x = F062 - F087$	0.0049	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.23, 0.60]$
$\mathbf{Y_{\mathrm{E}} = F106 + 0.0028 - 0.0751\,x + 0.0302\,x^{2}}$	$x = F087 - F106$	0.0005	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.08, 0.23]$
$\mathbf{Y_{\mathrm{E}} = F106 + 0.0001 - 0.0759\,x + 0.0288\,x^{2}}$	$x = F106 - F129$	0.0006	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.30, 0.08]$
$\mathbf{J_{\mathrm{E}} = F129 + 0.0030 - 0.3660\,x - 0.0942\,x^{2}}$	$x = F106 - F129$	0.0017	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.25, 0.12]$
$\mathbf{J_{\mathrm{E}} = F129 - 0.0082 - 0.2718\,x + 0.0494\,x^{2}}$	$x = F129 - F158$	0.0010	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.28, 0.07]$
$H_{\mathrm{E}} = F158 + 0.0889 - 0.3285\,x - 0.0346\,x^{2}$	$x = F129 - F158$	0.0022	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.23, 0.05]$
$\mathbf{H_{\mathrm{E}} = F158 - 0.0392 - 0.7789\,x + 0.3488\,x^{2}}$	$x = F158 - F184$	0.0014	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.25, -0.12]$
$\mathbf{H_{\mathrm{E}} = F184 - 0.0356 + 0.2246\,x + 0.2365\,x^{2}}$	$x = F158 - F184$	0.0013	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.23, -0.10]$
Euclid→JWST
$\mathbf{F070W = I_{\mathrm{E}} - 0.0274 - 0.0634\,x + 0.0455\,x^{2}}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0047	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, 0.50]$
$F090W = Y_{\mathrm{E}} - 0.0417 + 0.3531\,x$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0038	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.12, 0.45]$
$F090W = Y_{\mathrm{E}} + 0.0309 + 0.6553\,x$	$x = Y_{\mathrm{E}} - J_{\mathrm{E}}$	0.0044	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.15, 0.12]$
$\mathbf{F115W = J_{\mathrm{E}} - 0.0013 + 0.7393\,x - 0.0041\,x^{2}}$	$x = Y_{\mathrm{E}} - J_{\mathrm{E}}$	0.0011	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.33, 0.13]$
$F115W = J_{\mathrm{E}} + 0.0915 + 0.6571\,x - 0.1962\,x^{2}$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0026	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, -0.02]$
$\mathbf{F150W = H_{\mathrm{E}} - 0.0575 + 0.5115\,x + 0.0464\,x^{2}}$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0016	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.37, 0.00]$
$\mathbf{F200W = H_{\mathrm{E}} + 0.1667 - 0.1063\,x}$	$x = J_{\mathrm{E}} - H_{\mathrm{E}}$	0.0028	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, 0.00]$
JWST→Euclid
$I_{\mathrm{E}} = F070W + 0.0272 + 0.1018\,x - 0.1219\,x^{2}$	$x = F070W - F090W$	0.0047	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.18, 0.35]$
$Y_{\mathrm{E}} = F090W + 0.0492 - 0.5972\,x + 0.0506\,x^{2}$	$x = F070W - F090W$	0.0061	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.05, 0.30]$
$\mathbf{Y_{\mathrm{E}} = F090W - 0.0078 - 0.7143\,x - 0.0186\,x^{2}}$	$x = F090W - F115W$	0.0007	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.15, 0.16]$
$J_{\mathrm{E}} = F115W + 0.0278 - 0.0444\,x$	$x = F090W - F115W$	0.0013	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.13, 0.15]$
$\mathbf{J_{\mathrm{E}} = F115W - 0.0053 - 0.5769\,x + 0.0557\,x^{2}}$	$x = F115W - F150W$	0.0011	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.40, 0.15]$
$H_{\mathrm{E}} = F150W + 0.1226 - 0.4434\,x - 0.1389\,x^{2}$	$x = F115W - F150W$	0.0032	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.33, 0.10]$
$\mathbf{H_{\mathrm{E}} = F150W - 0.1370 - 0.8255\,x + 0.1066\,x^{2}}$	$x = F150W - F200W$	0.0023	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.43, -0.23]$
$H_{\mathrm{E}} = F200W - 0.1270 + 0.1764\,x$	$x = F150W - F200W$	0.0021	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.44, -0.23]$
Euclid→CSST
$r_{\mathrm{CSST}} = I_{\mathrm{E}} + 0.0230 + 0.3242\,x + 0.0370\,x^{2}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0042	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.35, 0.60]$
$i_{\mathrm{CSST}} = I_{\mathrm{E}} - 0.0461 - 0.2803\,x$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0045	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.22, 0.50]$
$W_{\mathrm{CSST}} = I_{\mathrm{E}} + 0.2775 + 1.3290\,x$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0521	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.02, 0.50]$
$W_{\mathrm{I\,CSST}} = I_{\mathrm{E}} - 0.0317 - 0.2324\,x + 0.0320\,x^{2}$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0040	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.20, 0.50]$
$z_{\mathrm{CSST}} = Y_{\mathrm{E}} - 0.0394 + 0.3470\,x$	$x = I_{\mathrm{E}} - Y_{\mathrm{E}}$	0.0050	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.08, 0.50]$
CSST→Euclid
$\mathbf{I_{\mathrm{E}} = r_{\mathrm{CSST}} + 0.0189 - 0.5315\,x - 0.1343\,x^{2}}$	$x = r_{\mathrm{CSST}} - i_{\mathrm{CSST}}$	0.0016	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.25, 0.50]$
$\mathbf{I_{\mathrm{E}} = W_{\mathrm{I\,CSST}} + 0.0118 + 0.4122\,x - 0.1454\,x^{2}}$	$x = r_{\mathrm{CSST}} - W_{\mathrm{I\,CSST}}$	0.0018	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.30, 0.45]$
$I_{\mathrm{E}} = z_{\mathrm{CSST}} + 0.0546 + 1.7254\,x$	$x = i_{\mathrm{CSST}} - z_{\mathrm{CSST}}$	0.0060	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.11, 0.19]$
$I_{\mathrm{E}} = z_{\mathrm{CSST}} + 0.0269 + 1.5353\,x$	$x = W_{\mathrm{I\,CSST}} - z_{\mathrm{CSST}}$	0.0052	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.13, 0.25]$
$I_{\mathrm{E}} = W_{\mathrm{V\,CSST}} + 0.0043 - 0.8509\,x - 0.0405\,x^{2}$	$x = W_{\mathrm{V\,CSST}} - W_{\mathrm{I\,CSST}}$	0.0054	$I_{\mathrm{E}} - Y_{\mathrm{E}} < 0.69$; $x \in[-0.40, 1.10]$

Survey and system characterisations

ID	Description
SURV-001	Depth Determines whether Euclid achieves the required survey depths and number counts, serving as a go/no-go criterion at the end of PV. One complete pass of the EDF-N and – if visible – the EDF-S provides sufficient data.
SURV-002	Relative photometric accuracy Assesses whether Euclid can achieve the required photometric precision for multiply detected sources across different images, serving as a go/no-go criterion at the end of PV. Estimated by analysing the consistency of flux measurements for sources detected multiple times in different exposures of the EDF-N during PV.
SURV-003	Sensitivity Uses the single pass of the EDF-N area in PV to verify the pre-launch estimates of the VIS and NISP detection limits for the chosen exposure times. This is another critical go/no-go criterion to commence the survey.
SURV-004	Zodiacal model Validates the zodiacal background model that is based on previous space missions, a go/no-go criterion to commence the survey. Pre-launch uncertainties in absolute flux calibration, intrinsic zodiacal variability, straylight modelling, and passband transformations to Euclid passbands limit the zodiacal uncertainties to the 5–10% range. The zodiacal background would need to be 30% higher than expected before necessitating changes to the exposure times in the ROS.
TEL-001	Telescope thermal timescale Determines the typical time-scale required for the telescope to reach sufficiently stable thermo-optical equilibrium for weak lensing measurements after considerable changes in attitude relative to the Sun. As a metric, we use the focus term derived from measurements of the diffraction spikes, together with thermistor readings from the PLM.
TEL-002	PLM contamination Aims to detect water ice on optical surfaces, using temporal changes in stellar photometry in the self-calibration field as well as comparisons of EWS images with Gaia photometry. This is a by-product of nominal calibration processing.
TEL-003	External straylight levels Must be obtained early in PV to map straylight contamination levels at various spacecraft attitudes, following the commissioning discovery that Sunlight enters the instrument cavity – among others – through a scattering chain originating from an illuminated thruster nozzle. Straylight levels are largely mitigated by rotating the spacecraft to move the nozzle into shadow. This characterisation determines the ‘dark’ AA and SAA subspace usable for survey operations.
TARGET-001	Colour gradient calibration fields Observes five CANDELS fields (AEGIS, GOODS-NORTH, COSMOS, UDS, GOODS-CDFS) to calibrate biases in VIS shear measurements caused by spatially variable SEDs within galaxies in conjunction with the chromatic VIS PSF. The AEGIS field must be completed in the first year due to its extensive HST observations; at least one field must be observed each following year.
TARGET-002	Photo-z calibration fields Observes four deep fields (COSMOS-wide, VVDS-Deep, CDFS, SXDS) with S/N 5 times better than the EWS. This yields NISP photometry in the $Y_{\mathrm{E}}$, $J_{\mathrm{E}}$, and $H_{\mathrm{E}}$ bands for approximately 20 000 galaxies with existing ground-based spectroscopy, enabling calibration of the photo-$z$ methods. The time-critical observation strategy requires the COSMOS and CDFS fields to be observed early-on, since these data also fulfill colour-gradient calibration requirements.
TARGET-003	Noise bias calibration fields Observes approximately 200 000 galaxies with S/N 6 times that of the wide survey across at least 40 deg$^{2}$. These data enable the characterisation and calibration of noise-induced biases in shape measurements, which must be corrected for in the shallower EWS data. The data are automatically provided by the EDS.
TARGET-004	Completeness and purity calibration Generates a spectroscopic subsample of at least 120 000 galaxies with 99% purity selected using identical criteria as in the EWS. Requires a survey area of at least 40 deg$^{2}$ based on expected galaxy number. The observations require multiple orientations (ten visits with minimum 5° separation angles plus four angles from the ROS) to enable optimal decontamination of overlapping slitless spectra. The full area is observed using all four red grisms and with the blue grism. A 3×3 field subset must be observed with the blue grism already during PV. The data enable the correction of selection biases in the EWS and ensure that the spectroscopic sample meets the requirements for galaxy clustering and redshift distribution measurements.

In-flight calibration observations (CALBLOCKs)

ID	Description
F-001	Self-calibration Objective: Provides baseline monitoring of Euclid’s spectrophotometric response to space weathering and outgassing through multiple dithered exposures of the self-calibration field near the North Ecliptic Pole. Observations: Executes the ROS_D1 at 60 pointings with specific grisms; four visits are required to observe all grisms at all dithers, including blue-grism observations with AA offsets of at least $\pm1.4$ for spectral decontamination. Cadence: Every 25–40 days. Duration: Approximately 23.5 hours per visit (94 hours total for four visits). The self-calibration observations support a large number of calibration products: NISP BFE (Table E.1), NISP-P Absolute Photometric Standards (Table F.1), NISP-P Focal Plane Astrometric Solution (Table F.1), NISP-P Large-Scale Flat (Table F.1), NISP-P PSF Model (Table F.1), NISP-P Dichroic Ghost Model (Table F.1), NISP-S Dichroic Ghost Model (Table G.1), NISP-S 0th Order Centroid SED Dependence Calibration (Table G.1), NISP-S Absolute Photometric Standards (Table G.1), NISP-S Focal Plane Distortion (Table G.1), NISP-S Large-Scale Flat (Table G.1), NISP-S Spectral Trace Calibration (Table G.1), NISP-S Wavelength Dispersion Check (Table G.1), VIS Absolute Photometric Standards (Table H.1), VIS BFE (Table H.1), VIS Cross-talk (Table H.1), VIS Focal Plane Astrometric Solution (Table H.1), VIS Large-Scale Flat (Table H.1), VIS Dichroic Ghost Model (Table H.1), and VIS Shuttermaps (Table H.1).
F-002	Noise-bias fields (Euclid Deep Fields) Objective: Obtains a sample of 200 000 galaxies with $S/N$ six times that of the EWS from a 40 deg$^{2}$ area (EDF-North, EDF-South, and EDF-Fornax) to quantify noise bias in weak lensing measurements. Observations: Uses the ROS with a blue:red grism exposure-time ratio of 5:3 to account for different zodiacal background and photon flux; regular grid dithering with spiral offsets between visits. Cadence: Continuous replanning throughout the mission to account for space-weather losses etc. Duration: Variable, depends on total area coverage required, and is partially fulfilled by CALBLOCK-F-003.
F-003	Completeness and purity calibration (CPC) Objective: Generates a spectroscopic subsample of at least 120 000 galaxies with 99% purity selected with the same criteria as the EWS, requiring observations with sufficiently many orientations for optimal slitless spectra decontamination. Observations: Obtained from EDF-N and EDF-S (43 deg$^{2}$ total); requires ten visits per field with minimum $5^\circ$ angle separation and the K-sequence angles (see [41]); The blue:red grism ratio is the same as for CALBLOCK-F-002, that is 5:3. Cadence: Continuous replanning throughout the mission. Duration: Variable based on completion requirements; partially fulfilled by CALBLOCK-F-002.
F-004	Colour-gradient calibration fields Objective: VIS and NISP observe five fields (Table 2) with extensive HST ancillary data to four times the mean EWS $S/N$ to calibrate for colour gradient effects in galaxies that bias shear measurements. Observations: Uses the ROS with a blue:red grism exposure-time ratio approximating 5:3; spiral dithering pattern; three of the fields are already covered by the photo-$z$ calibration observations (CALBLOCK-F-005). Cadence: The AEGIS field must be completed in the first year, followed by at least one additional field per following year. Duration: 63.8 hours (excluding 125.4 hours already covered by the photo-$z$ fields).
F-005	Photo-$z$ calibration fields Objective: Obtains NISP photometry in the $Y_{\mathrm{E}}$, $J_{\mathrm{E}}$, and $H_{\mathrm{E}}$ bands for 20 000 galaxies with existing ground-based spectroscopy at five times the EWS $S/N$ to calibrate Euclid photometric-redshift methods. Observations: Four fields (COSMOS-wide, VVDS-Deep, CDFS, SXDS) are observed with the ROS and the standard 5:3 blue:red grism exposure-time ratio; a spiral dither is employed; the COSMOS field spanning a $2 \times2$ mosaic has priority with a staged depth-completion strategy. Cadence: COSMOS started in first year with specific completion milestones; other fields distributed over following years. Duration: 745.3 hours total.
F-006	Wide Survey Objective: The EWS provides inline calibration products as part of the ROS including VIS flats, biases, darks, CTI measurements, trap pumping, and short science exposures that validate and update numerous calibration products. Observations: Executes the EWS with cycling VIS ROS-inline calibrations, the cycle repeats after 20 ROS; NISP acquires slew darks (Sect. 4.1) during field slews. Cadence: Continuous survey operations. Duration: Not applicable, as these are the intrinsic survey operations. For reference, one ROS lasts about 4300 seconds.
F-007	VIS nonlinearity, PTC, BFE and NISP LED flats Objective: VIS: Monitors nonlinearity, photon transfer curve (PTC), and BFE using on-sky stellar fields with special dither patterns and exposure times. NISP: Acquires light-emitting diode (LED) flats at 30 kADU fluence for monthly small-scale flat calibration and persistence monitoring. Observations: VIS executes 16 four-exposure nonlinearity sequences plus flat/short-science pairs at the self-calibration field; NISP takes 30 photo-MACC exposures per LED for all five LEDs, followed by 4.5 hours of darks for persistence decay. Cadence: Every 25–40 days, alternating between LED1 and LED4 variants (reference to CalF PDF excerpt). Duration: 10.2 hours.
F-008	VIS PSF calibration and NISP photo-nonlinearity Objective: VIS: Calibrates and validates the PSF model using stars in magnitude range $I_{\mathrm{E}} \in[17.5;21.0]$ at five times the EWS $S/N$. NISP: Takes parallel MACC(16,4) nonlinearity sequences in “engineering RAW” mode (see [25]) at various LED flux levels cycling through twelve configurations. Observations: VIS observes one of 24 PSF calibration fields 100 times with 289 s exposures using spiral dithering; NISP cycles through twelve LED flux variants (8, 20, 80, 200, 450, 550, 12, 40, 150, 350, 750, 950 e$^{-}$ s$^{-1}$) taking 10–30 ramps per detector pair followed by darks for persistence decay. Cadence: Approximately every 2–6 weeks, tied to survey thermal states to guarantee a stable PSF for VIS. Duration: 15 hours.
ID	Description
F-009	NISP spectro-nonlinearity Objective: Determines the nonlinearity for spectro-MACC(15,16,11) mode using LED exposures in engineering RAW mode; validates the model through observations with different exposure times. Observations: Validation: observes the self-calibration field in all filters with both photo-MACC(4,16,4) and photo-MACC(15,16,11). Measurement: two variants acquiring 15–30 spectro-MACC ramps per detector pair at five flux levels each (variant 1, having 8, 16, 30, 70, 150 e$^{-}$ s$^{-1}$; variant 2, having 12, 20, 50, 110, 175 e$^{-}$ s$^{-1}$) using LED C, followed by persistence decay darks. Cadence: Once per year for each variant, including validation. Duration: 340.1 hours (14.2 days) for both variants.
F-010	NISP wavelength dispersion and VIS BFE Objective: NISP: Observes a compact PN ([35]) with strong emission lines to obtain the field-dependent dispersion solution; transfers the solution to the self-calibration field for drift monitoring using stellar absorption features. VIS: Obtains flat-field pairs for BFE calibration. Observations: Establishes an astrometric reference with four ROS, then observes the PN at 80 positions (a quincunx pattern per detector) taking two spectro-MACC exposures per position, immediately followed by a slew to the self-calibration field; the grism wheel is not moved until all exposures of the PN and the self-calibration field are completed for a grism. VIS takes flat pairs at 10 kADU using LED1 and LED4 during the NISP spectro exposures. Cadence: Once after 11–16 months of science operations. Duration: 160.8 hours (6.7 days).
F-011	NISP reciprocity failure Objective: Validates reciprocity failure (or flux-dependent nonlinearity from charge trapping) for each detector by observing the self-calibration field with and without additional LED background illumination to test trap-passivation effects. Observations: Repeats the survey’s standard S-dither pattern ([41]) 18 times at the self-calibration field with small random offsets; odd repetitions without LED, even repetitions with LED D at increasing fluxes (2, 4, 8, . . . , 512 e$^{-}$ s$^{-1}$ pixel$^{-1}$) providing 220–56 000 e$^{-}$ pixel$^{-1}$ additional background in $H_{\mathrm{E}}$-band. Cadence: Yearly, beginning one year into the survey. Duration: 5.3 hours.
F-012	VIS flatband voltage shift Objective: Monitors the flatband voltage shift in charge-coupled device (CCD) metal-oxide semi-conductor (MOS) capacitors caused by ionising radiation. It measures the charge-injection turn-on/off point as a function of initial drain level (IDL) voltage, which increases with accumulated radiation dose. Observations: Uses a dedicated VIS commanding sequence, stepping the IDL voltage from 10.5 V to 12.5 V in 0.05–0.20 V increments (34 steps total), injecting into 100 lines and reading a windowed region after each voltage change; all data are taken with the shutter closed. Cadence: Every 2–6 weeks, jointly with CALBLOCK-F-008 and CALBLOCK-F-015. Duration: 0.2 hours.
F-013	Phase-diversity campaign (PDC) Objective: Determines the VIS wavefront error for PSF modelling as a function of various thermal states (AA and SAA combinations), observing stellar fields in defocused and focused states after thermal stabilisation. Observations: Requires a 5–7 day thermal stabilisation at each of ten different SAA/AA positions; three defocused AA values for low-polarisation fields plus one high-polarisation field; remainder varies SAA and AA in-focus. Executes eight ROS_D1 for photometric classification; molecular contamination is monitored at the self-calibration field via executions of CALBLOCK-F-014; the observations include defocused shutter maps (see also Sect. 7.4). Cadence: Once, immediately after PV-phase and before the survey starts; if necessary, scheduled again during future unallocated time. Duration: 60–105 days total depending on 5–10 day thermal-stabilisation periods.
F-014	Contamination scan Objective: Provides quick photometric monitoring of the self-calibration field at high cadence to track contamination evolution during the phase diversity campaign (CALBLOCK-F-013) when thermal stability is critical. Observations: Executes a single ROS at the centre of the self-calibration field. Cadence: Approximately weekly during PV (first two weeks) and PDC; must not disturb thermal stabilisation periods. Duration: 1.2 hours.
F-015	VIS Serial trap pumping Objective: Executes serial trap pumping (STP), which requires a readout-electronics soft reset that causes a short-term bias instability; provides all twelve STP variants for trap characterisation but must be followed immediately by PSF calibration (CALBLOCK-F-008) as subsequent VIS images will show the bias offset for several hours; the PSF calibration is insensitive to the transient bias instability. Observations: Acquires a full set of twelve different STP configurations (dwell times, shuffle parameters, etc), followed by three trash biases to clear this bias perturbation. Cadence: Scheduled every 2–6 weeks immediately prior to CALBLOCK-F-008, then followed by CALBLOCK-F-012. Duration: 1.9 hours plus 3 dithers.
F-016	VIS charge-injection timing Objective: Gathers statistics on charge-injection temporal consistency and adds information about trap capture/release times by varying the charge-injection speed. Observations: Executes a single block with six different charge-injection configurations. Cadence: Every 5–7 months. Duration: 2.5 hours.
F-017	Grism sanity Objective: This is a technical observation consisting of a single red-grism ROS. It has to be scheduled on rare occasions to resolve a look-ahead issue concerning red/blue grism that can occur in the survey. Observations: Executes a single red-grism ROS. Cadence: When needed. Duration: 1.2 hours.
F-018	NISP persistence Objective: Monthly observations to update persistence parameters due to considerable time variability in the persistence behaviour. Observations: Takes three trash darks for persistence decay from preceding observations, then repeats twice: a single LED E exposure with a fluence of 30 kADU (690 e$^{-}$ s$^{-1}$) followed by five MACC(4,16,4) darks and four MACC(15,16,11) darks. Cadence: Monthly, scheduled between CALBLOCK-F-001 and CALBLOCK-F-007 to avoid affecting either and allow persistence decay before the survey resumes. Duration: 1.8 hours.

NISP common calibration products

ID / Cadence	Description
NI-001<br>Continuous	NISP bad pixel map Continuously updated, aggregating information from baseline maps, dark current, flats, slew darks, and permanently disconnected pixels. Both intermittent and permanently bad pixels enter the map. random telegraph noise (RTN) pixels are currently not covered by processing, and pixels affected by persistence are masked while exceeding thresholds.
NI-002<br>Once at commissioning; once during PV	NISP baseline map Measures the bias value for each pixel after reset and is affected by kTC noise from thermodynamically created charges in (disconnected) capacitors. Exhibits a temperature dependence of approximately $-100\ \mathrm{ADU\ K^{-1}}$, with reference pixels set to 5.5 kADU and science pixels to 10 kADU. Updated during commissioning and PV to account for temperature changes. The baseline map ensures that the vast majority of pixels operate within the linear range of the analogue-digital chain; it is stable for each up-the-ramp acquisition and thus removed through ramp fitting. Potential baseline drifts can be detected in the slew darks in conjunction with five raw lines whose MACC groups are always fully downlinked.
NI-003<br>Monthly	NISP brighter-fatter effect BFE can be characterised from flat fields, and validated by comparing observed FWHM of stars at different magnitudes; BFE is partially masked by interpixel capacitance and charge diffusion. We presume the BFE is constant.
NI-004	NISP colour-gradient calibration fields Replaced by CHAR-TARGET-001.
NI-005	NISP crosstalk Channel crosstalk has an amplitude of about 0.001%, detectable only for super-saturated stars. This calibration identifies the threshold at which bright stars cause a detectable signal in the form of dark rings in neighbouring channels. Crosstalk is unproblematic in the EWS, but masking will be needed for DR2 in the Euclid Deep Fields (EDF).
NI-006<br>Once during PV	NISP dark current Determines the true inflight amplitude of the dark current, which could not be measured accurately on ground. The dark current is very low (approximately $0.02\ \mathrm{e^{-}\ s^{-1}}$), and requires hundreds of exposures in the two MACC readout modes. It was measured once during PV, after permitting for sufficient persistence decay prior to the measurements.
NI-007	NISP intrapixel sensitivity variation IPSV was characterised on ground using spot scan measurements with flight-like detectors. In flight, a mean intrapixel sensitivity variation (IPSV) profile can be determined by measuring flux variations of stable sources in the self-calibration field as a function sub-pixel position known from Gaia astrometry. This has not yet been performed.
NI-008	NISP inter-pixel capacitance verification IPC is local electronic crosstalk caused by capacitive coupling between adjacent pixels that spreads charge prior to readout. This broadens the PSF and smooths Poisson noise, requiring correction before deriving, for example, the gain. The inter-pixel capacitance (IPC) distribution is narrow and expected to change minimally with detector position and over time (see [see 62]). It is wavelength-independent and must be distinguished from charge diffusion; the latter has a similar appearance but depends on the wavelength-dependent absorption-length of photons.
NI-009	NISP inoculation Deactivated. Deep saturation of detectors has been shown to inactivate charge traps on time scales of weeks and months ([70]). This ‘inoculation’ effect has been demonstrated on ground for the NISP H2RGs. However, it is even less understood than charge persistence itself. There is a concern that massive flooding of all pixels could alter the detector’s diode junction through strong electrostatic fields, causing effects not seen in normal survey data and possibly voiding other calibration data. Deliberate inoculation was therefore never attempted with NISP.
NI-010<br>Monthly	NISP LED flats 30 flats per LED, with a flux of 90 $\mathrm{e^{-}\ s^{-1}}$ yielding 4 kADU with the standard MACC(4,16,4) readout. Higher fluences tested during PV phase resulted in poorly understood flats in some configurations, and have not been pursued further.
NI-011<br>Monthly: MACC(4,16,4); Yearly: MACC(15,16,11)	NISP nonlinearity verification Nonlinearity data for the MACC(4,16,4) mode are obtained at twelve flux levels ranging from 8 $\mathrm{e^{-}\ s^{-1}}$ to 950 $\mathrm{e^{-}\ s^{-1}}$. The fluxes are interleaved and distributed across two blocks that are observed monthly in an alternating sequence. To sample the nonlinearity as charge accumulates in the detector, all groups in the ramp must be downloaded. Due to bandwidth limitations, only two detectors can be calibrated simultaneously, meaning we need 8 times as many exposures as normally. This makes nonlinearity calibration one of the more time-consuming in-orbit calibration activities. The nonlinearity of the MACC(15,16,11) mode is characterised in an analogous manner. We use ten flux levels between 8 $\mathrm{e^{-}\ s^{-1}}$ and 175 $\mathrm{e^{-}\ s^{-1}}$, distributed over two blocks that must be executed once per year, but not necessarily back-to-back since the nonlinearity appears to be very stable. Validation is performed by taking underspersion measurements in the self-calibration field with both MACC(4,16,4) and MACC(15,16,11) modes. After nonlinearity calibration, non-saturated stars within the overlapping dynamic range of the two exposure modes should have consistent fluxes within the measurement uncertainties.
NI-012<br>Once during PV (initial); monthly thereafter	NISP persistence This calibration provides the time constants and amplitudes of charge traps. They were obtained for five different fluence levels in PV using the reddest LED E. In the early phase of the nominal survey, it became evident that the persistence parameters change slowly depending on their illumination history, and eventually return to their nominal state. We update the persistence parameters now on a monthly basis, using a single fluence of 30 kADU.
ID / Cadence	Description
NI-013<br>Yearly	NISP reciprocity failure Reciprocity failure describes flux-nonlinearity from charge-trapping effects. Basically, a brighter source fills traps more quickly and therefore enters a linear regime sooner. We calibrate this on a yearly basis by observing the self-calibration field without and with additional, different light levels from LED D.
NI-014<br>Per field slew	NISP slew darks A single slew dark is taken during field slews. It provides continuous monitoring of detector performance throughout the mission without dedicated observing time. Slew darks are contaminated by persistence from both the zodiacal background, cosmic ray hits, and astrophysical sources in the previous exposures. While median combination removes the persistence from compact sources, the spatially uniform zodiacal contribution remains.
NI-015	NISP snowballs Snowballs are highly energetic events that appear as radially symmetric or elliptic, extended features that may saturate the detectors. They likely originate from energetic cosmic rays, and possibly radioactive decay in or near the detectors. Snowballs may be surrounded by a number a numbers of hot pixels. Affected pixels need to be masked in consecutive exposures. This is not a calibration product as such, and was introduced to track this effect.

NISP-P calibration products

ID / Cadence	Description
NP-001 Once during PV; optional repeats in routine phase	Absolute photometric standards This calibration obtains absolute flux standard measurements across the NISP focal plane by observing spectrophotometric standard stars on a $4 \times4$ grid centred on each detector. The observations characterise the entire optical path and provide independent verification of global system response, including persistence and trap filling effects. Executed once during PV-phase with high spatial sampling using $Y_E$, $J_E$, and $H_E$ filters, with optional low-sampling repeats during routine phase if photometric discrepancies arise.
NP-002 Monthly	Focal plane astrometric solution This product determines relative detector positions, orientations, and optical distortion by comparing source positions against Gaia astrometric references in multiply dithered self-calibration observations. While optical distortion is expected to remain stable, relative detector positions may drift due to outgassing, mechanical relaxation, and thermal effects, requiring monthly monitoring through self-calibration observations with approximately 140 unsaturated Gaia stars per detector providing sufficient accuracy.
NP-003 Monthly	Large-Scale Flat Monthly reconstruction of matching broad-band large-scale flat fields for $Y_E$, $J_E$, and $H_E$ filters using stellar photometry from self-calibration observations. These flats remove residual large-scale non-uniformity in the LED flats and any large-scale illumination effects from the payload module, ensuring photometric consistency across the focal plane.
NP-004	Dichroic ghost model In-orbit validation of the ground-based dichroic ghost model (based on Zemax simulations) to confirm location and amplitude of ghosts requiring masking for sources brighter than 14.6 AB mag with amplitude ratio of $8 \times10^{-8}$. While size, shape, and wavelength dependence are expected to match simulations, any inaccuracies will be evident in masked pixel data allowing model updates, with optical ghosts from NISP internal optics considered negligible.
NP-005	Photo-z calibration fields Replaced by CHAR-TARGET-002, this involved observing five fields (COSMOS, VVDS-Deep, CDFS, SXDS) to obtain NISP-P photometry in $Y_E$, $J_E$, and $H_E$ bands for 20 000 galaxies with existing spectroscopy at 5 times the EWS S/N. VIS data at minimum two times the EWS depth aids target detection and deb­lending, while observations conducted using the ROS for practical and legacy purposes, achieving 5 times EWS depth for both instruments.
NP-006	PN reference photometric catalogue This product provides a reference photometric catalogue supporting NISP-S spectral observations of planetary nebulae, particularly for determining wavelength zeropoints in the spectral dispersion solution. Observations use the ROS at each dither position across five pixels per detector for each prism, with NISP-P data used to determine accurate stellar positions since the bright target will be saturated in NISP images.
NP-007 Monthly	PSF model Monthly calibration and validation of the NISP PSF model using stars in the 16–21 AB mag range, with approximately 12 000 stars per detector per filter after 60 dithers in the self-calibration field providing sufficient sampling. The PSF model must be robust against undersampling and intra-pixel sensitivity variations to achieve position accuracy better than 0.1 pixel for brighter NISP sources, though no formalised requirements currently exist.
NP-008 Once during PV	Straylight check Once during PV-phase, this calibration tests for glints (parasitic reflections from mechanical features like detector edges) and validates the normalised detector irradiance (NDI) model for unbaffled scattered light. The glint check places a medium-bright star (around 6 AB mag) adjacent to the focal plane using the ROS_D1, while NDI validation places very bright stars (0–2 AB mag) at logarithmically decreasing separations from the focal-plane edge to verify the bright-star avoidance radius.
NP-009 Monthly	Small-scale flat Monthly reconstruction of matching broad-band small-scale flat fields for $Y_E$, $J_E$, and $H_E$ filters from NISP LED flats using ground-based quantum efficiency maps. These flats correct pixel response non-uniformity (PRNU) and small-scale features in photometric data, complementing the large-scale flats derived from self-calibration observations.
NP-010	Zodiacal model Replaced by CHAR-SURV-004, this involved updating the zodiacal background model based on routine wide survey fields sampling the regularly observed ecliptic longitude and latitude range. While large deviations are not expected given well-characterised zodiacal backgrounds from other missions, validation ensures backgrounds aren’t substantially higher than predicted, which could impact survey depth and require operational adjustments.
NP-011	Colour-gradient calibration fields Replaced by CHAR-TARGET-001.

NISP-S calibration products

ID / Cadence	Description
NS-001	0th-order centroid SED dependence This calibration characterises how source SEDs affect the 0th-order centroid position, which appears as a double-peaked peanut-like shape due to prism dispersion and grating blaze effects. Since the main SED impact changes the ratio of the two peaks (and thus the inferred mean position), this dependence must be quantified using wide survey data as sources with different SEDs are mapped at different FPA positions, with no immediate PV calibration required.
NS-002<br><br>Once during PV	Absolute photometric standards Absolute flux standard measurements obtained once during PV by observing HST CALSPEC spectro-photometric standards as a function of detector position for each grism, with subsequent self-calibration field monitoring providing relative throughput. The star is stepped across five positions per detector using ROS_D1 observations. The grisms are scheduled to minimise spectra order contamination from other sources.
NS-003<br><br>$\ge$ every 6 months (self-calibration);<br>once during PV and once after 1 yr (PN)	Focal-plane distortion This calibration determines the distortion pattern across the NISP spectroscopic focal plane through three mappings: sky (Gaia references) to 0th-order using bright stars; 0th-order to reference wavelength in the 1st order; and reconstruction of the wavelength dispersion solution from PN observations. Monitoring occurs at least every six months using secondary wavelength calibrators (absorption line systems) in the self-calibration field, with full PN-based wavelength dispersion determined once during PV and once after one year.
NS-004<br><br>Monthly	Large-scale flat Monthly reconstruction of matching spectroscopic large-scale flat fields for all four grisms from self-calibration observations. These flats correct for large-scale illumination variations and vignetting effects across the focal plane, ensuring consistent flux calibration across the spectroscopic field of view.
NS-005<br><br>Monthly	Small-scale flat Monthly reconstruction of matching spectroscopic small-scale flat fields for all four grisms from NISP LED flats using ground-based characterisation data. These flats correct pixel-to-pixel response variations and small-scale sensitivity features in the spectroscopic data, complementing the large-scale corrections.
NS-006<br><br>Every 6 months	Spectral trace calibration This product determines the relative positions of 0th, 1st, and 2nd spectral orders as a function of detector position, monitored every 6 months using self-calibration observations. The trace calibration ensures accurate extraction of spectroscopic data by characterising how spectral orders are positioned and oriented across the FPA.
NS-007<br><br>Monthly	Wavelength dispersion check Monthly validation of the wavelength dispersion model using secondary wavelength calibrators (stellar absorption-line systems) in the self-calibration field, identified during PV-phase when primary dispersion was determined from PN observations. While full dispersion solutions may not be derivable from secondary standards due to insufficient spatial density, they can reveal considerable drifts indicating need for PN-based recalibration.
NS-008<br><br>Once during PV;<br>once after 1 yr	Wavelength dispersion with Planetary Nebulae This calibration observes compact PNe with strong emission lines at 80 positions across each detector to obtain wavelength dispersion solutions for all grisms, executed once during PV and once after one year. The dispersion solution is immediately transferred to the self-calibration field without intermediate grism-wheel movements to maintain thermal stability, establishing secondary standards for routine drift monitoring.
NS-009	Wavelength Transfer to self-calibration field The wavelength dispersion established using the PN observations is transferred to the self-calibration field by identifying suitable secondary standards (point-like emission-line galaxies and stellar absorption-line systems) for regular monitoring. Self-calibration observations must immediately follow or precede planetary nebula observations to avoid thermal state changes that could alter the wavelength dispersion before transfer.
NS-010	Completeness and Purity Calibration Replaced by CHAR-TARGET-004.
NS-011<br><br>Once during PV	Dichroic Ghost Model Once during PV phase, this validates the on-ground dichroic ghost model (based on Zemax simulations) by confirming ghost location and amplitude for sources requiring masking, with amplitude ratio requirement to be determined ($8 \times10^{-8}$ for NISP-P provides reference). While size, shape, and wavelength dependence are expected to match simulations, any inaccuracies will be evident in masked pixel data allowing model updates, with optical ghosts from NISP internal optics considered negligible.
NS-012<br><br>Once during PV	Grism Centers of Rotation This calibration determines the grism-wheel rotation centres by observing a stellar field five times per grism setup with independent random realisations of the jitter angle (typically $0.1^\circ$ RMS, occasionally up to $0.5^\circ$), executed once during PV. While the dispersion pattern does not purely rotate due to optics chromaticity, to first order in small jitter angles the pattern can be derived from reference calibration as simple rotation, requiring full grism-wheel rotation between exposures to ensure independent angle realisations.

VIS calibration products

ID / Cadence	Description
VS-001<br>Once during PV; optional low-sampling repeats in routine phase	Absolute photometric standards Obtains absolute flux calibrations across the VIS focal plane by observing spectro-photometric standards on a $4 \times4$ grid centred on each NISP detector (yielding approximately seven positions per VIS CCD due to different FPA layouts). It characterises the entire optical path and checks system response including persistence and trap filling. Executed once during PV with high spatial sampling using short-science exposures to avoid saturation, with optional low-sampling repeats during routine phase should photometric discrepancies arise.
VS-002	Passband variations Deactivated. Determines dichroic passband variations from stellar observations with known SEDs, though VIS passband variations are quantified on ground using optical raytracing and dichroic reflectance measurements.
VS-003<br>Once early in PV; monthly in routine ops	Bias The VIS detector bias level may vary over time and exhibits spatial variations at readout beginning (bias stabilisation), with the spatial pattern expected to be stable within each quadrant and accurately computed from series of individual bias exposures. One master bias is created early in PV-phase, with routine operations acquiring bias frames during the ROS for monthly master bias computation.
VS-004<br>Once during PV	Brighter-fatter effect This calibration characterises the BFE where increasing charges in pixel potential wells change physical pixel boundaries, causing photons arriving late in exposures to create photo-electrons in neighbouring pixels. This is detectable as increasing stellar FWHM with flux and decreasing pixel-to-pixel variance in high-fluence flats. Executed once during PV using self-calibration stellar field observations and flat-field pairs, this achromatic effect directly impacts PSFs models and shear measurements and may be subject to radiation damage.
VS-005<br>6× early in PV; every 1–2 weeks during PV; daily in routine ops	CTI charge-injection lines This calibration provides the CTI model for correcting charge transfer inefficiency due to radiation damage, based on charge-injection line (CIL) data and flat fields. The full sequence is executed six times early in PV with 42 CIL patterns plus once every 1–2 weeks during PV, while daily routine operations acquire eight CIL images through the ROS with 32 different patterns cycling completely over approximately four days.
VS-006	Colour-gradient calibration fields Replaced by CHAR-TARGET-001.
VS-007<br>Monthly	Cosmetic map Monthly update of the cosmetic map/bad pixel map based on various data sources acquired during routine operations and dedicated calibrations. This continuously maintained product tracks detector defects, hot pixels, dead pixels, and other cosmetic issues affecting data quality.
VS-008<br>Monthly	Crosstalk Monthly determination and updating of the VIS crosstalk matrix from self-calibration and wide-survey observations, allowing correction of crosstalk from unsaturated stars and masking of cross-talk from saturated sources in the science pipeline. The calibration monitors electronic crosstalk between detector readout channels and components.
VS-009<br>Once during PV; daily in routine ops	Darks The VIS dark current is expected to be extremely low ($\leq0.01\ \mathrm{e^{-}\ s^{-1}\ pix^{-1}}$) and could not be fully characterised on ground, though tests showed $< 0.1$ ADU over 350 s integration. Measured once during PV with routine operations acquiring four darks daily ($2 \times325$ s, $1 \times200$ s, $1 \times95$ s) through the ROS, these exposures primarily measure stray sunlight (which never completely disappears even at operational attitudes). They also probe cosmic-ray flux, removal routines, and are used to update the cosmetic map.
VS-010<br>Once (stitch alignment); monthly (self-calibration positions); during PV (distortion)	Focal-plane astrometric solution This calibration has three goals: determining relative stitch block alignment ($256 \times512$ pixel blocks, 128 per detector, measured once using $\geq1000$–$2000$ Gaia stars per detector at 25–50 arcmin$^{-2}$ density); determining relative detector positions/orientations subject to drift (monitored monthly via self-calibration); determining the optical distortion with high angular resolution during PV. The astrometric solution may vary due to detector drift, mechanical relaxation, and thermal effects.
VS-011<br>Weekly in routine phase; twice during PV (4 weeks apart)	Photon transfer curve Weekly calibration during routine phase obtaining the PTC for every detector quadrant to determine the gain, executed twice during PV phase four weeks apart. The PTC curve is affected by electronic nonlinearity (independent of differential nonlinearity) and can yield nonlinearity estimates for comparison with primary methods from on-sky and flat-field measurements.
VS-012<br>Once during PV	Dichroic ghost model Once during PV phase, this validates the ground-based dichroic ghost model (based on Zemax simulations and PLM end-to-end tests) by confirming ghost location and amplitude for sources brighter than 15 AB mag, with the main ghost source being dichroic backside reflection having amplitude ratio requirement $< 2 \times10^{-6}$. While size, shape, and wavelength dependence are expected to match simulations, any inaccuracies will be evident in masked pixel data, with focal plane reflection ghosts already stopped by design and NISP reflection ghosts about four orders of magnitude lower.
ID / Cadence	Description
VS-013<br>Monthly	Large-scale flat Monthly reconstruction of matching broad-band large-scale flats for VIS from self-calibration stellar photometry, removing simultaneously any residual large-scale non-uniformity in LED flats and large-scale illumination effects from baffle vignetting, mirror edges, coating variations on folding mirrors, molecular contamination, VIS shutter map, and apodisation. The large-scale flat characterises illumination variations that cannot be captured by the LED flat-field system alone.
VS-014<br>Once during PV; regularly<br>in routine ops	LED pixel response non-uniformity This calibration obtains zero-signal PRNU for each of six LED wavelengths by extrapolation to zero signal or brighter-fatter effect correction, executed once during PV phase and regularly during ROS routine operations. The calibration flags cold/dead pixels and hot pixels while characterizing pixel-to-pixel response variations across the focal plane using the VIS calibration unit’s monochromatic LEDs.
VS-015	Noise bias calibration fields Replaced by CHAR-TARGET-003.
VS-016	Nonlinearity This calibration characterises the nonlinearity of the VIS CCD response over the full dynamic range. It validates and updates the ground-derived nonlinearity correction using in-flight flat-field data acquired at multiple fluence levels. Accurate nonlinearity correction is essential to preserve photometric accuracy and PSF stability.
VS-017	Phase diversity This product derives wavefront-error information from focussed and defocussed images. The resulting phase diversity analysis constrains low-order optical aberrations and supports the validation of the VIS PSF model. It also provides diagnostic input for monitoring optical stability over time.
VS-018	Extended pixel edge response (CTI-EPER) This calibration measures charge trailing caused by CTI using extended pixel-edge response (EPER) techniques. The data quantify deferred charge and trap-related effects that impact photometry, astrometry, and shape measurements. The results are used to validate and refine the VIS CTI correction models.
VS-019	PSF calibration This product provides an empirical characterisation of the VIS PSF using stellar observations. It captures spatial and temporal PSF variations across the FPA under nominal observing conditions.
VS-020	Blooming threshold This calibration determines the signal level at which blooming occurs, using high-fluence flat fields and, optionally, stellar observations.
VS-021	Straylight check This product assesses the level of straylight contamination in VIS images. It checks for glints from stars falling onto physical features in the FPA, such as detector or baffle edges. It also assesses the normalised detector irradiance (NDI) reaching the detectors from in-field and out-field sources.
VS-022	Small-scale flat field This calibration provides the PRNU and other small-scale variations that are not captured by the LSF. The PRNU map is computed from different LED flats to be representative for the broad VIS passband.
VS-023	Zodiacal model Replaced by CHAR-SURV-004.
VS-024	Flatband voltage shift This calibration assesses radiation damage in the detectors’ metal-oxide semiconductors, leading to a shift in the flat-band voltage.
VS-025	Trap pumping This product identifies and characterises charge traps using trap-pumping techniques. It measures trap densities, release timescales, and spatial distributions. The results feed directly into CTI correction models and radiation damage monitoring.
VS-026	CTI-induced covariance This calibration quantifies pixel-to-pixel noise correlations induced by CTI effects. Such covariances impact shape measurements and noise modelling if left uncorrected.
VS-027	Persistence This determines any residual signal persistence following illumination by bright sources. It measures persistence amplitudes and decay timescales under in-flight conditions.
VS-028	Shuttermap Provides a map of the effective exposure-time modulation by the VIS blade shutter. This map is required to separate mechanical from optical effects in the LSF, and to achieve consistent photometric zero-points across exposures with different integration time.