POLAR-2: Next-Gen GRB & Polarimetry Mission

• POLAR-2 is a next-generation wide-field Compton polarimeter designed to make prompt GRB polarization a routine observable while extending pulsar studies.
• It employs 100 detector modules with shortened scintillator bars and SiPM arrays, significantly boosting light yield, effective area, and polarimetric precision.
• International collaboration and comprehensive calibration efforts ensure integrated spectroscopy, rapid transient alerts, and improved GRB polarization measurements.

POLAR-2 is the successor to the POLAR gamma-ray polarimeter and is described in the literature as a next-generation wide-field Compton polarimetry mission for the China Space Station (CSS), developed primarily to make prompt gamma-ray burst (GRB) polarization a routine observable and, secondarily, to extend POLAR’s timing, spectroscopic, and polarimetric studies of pulsars and other high-energy transients (Kole et al., 2021, Li et al., 2019).

1. Mission definition and development trajectory

POLAR-2 is consistently presented as an international mission involving institutes in Switzerland, China, Poland, and Germany, with CSS hosting, power, data downlink, and onboard computing support for transient searches and rapid alerts (Kole et al., 2021, Angelis et al., 2021). The proposal was accepted in July 2019, and published mission schedules subsequently appeared in several forms: a launch planned in early 2024, operations starting around 2025, launch to CSS targeted for mid/end 2027, and deployment around 2028 in a later subsystem-focused prospects paper (Kole et al., 2021, Produit et al., 2023, Angelis et al., 8 Sep 2025, Hulsman et al., 12 Nov 2025, Gill et al., 20 Jan 2026). This published record shows POLAR-2 as a moving development program rather than a single frozen configuration.

The mission concept also broadened over time. Early pulsar-focused treatments described POLAR-2 mainly as a scaled-up version of POLAR with four times the number of detector modules, retaining the same basic detection principle and energy band assumptions used for POLAR (Li et al., 2019). Later mission papers described a larger platform in which the High-energy Polarimetry Detector (HPD) is complemented by a Broad-band Spectrometer Detector (BSD) and a Low-energy Polarization Detector (LPD), so that the polarimetric payload is coupled to onboard spectroscopy, localization, and soft-X-ray polarimetry (Gill et al., 20 Jan 2026, Angelis et al., 8 Sep 2025, Sun et al., 21 Apr 2026).

2. Detection principle and hardware architecture

Like POLAR, POLAR-2 is based on Compton-scattering polarimetry in segmented plastic scintillators. For linearly polarized radiation, the detected azimuthal scattering distribution is modeled as

$$N(\phi) = A \left[ 1 + \mu P \cos 2(\phi - \phi_0) \right],$$

where $P$ is the polarization degree, $\phi_0$ is the polarization angle, $\mu$ is the modulation factor for 100% polarized radiation, and $A$ is a normalization factor (Kole et al., 2021, Angelis et al., 2023). The basic event topology is a two-bar Compton event: an incoming photon Compton scatters in one scintillator bar and is photo-absorbed or scatters again in a second bar, from which the azimuthal scattering angle is reconstructed (Kole et al., 2021).

Across the mission papers, the core polarimeter is described as 100 detector modules containing 6400 plastic scintillator bars, with 64 bars per module in an $8\times8$ geometry (Kole et al., 2021, Angelis et al., 2021, Produit et al., 2023, Angelis et al., 8 Sep 2025). The bar dimensions are repeatedly given as $5.9\times5.9\times125~\mathrm{mm}^3$, reflecting a shortening relative to POLAR’s 176 mm bars to improve signal-to-noise ratio and reduce background (Kole et al., 2021, Angelis et al., 2021, Produit et al., 2023). The readout architecture replaces POLAR’s multi-anode PMTs with SiPM arrays, with dedicated front-end electronics per module, carbon-fibre mechanical housing, an aluminium support grid, and Peltier-based thermal control for the SiPMs (Kole et al., 2021, Angelis et al., 2021, Produit et al., 2023). One design paper gives payload dimensions of about $590 \times 648 \times 700~\mathrm{mm}^3$, a mass of about 150 kg, and power consumption of about 300 W (Kole et al., 2021, Angelis et al., 2021).

Published energy-range descriptions depend on development stage and subsystem definition. Pulsar detectivity papers treated POLAR-2 as preserving POLAR’s 15–500 keV operating band (Li et al., 2019). GRB-oriented mission papers described design or sensitivity emphasis in the 10–1000 keV, 20–800 keV, 30–800 keV, and 40–1000 keV ranges (Kole et al., 2021, Angelis et al., 8 Sep 2025, Hulsman et al., 12 Nov 2025, Gill et al., 20 Jan 2026). Later mission architecture papers separated these by subsystem.

Subsystem	Function	Energy range stated in the literature
HPD	High-energy polarimetry	30–800 keV; 40–1000 keV
BSD	Broad-band spectroscopy and localization	10–1000 keV
LPD	Low-energy polarimetry	2–10 keV

This multiplicity of quoted bands reflects the difference between early scaled-up POLAR descriptions, later mission-level summaries, and subsystem-specific performance papers (Li et al., 2019, Angelis et al., 8 Sep 2025, Gill et al., 20 Jan 2026, Sun et al., 21 Apr 2026).

3. Performance envelope and calibration program

The basic hardware scaling is explicit: POLAR-2 has four times the number of detector modules as POLAR (Li et al., 2019, Angelis et al., 2021). The performance improvement, however, is not presented as a purely geometric factor. Multiple papers state that the effective area for polarization events is approximately one order of magnitude larger than POLAR’s, owing to the combination of larger detector area, shortened bar geometry, improved optical design, and SiPM readout (Kole et al., 2021, Angelis et al., 2021, Produit et al., 2023). A later mission overview states that POLAR-2 exceeds $2000~\mathrm{cm}^2$ at 100 keV and detects about an order of magnitude more photons than POLAR for similar bursts (Angelis et al., 8 Sep 2025).

A major part of this gain is the increase in light yield. POLAR is quoted at about 0.3 photo-electrons/keV, whereas POLAR-2 is quoted at about 1.6 photo-electrons/keV in design studies and about 1.5 photo-electrons/keV in ESRF-validated later development papers (Kole et al., 2021, Produit et al., 2023, Hulsman et al., 12 Nov 2025). This supports substantially lower thresholds: one design paper gives a single-channel low-energy threshold of about 2.5 keV for a 4 photo-electron trigger, compared to about 12.5 keV in POLAR (Produit et al., 2023). Prototype tests are also summarized with an 8 keV single-channel threshold and 5 keV double-channel threshold at room temperature, improving to about 5 keV and 3 keV at about $0^\circ$C (Produit et al., 2023).

Calibration and qualification are correspondingly extensive. Development papers report polarized-beam tests at ESRF, detailed Geant4 optical modeling, proton and neutron irradiation campaigns, vibration and shock testing along all three axes, and thermal-vacuum cycling from $P$0C to $P$1C on a 3×3 sub-polarimeter; the module design is stated to have reached TRL 7 (Angelis et al., 8 Sep 2025). ESRF measurements on prototype detector modules reported a modulation factor of about 15% at 40 keV and about 20%, 30%, and 27% at 60, 100, and 120 keV, respectively, for a single module (Produit et al., 2023). Optical cross talk is described as well below 1% in prototype measurements in one paper, while another reports 1.5–2.5% nearest-neighbor cross talk with optical grease and <4% with first RTV pads (Produit et al., 2023, Angelis et al., 2021). This suggests ongoing optimization rather than a contradiction in the instrument concept.

Radiation-hardness studies support the expected mission lifetime. Proton irradiation of candidate scintillator bars up to 18.7 Gy showed no significant degradation in light yield, emission spectra, or absorption spectra, and that dose corresponds to about 126 years in the “Full Instrument + CSS” scenario (Mianowski et al., 2023). Proton irradiation of Hamamatsu SiPM arrays to 4.96 Gy was reported as equivalent to 62.9 years for silicon photomultipliers inside POLAR-2 with shielding, or 1.78 years when disregarding shielding from the instrument (Mianowski et al., 2022). The dominant effects are increased dark current and dark counts, partially mitigated by low-temperature operation and annealing (Mianowski et al., 2022).

4. GRB science objectives and expected return

The scientific motivation for POLAR-2 is rooted in POLAR’s GRB results. POLAR detected 55 GRBs in about six months, and detailed analyses of the brightest 14 showed time-integrated prompt emission that is generally lowly polarized or fully unpolarized, while time-resolved analyses provided hints of an evolving polarization angle within single pulses (Kole et al., 2021, Angelis et al., 8 Sep 2025, Angelis et al., 2023). Energy-resolved polarization analyses with POLAR did not produce constraining results because of limited statistics (Angelis et al., 2023, Angelis et al., 8 Sep 2025). POLAR-2 is explicitly framed as the instrument required to move from low-statistics integrated measurements to high-quality time-resolved and energy-resolved spectro-polarimetry.

The mission goals are repeatedly stated in terms of three observables: the polarization degree, the polarization angle, and their evolution over time and energy (Kole et al., 2021, Angelis et al., 2021). The intended applications are discrimination among synchrotron, photospheric, and Compton-drag scenarios; tests of ordered versus tangled magnetic fields; constraints on jet geometry and viewing angle; and population studies of prompt polarization distributions (Kole et al., 2021, Angelis et al., 2021, Angelis et al., 8 Sep 2025). One explicit controversy in the literature is the contrast between POLAR’s low-polarization catalog and AstroSat CZTI analyses favoring polarization degrees above about 50% for several bursts; POLAR-2 is designed to resolve such tensions with larger samples and better per-burst precision (Kole et al., 2021).

Several quantitative forecasts recur. POLAR-2 is expected to perform at least 50 high-quality polarization measurements per year with precision equal to or better than the best POLAR measurements, and to probe polarization degrees down to about 10% for about 10 GRBs per year (Kole et al., 2021, Angelis et al., 2021, Angelis et al., 8 Sep 2025, Hulsman et al., 2021). A later overview states that the mission should detect at least about 200 GRBs per year (Hulsman et al., 12 Nov 2025). The HPD prospects paper goes further, coupling synthetic sources to a time-resolved spectro-polarimetric model and an unbinned event-level likelihood. For a pulse fluence

$P$2

it reports that the time-integrated polarization degree can be constrained to an absolute accuracy of about

$P$3

at $P$4, as long as source photons dominate over the background (Gill et al., 20 Jan 2026). The same study estimates about 78 GRBs per year with $P$5 in the HPD bright-burst regime (Gill et al., 20 Jan 2026).

Multi-messenger science is a recurrent secondary driver. POLAR-2 papers state that the instrument should detect GRBs as faint as GRB 170817A and provide rapid alerts to ground-based follow-up facilities, with alert latencies described as within a minute in one overview and about 2 minutes in another (Kole et al., 2021, Produit et al., 2023). This suggests that the mission is intended not only as a polarimeter but also as a prompt-transient node in a broader GRB follow-up network.

5. Pulsar and persistent-source capabilities

Although POLAR-2 is primarily a GRB mission, the Crab-pulsar studies with POLAR established a second line of development. In the pulsar-focused analysis, POLAR-2 is treated as a scaled-up wide-field detector with the same basic 15–500 keV detection principle as POLAR, but with four times as many modules (Li et al., 2019). Geant4 simulations of this preliminary geometry led to the result that, after folding with the phase-averaged Crab spectrum inferred from POLAR data, “the amount of measured photons of POLAR-2 is approximately 10 times that of POLAR in the 15–500 keV energy range,” while the background was conservatively assumed to scale as four times POLAR’s level (Li et al., 2019).

The pulsation significance metric used in both POLAR and POLAR-2 simulations is

$P$6

with $P$7 the total counts in the pulsed phase interval, $P$8 the total counts in the background phase interval, and $P$9 for the Crab (Li et al., 2019). Starting from POLAR’s Crab result, where about 400 hours of exposure produced a total pulsation significance of 58.14 and enabled 40-bin phase-resolved spectroscopy, the simulations showed that POLAR-2 would exceed this significance in only 33 hours of Crab exposure (Li et al., 2019). The same paper states that the averaged pulse significance of POLAR-2 can be seen to be about five times that of POLAR at a given exposure time, and that a three-hour POLAR-2 observation yields a much more significant pulsation than a three-hour POLAR observation (Li et al., 2019).

The longer-term implication is explicit: using about two years of POLAR-2 data, “a significantly more detailed spectral analysis of the Crab pulsar, even of other PSRs, can therefore be performed” (Li et al., 2019). Extrapolating from POLAR’s detection of PSR B0531+21 and PSR B1509-58, the same study concludes that POLAR-2 is promising to detect about 10 pulsars (Li et al., 2019). A companion paper on Crab polarimetry with POLAR emphasizes that the joint-fitting and de-rotation methodology developed for a wide-field, non-pointing Compton polarimeter is directly applicable to POLAR-2 and should support much better phase-resolved and energy-resolved pulsar polarimetry once the higher statistics become available (Li et al., 2021).

6. Auxiliary instrumentation, operations, and scope

The most substantial subsystem addition beyond the scaled-up POLAR concept is BSD, the Broad-band Spectrometer Detector. BSD is a coded-aperture mask instrument using pixelated GAGG scintillation crystals, with a wide half-coded field of view of about $\phi_0$0, an operational energy range of 10–1000 keV, and localization accuracy of approximately $\phi_0$1 for faint GRBs similar to GRB 170817A (Sun et al., 21 Apr 2026). Its simulated on-axis effective area peaks at about $\phi_0$2, and in the 50–300 keV band the 5σ sensitivity is about $\phi_0$3 at $\phi_0$4 keV (Sun et al., 21 Apr 2026). The same study estimates about 127 GRBs per year detected at ≥5σ and about 78 per year within the 120° partially coded field of view, explicitly because BSD’s localization and spectral measurements are required inputs for accurate HPD polarimetry (Sun et al., 21 Apr 2026).

BSD is also described as having moderate capability for GRB polarimetry, particularly at several hundred keV energy (Sun et al., 21 Apr 2026). This does not make it a replacement for HPD; rather, it adds an internally consistent localization-and-spectroscopy channel that POLAR did not possess. A plausible implication is that a large fraction of POLAR-2’s final polarization precision will depend as much on BSD-constrained response modeling as on raw HPD event statistics.

One terminological ambiguity merits explicit clarification. In gamma-ray astronomy, “POLAR-2” denotes the CSS GRB polarimetry mission discussed here. It is unrelated to “POLARBEAR-2,” a ground-based cosmic microwave background polarization receiver that forms part of the Simons Array experiment (Suzuki et al., 2015). The shared shorthand has occasionally created nomenclatural confusion, but the instruments, energy ranges, and scientific goals are entirely different.