SST-1M: Single-Mirror Cherenkov Telescope

• SST-1M is a small Imaging Atmospheric Cherenkov Telescope with a single-mirror Davies–Cotton configuration and SiPM-based, fully digital camera.
• The design emphasizes mechanical simplicity, cost-effectiveness, and reliable performance under elevated night-sky backgrounds, enabling stereoscopic operations.
• Commissioning results demonstrate competitive angular and energy resolutions in mono and stereo modes for TeV–multi-TeV gamma-ray astronomy.

The Single-Mirror Small Size Cherenkov Telescope (SST-1M) is a small Imaging Atmospheric Cherenkov Telescope based on a single-mirror Davies–Cotton optical configuration and an SiPM-based, fully digitizing camera. It was developed by a consortium of institutes in Switzerland, Poland, and the Czech Republic as a prototype SST concept for very-high-energy gamma-ray astronomy, with particular emphasis on the TeV and multi-TeV domain, wide-field observations, and industrially reproducible construction. In its present realized form, SST-1M is represented by a two-telescope stereoscopic system at the Ondřejov Observatory in the Czech Republic, where commissioning, calibration, and astrophysical observations have been carried out in mono and stereo modes (Tavernier et al., 2024, Alispach et al., 2024).

1. CTA context and instrument concept

SST-1M was conceived within the Small-Sized Telescope segment of the Cherenkov Telescope Array program, where SSTs were intended to dominate sensitivity at the highest energies. The literature describes this role in slightly different energy-language depending on epoch: the 2016 design paper describes SST-1M as optimized for gamma rays from 5 to 300 TeV, whereas later commissioning papers state a design goal of 1 to 300 TeV (Heller et al., 2016, Tavernier et al., 2024). This shift in wording is explicit in the published record and should not be conflated with a single fixed performance number.

The design philosophy emphasized a compact, simple, reliable, and cost-effective single-mirror solution, in contrast to the dual-mirror Schwarzschild–Couder SST concepts discussed contemporaneously in CTA studies. The project literature repeatedly associates SST-1M with mechanical simplicity, off-the-shelf industrial components, and a fully digital SiPM camera architecture intended to support large-scale deployment and operation under elevated night-sky background. At the same time, the collaboration’s later performance paper notes that SST-1M had earlier been judged compliant with CTA requirements during Kraków commissioning, but was not selected as the final CTA SST design (Montaruli et al., 2015, Jurysek et al., 2023).

The realized system therefore occupies a dual position in the literature: it is both a CTA-derived prototype lineage and an operational stereoscopic research instrument in its own right. This distinction matters because some papers discuss projected CTA-style array roles, whereas later papers report measured performance from the Ondřejov two-telescope system under low-altitude observational conditions (Alispach et al., 2024, Alispach et al., 2 Jun 2025).

2. Optical configuration and structural realization

SST-1M uses a single-mirror Davies–Cotton reflector with a 4 m class dish, a focal length of 5.6 m, and a field of view of about 9°. The reflector is segmented into 18 hexagonal facets, each 78 cm flat-to-flat, mounted on a spherical surface; the facet radius of curvature is 11.2 m, consistent with the Davies–Cotton prescription. The geometrical mirror area is reported as 9.42 m², while effective reflecting area after shadowing and reflectivity is reported as approximately 6.47 m² or ≈6.5 m² depending on the source and rounding convention (Tavernier et al., 2024, Seweryn et al., 2015, Alispach et al., 2024).

Optical performance was a central design criterion. Commissioning and optical-development papers report an optical PSF smaller than 0.25° over the required off-axis range, with representative values of 0.09° on-axis and 0.21° at 4° off-axis in commissioning summaries, and 0.08° on-axis with 0.21° at 4° off-axis in later stereoscopic-system descriptions (Tavernier et al., 2024, Alispach et al., 22 Jul 2025). Earlier optical-development work reported on-axis D80 values around 0.07° and emphasized compliance with CTA-style off-axis requirements (Seweryn et al., 2015). The optical time spread is likewise reported as very small: <0.84 ns in the commissioning overview, 0.244 ns on-axis in camera/optics studies, and <2 ns across the camera in the validated end-to-end description (Tavernier et al., 2024, Heller et al., 2016, Alispach et al., 2024).

The mechanical structure is predominantly steel, with a total mass reported at about 8.6 metric tons. The mount is Alt–Az, with elevation and azimuth drives based on slew drives and dual servo motors. Reported motion ranges are −14° to +91° in elevation and −270° to +270° in azimuth; rapid slewing to any direction in less than 1 minute is stated in the design papers. Pointing accuracy is quoted as 7 arcsec, and tracking accuracy as <5 arcmin in the 2016 system paper; later end-to-end validation reports pointing accuracy after model application within ≈70 arcsec (Heller et al., 2016, Alispach et al., 2024). Structural FEM and experimental studies in Kraków reported low relative camera–dish deformations, survival under 200 km/h winds in parked configuration, and lowest modal frequencies above the CTA minimum requirement discussed in early design papers (Niemiec et al., 2015, Niemiec et al., 2013).

Alignment is implemented with actuator-based facet control. The published descriptions mention an automatic optical feedback system, Bokeh-based initial alignment, later consideration of the NAMOD technique, a PSF-monitoring screen on the camera lid, and CCD-based optical feedback from the dish center (Heller et al., 2016, Alispach et al., 2024). The literature does not present a single immutable alignment workflow; rather, it documents an evolving alignment stack across prototype and operational phases.

3. DigiCam camera, SiPM photodetection plane, and triggering

The SST-1M camera, DigiCam, is built around a compact Photo-Detector Plane composed of 1296 pixels in a hexagonal arrangement. In the detailed camera papers, these 1296 pixels are organized into 108 modules of 12 pixels each. Each pixel couples a hexagonal light concentrator or hollow light guide to a custom large-area hexagonal Hamamatsu SiPM, and the linear entrance size is reported as 23.2 mm, corresponding to an angular pixel size of approximately 0.237° or, in some CTA-wide summaries, 0.24° (Tavernier et al., 2024, Heller et al., 2016, Heller et al., 2016).

The camera enclosure uses a Borofloat entrance window with anti-reflective coating and a long-wavelength cutoff to suppress NSB. Different papers report this cutoff as above 540 nm or above 550 nm, reflecting different generations or descriptions of the optical filter stack (Heller et al., 2016, Alispach et al., 2024). The camera is water-cooled, with cooling power around 2 kW in the detailed design reports, and is housed in an IP65-compliant compact enclosure with mass below 200 kg (Heller et al., 2016, Heller et al., 2016).

DigiCam implements fully digital readout and triggering with FPGA-based processing. The canonical sampling rate is 250 MHz with 12-bit FADCs. The detailed architecture comprises three microcrates, each with nine digitizer boards and one trigger board; each digitizer board handles 48 pixels. Signals are continuously digitized into ring buffers, allowing dead-time-free or almost deadtime-free operation up to the quoted rate domain of a few kHz to several kHz, depending on paper and operational context (Heller et al., 2016, Alispach et al., 22 Jul 2025). Trigger formation is described at several levels in the literature: early camera commissioning emphasizes patch-based and cluster-based FPGA logic; the 2016 design paper describes pattern-recognition trigger algorithms for circular, elliptical, and ring-like topologies; and performance studies discuss 7-pixel or 7-patch configurations as practical operating modes under NSB (Samarai et al., 2017, Heller et al., 2016, Moderski et al., 2015).

SiPM operation under elevated NSB is one of the defining SST-1M themes. The camera papers emphasize robust performance under dark-sky, moonlight, and half-moon conditions, DC coupling for baseline-based NSB monitoring, active bias compensation using per-pixel NTC readouts, and the possibility of dynamic threshold control (Schioppa et al., 2015, Heller et al., 2016). Later commissioning work at Ondřejov also documents NSB-induced gain and PDE drops, different bias-resistor choices between the two cameras, and run-wise corrections based on pedestal baseline shifts (Alispach et al., 2 Jun 2025).

Stereo synchronization evolved across the project. A 2024 commissioning-data analysis states that no dedicated stereoscopic hardware trigger was available during the March 2023 Crab campaign and that stereo coincidence was reconstructed from local timestamps (Tavernier et al., 2024). Later commissioning and performance papers describe SWAT-based stereo triggering and White Rabbit synchronization with nanosecond precision (Tavernier et al., 2024, Juryšek et al., 22 Jul 2025). This sequence is a developmental chronology rather than a contradiction.

4. Calibration, reconstruction pipeline, and data products

SST-1M data reduction is organized around sst1mpipe, a pipeline built on ctapipe that transforms raw instrument data into DL3 products compliant with GADF standards; higher-level products such as sky maps and spectra are generated with gammapy (Tavernier et al., 2024, Jurysek et al., 2023). In later performance work, IRFs are computed with pyirf, and the full reconstruction chain is anchored to Monte Carlo simulations generated with CORSIKA and sim_telarray (Alispach et al., 2 Jun 2025).

Calibration proceeds on several levels. Dark runs at the beginning and end of each night are used for per-pixel gain, dark-count, and optical-crosstalk characterization. One commissioning paper models multi-photoelectron spectra with a generalized Poisson distribution and defines the crosstalk-corrected conversion factor as

$g^*=\frac{1-\mu}{g},$

with charge extraction

$$Q=\sum_i (s_i-p_i), \qquad N_{\rm pe}=\frac{Q}{g^*}.$$

The same paper reports that 20–30 pixels in SST-1M-2 exhibited lack of gain during part of commissioning, with maintenance ongoing (Tavernier et al., 2024).

Muon-ring analysis constitutes the main end-to-end optical-throughput calibration strand. The dedicated muon-calibration study had already identified muon rings as a throughput calibrator for SST-1M in simulation (Toscano et al., 2015). Ondřejov commissioning papers then used muon rings operationally, reporting D80 values of 10.5 ± 0.1 mm for SST-1M-1 and 10.8 ± 0.1 mm for SST-1M-2 from muon data, together with an MC optical efficiency 10–20% higher than measured in both telescopes if uncorrected (Tavernier et al., 2024). The 2025 Crab performance paper extends this with continuous muon-ring throughput calibration and atmosphere-tuned MC at low altitude (Alispach et al., 2 Jun 2025).

Event reconstruction is based on calibrated waveforms, image cleaning, Hillas parameterization, and random-forest regression and classification. In the detailed 2025 performance paper, charge extraction uses LocalPeakWindowSum with an 8-sample window; cleaning uses adaptive two-stage tailcuts with temporal coincidence; mono direction reconstruction uses DISP random forests; stereo reconstruction combines timestamp matching, axis-crossing geometry, and stereo DISP; and energy plus gamma–hadron separation are trained on MC features including intensity, width/length, timing slope, skewness, kurtosis, leakage, centroid coordinates, shower-maximum height, and impact distance (Alispach et al., 2 Jun 2025). The basic IRF definition is given as

$$A_{\rm eff}(E)=\frac{N_{\gamma,\rm sel}(E)}{N_{\gamma,\rm gen}(E)}\times A_{\rm gen},$$

and the energy-resolution variable is

$$\Delta E/E=\frac{E_{\rm rec}-E_{\rm true}}{E_{\rm true}}.$$

Background estimation depends on analysis product. Commissioning sky maps use the ring background method, with one 2024 proceedings paper specifying inner and outer radii of 1.0° and 1.3° for Crab maps (Tavernier et al., 2024). The later Crab performance paper uses reflected-region background estimation for wobble observations, and the VER J2019+368 analysis uses ring background, FoVBackgroundMaker, and reflected-region methods in different stages (Alispach et al., 2 Jun 2025, Juryšek et al., 22 Jul 2025).

5. Simulated expectations and measured low-altitude performance

Pre-commissioning simulations described a single SST-1M telescope with an energy threshold of 0.3 TeV, defined as the energy at which the differential gamma trigger rate peaks, and a Crab-like single-telescope trigger rate of 0.27 Hz. The same study derived a safe threshold of 17.2 photoelectrons under the adopted NSB prescription and an on-axis whole-mirror PSF of 0.083 ± 0.005 deg in angular D80 (Moderski et al., 2015). These values belong to an early simulation phase and should not be read as the measured low-altitude stereoscopic performance of the Ondřejov system.

Measured performance at Ondřejov is substantially shaped by site altitude and atmospheric transmission. The detailed Crab validation paper states explicitly that the low altitude of 510 m a.s.l. increases the telescope–shower-maximum distance, reduces Cherenkov photon density at ground level, and raises trigger and analysis thresholds compared with mountain sites (Alispach et al., 2 Jun 2025). For zenith angles below 30°, that paper reports analysis-level thresholds of approximately 1.0 TeV in mono mode and 1.3 TeV in stereo mode. It further reports energy and angular resolutions of approximately 20% and 0.18° for mono mode, and 10% and 0.10° for stereo mode, respectively (Alispach et al., 2 Jun 2025).

These measured results are consistent in direction with earlier 2023 mono/stereo performance studies, which had already found trigger-level thresholds around 1 TeV at 20° zenith, stereo angular resolution around 0.1° above threshold, mono angular resolution below about 0.2°, energy bias below 5% up to about 200 TeV, and roughly a factor-of-two stereo improvement in differential sensitivity relative to mono for 20°–40° zenith (Jurysek et al., 2023). The later low-altitude Crab paper quantifies integral stereo sensitivity above threshold at approximately 7% Crab Units in 50 h and notes that stereo improves differential sensitivity by about a factor of two versus mono (Alispach et al., 2 Jun 2025).

The literature therefore separates naturally into two performance regimes. The first is pre-commissioning design/simulation, where optics, thresholds, and rate models were established under CTA-like assumptions (Moderski et al., 2015). The second is measured stereoscopic operation at Ondřejov, where thresholds are higher but calibration, MC agreement, and astrophysical detections demonstrate a mature instrument response (Alispach et al., 2 Jun 2025).

Published site parameters also differ slightly across papers. Ondřejov is quoted at 500 m in one commissioning summary and 510 m a.s.l. in later papers; telescope separation is quoted as 152.5 m, 155 m, or 155.2 m in different reports (Tavernier et al., 2024, Alispach et al., 2 Jun 2025, Juryšek et al., 22 Jul 2025). This suggests minor differences in reporting convention rather than an instrumental-performance ambiguity.

6. Commissioning observations, astrophysical results, and current limitations

The Crab Nebula has been the principal validation source. In the 2024 commissioning proceedings, stereoscopic Crab observations from November 2023 to March 2024 yielded a total gamma-ray excess of 176.1 events and a Li & Ma significance of 21.3 σ after 22.03 h of stereo livetime following quality selection; the same paper notes that its conclusion section quotes 22.5 h without explaining the discrepancy (Tavernier et al., 2024). An earlier commissioning-data paper based on March 2023 data, before a dedicated stereo trigger was implemented, reported a 5.21 σ Crab detection in 2 h 38 m, with 18 excess events and 28 ± 5 expected from MC (Tavernier et al., 2024). The later full Crab analysis reported 25 h of stereo data after quality selection and measured a power-law spectrum over 2.5–50 TeV with stereo parameters

$$\phi_0=(1.76 \pm 0.12_{\rm stat} \pm 0.40_{\rm syst}) \times 10^{-13}\ {\rm cm^{-2}\ s^{-1}\ TeV^{-1}},$$

$$\Gamma = 2.78 \pm 0.10_{\rm stat} \pm 0.08_{\rm syst},$$

at reference energy 7 TeV (Alispach et al., 2 Jun 2025).

Extragalactic-source monitoring has likewise produced commissioning-era detections. The 2024 proceedings reported a Markarian 421 flux increase on 13 March 2024, with 3.3 h livetime, 6.8 σ significance, emission observed up to 7 TeV, and flux close to the Crab; a further 5.0 h follow-up was obtained on 17 March 2024 (Tavernier et al., 2024). A later “first results” contribution expanded this into a 23.5 h good-quality stereo dataset, reporting an EBL-corrected power-law spectral index of $\Gamma = 3.22 \pm 0.3$ over 1–50 TeV and noting that a fixed-cutoff model gave $\Gamma = 2.6 \pm 0.3$ (Juryšek et al., 23 Jul 2025).

Extended-source capability has been tested on VER J2019+368. The 2025 source paper reported 97 h of good stereoscopic data from April to November 2024, nearly flat gamma-ray acceptance up to about 2.5° offset with only about a 10% drop thereafter, and an extended-search local significance of 8.1σ, corresponding to a 6.5σ post-trial significance (Juryšek et al., 22 Jul 2025). The same analysis derived an asymmetric-Gaussian morphology and a power-law spectrum up to 100 TeV with

$$\Gamma = 2.44 \pm 0.13, \qquad \phi_0 = (3.17 \pm 0.41)\times 10^{-14}\ {\rm cm^{-2}\ s^{-1}\ TeV^{-1}}$$

at 7 TeV (Juryšek et al., 22 Jul 2025). This source study is important because it operationalizes a claim long present in SST-1M design papers: that the wide field of view and controlled off-axis aberrations are particularly suited to extended Galactic sources (Montaruli et al., 2015, Juryšek et al., 22 Jul 2025).

The current literature is also explicit about limitations. Poor weather limited data taking during the winter 2023/2024 campaign; low altitude raises thresholds and increases sensitivity to atmospheric transmission; NSB is handled operationally by higher trigger thresholds or NSB-dependent calibration corrections; and several commissioning papers state that pipeline development and MC tuning remain ongoing (Tavernier et al., 2024, Alispach et al., 2 Jun 2025). Early proceedings often omit final quantitative sensitivity curves, standardized angular-resolution definitions, or fully tabulated IRF components, whereas the later 2025 Crab paper provides a substantially more complete response characterization (Tavernier et al., 2024, Alispach et al., 2 Jun 2025).

A common misconception is to treat every published SST-1M number as if it described a finalized, single configuration. The published record instead spans prototype optics studies, camera-laboratory validation, Kraków structure and Monte Carlo work, early Ondřejov commissioning, and later stereoscopic performance papers. Another common misconception is to read Ondřejov results as direct surrogates for high-altitude CTA-site performance. The low-altitude Crab study explicitly states the opposite: the Ondřejov environment raises thresholds, yet still permits calibrated multi-TeV astrophysics and detailed MC–data validation (Alispach et al., 2 Jun 2025). In that sense, SST-1M is best understood not merely as a historical CTA prototype, but as a technically mature stereoscopic Cherenkov system whose operational literature documents the transition from design concept to instrumented astrophysical observatory.