SOLO: wide-field asteroid light curve monitoring system for SPHEREx Bumhoo Lim, Seungwon Choi, Yoonsoo P. Bach, Masateru Ishiguro, Sunho Jin, Carey M. Lisse, Max Mahlke, Jooyeon Geem, Jinguk Seo, Sihu Ahn, Hangbin Jo Abstract 0000-0002-8244-4603 0009-0006-1028-1653 0000-0002-2618-1124 0000-0002-7332-2479 0000-0002-0460-7550 0000-0002-9548-1526 0000-0003-2831-0513 0000-0002-3291-4056 0009-0004-9591-8646 1]Department of Physics and Astronomy, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea 2]SNU Astronomy Research Center, Department of Physics and Astronomy, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 08826, Republic of Korea 3]Korea Astronomy and Space Science Institute, Daejeon 305-348, Republic of Korea 4]Johns Hopkins University Applied Physics Laboratory, 11100 Johns Hopkins Rd, Laurel, MD 20723, USA 5]Université Marie et Louis Pasteur, CNRS, Institut UTINAM (UMR 6213), équipe Astro, F-25000 Besançon, France 6]Asteroid Engineering Laboratory, Luleå University of Technology, Kiruna, Sweden \jkashead 1 Introduction Small bodies in the Solar System (e.g., asteroids and comet nuclei) are primitive remnants of its formation. Most of them are smaller than a few tens of kilometers in diameter and therefore typically exhibit irregular shapes that cause periodic brightness variations due to their rotation or tumbling motion (the so-called light curves) and due to changes in the solar phase angles (i.e., the angle between Sun—target—observer). Light curves reveal key physical properties of asteroids, such as shape, rotation period ( P_{\mathrm{rot}} ), and spin-axis orientation. Since their rotational states reflect collisional evolution and thermal effects such as the YORP torque, investigating their light curves is of great importance for understanding the physical and dynamical evolution of small bodies in the Solar System (e.g., Davis et al. 1985 ; Pravec et al. 2008 ). To obtain reliable light curves, it is important to measure asteroid fluxes at short time intervals of a few minutes, given that typical rotation periods range from several hours to several tens of hours ( Pravec & Harris 2000 ; Warner et al. 2009 ) . Except for a small number of fast-rotating asteroids with periods below one hour, such temporal resolution is sufficient for resolving rotational brightness variations. Conventionally, light curve observations have been performed for a limited number of targets of particular interest using "relative" photometry with comparison stars in the same field. These relative photometric data have been compiled in the Database of Asteroid Models from Inversion Techniques (DAMIT; Ďurech et al. 2018 ) and the Asteroid Lightcurve Database (LCDB; Warner et al. 2009 ), which have been used to reconstruct three-dimensional asteroid shape models through light curve inversion techniques ( Durech et al. 2010 ) . Even this most comprehensive catalog (i.e., DAMIT) contains, as of January 2026, well-established light curves for only \sim 300 asteroids (defined by the quality flag 1 1 1 https://damit.cuni.cz/projects/damit/pages/documentation \geq 2.5 ), representing only \sim 0.01~\% of the known population. This clearly demonstrates that the number of asteroids with reliably determined light curves remains very limited. In recent years, large-scale sky surveys such as the Pan-STARRS ( Chambers et al. 2016 ) , the Zwicky Transient Facility (ZTF; Bellm et al. 2019 , and the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST; Ivezić et al. 2019 ) have made it possible to provide time-series photometric data of hundreds of thousands of asteroids with excellent sensitivity. However, their cadence is generally limited (on the order of a few observations per day), which is insufficient to construct accurate light curves. Nevertheless, steady efforts have been made to estimate spin-axis orientations and rotation periods by simultaneously modeling the three-dimensional shape and the phase function (see, e.g., Carry et al. 2024 ). The rotational properties inferred from asteroid light curves are essential to obtain further physical properties. In 2024, our research group launched a program to construct reflectance spectra of small Solar System bodies using data from the Spectro-Photometer for the History of the Universe, Epoch of Reionization, and Ices Explorer ( SPHEREx ) mission ( Crill et al. 2020 ; Ivezić et al. 2022 ; Bock et al. 2025 ) . Launched in March 2025, SPHEREx will carry out a long-term spectroscopic survey for 2 years covering the wavelength range of 0.75–5.0 \micro m, which will make it possible to produce spectral data of individual objects by measuring their fluxes at different wavelengths and at different epochs during repeated all-sky scans. However, to utilize these SPHEREx data for studies of asteroids, it is essential to characterize their rotational brightness modulation and changes in phase angle. This is because, during the SPHEREx spectral scans, asteroids rotate multiple times and the phase angle changes, producing flux variation that cannot be readily distinguished from intrinsic spectral features. SPHEREx will observe more than \approx 10^{5} small Solar System bodies ( Crill et al. 2020 ) , which is much larger than the number of asteroids with observed light curves. To enable detailed spectral characterization via light curve analysis, we constructed a new telescope system, the Solar system Objects Light curve Observatory (SOLO), at the Sierra Remote Observatories (SRO) in California, USA, in July 2025. Equipped with a wide field-of-view (FoV) of approximately 11.6~\mathrm{deg}^{2} , SOLO is dedicated to monitoring asteroid light curves with high cadence. In particular, over the next two years, SOLO aims to provide ancillary light curve data for the SPHEREx Solar System Objects Catalog ( SPHEREx SSOC). This will be achieved by coordinating nearly simultaneous ground-based observations with the SPHEREx survey regions, enabling us to trace and correct the flux variability of asteroids within the SPHEREx field of view and thereby improve the reliability of the corrected SPHEREx spectral database (Section 4 ). This paper presents an overview of the SOLO system. Section 2 summarizes the system’s configuration and technical characteristics. Section 3 describes the system performance, the data calibration pipeline, and presents example asteroid light curves obtained with SOLO. Section 4 outlines our observing strategy to maximize light curve acquisition and highlights SOLO’s contribution to correcting SPHEREx data. 2 System Characteristics In this section, we summarize the system characteristics of SOLO. Section 2.1 provides the specifications of the telescope, detector, and filter, and the details of the observatory site. Section 2.2 describes the operational system and data reduction pipeline. 2.1 Site and Instrument Overview Figure 1: Overview of the SOLO observation system after its installation at the Sierra Remote Observatories (SRO) in California, USA, in July 2025. Table 1: Specifications of the SOLO telescope (RASA-11) and CMOS camera (Kepler 4040FI) Effective diameter 279 mm (11-inch) Focal length 620 mm f-ratio f/2.22 Pixel array 4096 \times 4096 Pixel size 9 \mathrm{\mu m} Pixel scale 2^{\prime\prime}.97 \mathrm{pixel^{-1}} FoV 3°.4 \times 3°.4 (11.6 \deg^{2} ) Readout time \sim 2 s Readout noise 34.1 \mathrm{e^{-}} Full well 4096 ADU Linearity \sim 3800 ADU Effective gain 18.69 \mathrm{e^{-}/ADU} Dark current a 0.08 \mathrm{e^{-}\ pixel^{-1}s^{-1}} \tabnote The camera specifications (readout noise, full well, linearity, effective gain, and dark current) are the values at 1\times 1 HDR mode and 2.8\times PGA gain setting. a The dark current was measured at -10\degree C. Figure 2: Design of the custom spider assembly for SOLO (shown in red). The circular interface on the front-left side attaches to the telescope entrance aperture, while the dark-gray structure on the rear-right side represents the mounted CMOS camera. A faint outline of the spider is shown in the background to illustrate its projected side view and overall geometry. Figure 3: Expected system throughput of SOLO, considering detector QE (dashed), filter transmittance (dotted), and atmospheric transmission (dash-dotted). The black solid line shows their product (i.e., system throughput). For comparison, the green line shows the filter transmittance of Gaia G -band. We adapted the atmospheric model constructed by lowtran ( Kneizys et al. 1988 ) . Figure 1 shows the SOLO system after its installation. Considering the short development period of about one year, the observing system was designed using a combination of commercially available components. Based on the compatibility between available commercial telescopes and instruments, as well as the FoV matching with single detector of SPHEREx (3°.5 × 3°.5; Bock et al. 2025 ), we adopted an 11-inch Rowe-Ackermann Schmidt Astrograph (RASA-11) manufactured by Celestron 2 2 2 https://www.celestron.com/ . The telescope has a prime-focus optical system with an effective diameter of 0.28 meters (11 inches). With this design, the telescope achieves fast optics with an effective focal ratio of f/2.2, providing good image quality over a 3-degree diameter circle without notable distortion. Combined with ancillary corrective lenses, it delivers a coma-free flat field without significant optical aberration over a 43 mm diameter circle on the focal plane. The telescope is paired with a RainbowAstro 3 3 3 https://www.rainbowastro.com/ RST-400 mount. Because the fast optics provide relatively low spatial resolution and the mount offers sufficient tracking performance within the given exposure times (see below), we have not implemented an auto-guiding system at this stage in order to maintain simplicity of operation. We adopted the Kepler 4040 FI camera manufactured by Finger Lakes Instrumentation 4 4 4 https://www.flicamera.com/ (FLI) as the focal-plane detector, which provides a large-format sensor while being relatively easy to obtain. The camera employs a GSense4040 front-illuminated CMOS sensor with a peak quantum efficiency (QE) of 70 %, featuring a 4096\times 4096 pixel array with 9 µm pixels. This configuration yields a pixel scale of 2^{\prime\prime}.97 and provides a wide FoV of 3°.4 × 3°.4. The key system specifications are summarized in Table 1 . It should be noted that when this large-format camera is coupled with this fast optical system, mechanical flexure of the assembly can cause focal-plane tilt or small variations in focus depending on the telescope orientation. To ensure mechanical stability, we developed a custom-designed camera holder (or “spider”, Fig. 2 ) that rigidly supports the CMOS camera at the entrance aperture and maintains its parallelism with the focal plane. Finite-element calculations indicate that flexure in the X–Y plane corresponding to field shifts on the detector is more than two orders of magnitude smaller than the pixel size and, therefore, negligible. Potential flexure along the Z direction (i.e., focus shift) was assessed experimentally. As shown in Section 3.1 , the point spread function (PSF) remains stable across the FoV, demonstrating that any focus variation induced by the camera load is negligible in practice. These facts confirm that the spider assembly introduces no measurable degradation in either PSF uniformity or photometric performance. Since the primary goal of SOLO is to correct the apparent brightness variations of asteroids caused by their rotation (i.e., changes in projected cross section), we adopted a single-filter system to achieve high photometric precision and maximize observing efficiency. We installed a SCHOTT WG320 long-pass filter manufactured by Edmund Optics 5 5 5 https://www.edmundoptics.co.kr . This UV-cut clear filter provides high optical throughput across the visible range, which is well suited for maximizing the photometric sensitivity of SOLO. The combined system throughput (taking into account the detector quantum efficiency, filter transmittance, and the typical atmospheric transmission) is shown in Figure 3 . Although the detailed efficiency curves for the RASA-11 corrector plate and primary mirror are not publicly available, the manufacture (Celestron) indicates that the telescope efficiency decreases toward wavelengths below 400 nm. Overall, as a result, the effective bandpass of SOLO is designed to roughly match that of the Gaia G -band (green line in Figure 3 ), and this consistency is quantitatively demonstrated in Section 3.2 through an absolute photometric calibration between SOLO and the Gaia G magnitudes at the \approx 3\% level. The system is hosted at SRO in the Sierra Nevada Mountains, California, USA (37°.07N, 119°.40W; 1405 m altitude). According to information provided by SRO 6 6 6 https://www.sierra-remote.com/ , the site offers excellent observing conditions, with approximately 290 clear nights annually and a measured sky brightness of 21.78 \rm{mag\ arcsec^{-2}} in the V -band. The typical seeing is \sim 1.5^{\prime\prime} . The selection of the SRO site was primarily motivated by its high clear-sky fraction and dark sky background. Because the primary goal of SOLO is to obtain time-series photometry in coordination with the SPHEREx observations over as many nights as possible, a site with excellent weather statistics was essential. In addition, our detector (see below) has a relatively large pixel scale, making the signal-to-noise ratio (S/N) of faint targets highly sensitive to the night-sky brightness. Considering these requirements, we evaluated several candidate sites both within and outside South Korea, and finally selected SRO as the site that best satisfies these conditions. SRO provides roll-off-roof dome infrastructure and on-site maintenance as part of its telescope-hosting service and, we operate SOLO alongside other remotely managed telescopes within this facility. 2.2 System operation and data reduction Figure 4: Schematic diagram of the SOLO system configuration and data-processing pipeline. Solid and dashed lines indicate power and data connections, respectively. The three observational subsystems (Equipments), the RST-400 mount, the Kepler 4040 FI camera, and the ZWO EAF focuser, are installed at Sierra Remote Observatories (SRO) and are controlled by an on-site PC located in the control box. The SRO PC is dedicated to device control and data acquisition and receives power supply and internet connectivity from the SRO infrastructure. The system is remotely operated from Seoul National University (SNU), South Korea, via the internet. The raw data acquired at SRO are transferred to a dedicated SOLO processing server at SNU, where the primary data reduction and photometric analysis are performed, and the resulting data products are stored in the SOLO data server. Figure 5: An example image from SOLO, taken on July 22, 2025, with 180 sec exposure. (a) Raw image, (b) calibrated image, and (c) mask image. In panel (a), strong vignetting is visible in the four corners, along with additional shading on the top, bottom, and left edges caused by the square filter holder. After applying the flat-field correction, these effects are largely removed in panel (b), resulting in a much more uniform background across the entire field of view. Panel (c) shows an example of the mask image, where bright stars, satellite trails, and bad pixels are masked out to exclude them as outliers during the data calibration. Overall operation and data process is shown in Figure 4 . SOLO is operated remotely via a custom telescope control system (TCS). Located at SRO, the system is managed through remote access to an on-site host computer ("Control Box" in Figure 4 ). The three instruments, comprising the mount (RST-400), the camera (Kepler 4040 FI), and ZWO 7 7 7 https://www.zwoastro.com/ EAF focuser, are controlled via the Astronomy Common Object Model 8 8 8 https://ascom-standards.org/ (ASCOM) Alpaca protocol, which provides a standardized interface for astronomical device communication. Using this protocol, we constructed a Python-based TCS that issues Alpaca commands through a command-line interface. Through this interface, we not only control basic device functions but also execute complex routines such as auto-focusing, field identification, and acquisition of calibration frames (e.g., bias and dark images). This system enables fully robotic operation of the SOLO survey observations (Section 4 ). At the end of each night, the SOLO data are processed through a dedicated reduction pipeline that performs the standard calibrations required for astronomical CMOS data. To reduce network traffic and computational load at SRO, these data processing tasks are executed on a dedicated SOLO processing server installed at Seoul National University (SNU; see the right panel of Figure 4 ). The procedure begins with astrometric calibration (WCS solution), followed by bias subtraction, dark correction, and flat-fielding (Figure 5 b). We then apply a bad-pixel mask to exclude problematic regions such as saturated pixels, hot pixels (dark current >10\ \mathrm{e^{-}\ pixel^{-1}\ s^{-1}} ), and sensor edges (Figure 5 c). Here, we also remove the linear features (primarily resulting from satellite trails) and pixels likely affected by nearby bright sources identified using sep ( Bertin & Arnouts 1996 ; Barbary 2016 ) . Following calibration, we determine the photometric zero point by cross-matching field stars with Gaia Data Release 3 (DR3) G -band sources ( Gaia Collaboration et al. 2023 ) , excluding bad pixels from this estimate (Lv.1 data in Figure 4 ). Afterward, asteroid candidates in the field are identified, and aperture photometry is performed to derive their instrumental magnitudes. These are then calibrated to apparent magnitudes using the derived zero point (Lv.2 data in Figure 4 ). All data reduction processes and the documentation are implemented in the Python-based pipeline solopy 9 9 9 https://github.com/bumhooLIM/solopy , which makes extensive use of widely adopted astronomical libraries such as astropy ( Astropy Collaboration et al. 2013 ) and astroquery ( Ginsburg et al. 2019 ) , along with other supporting packages ( Bertin & Arnouts 1996 ; Craig et al. 2015 ; Barbary 2016 ) . 3 Performance evaluation In this section, we report the initial performance analysis of the SOLO system. Given its wide field of view and fast optics, we first investigated its optical performance, including vignetting and the PSF, in Section 3.1 . We then performed the absolute photometric calibration against the Gaia sources in Section 3.2 . Finally, we estimate the system’s limiting magnitude (Section 3.3 ) and present sample asteroid light curves (Section 3.4 ). 3.1 Optical performance Figure 6: Median-stacked and normalized master sky flat of SOLO. Each individual frame was normalized by its respective median value prior to stacking, resulting in a master flat with a median value of approximately unity. Toward the field edges, the relative throughput decreases to \lesssim 70-80 % of the reference value; however, this spatial sensitivity variation can be properly corrected using this median-stacked flat field, albeit with a corresponding reduction in signal-to-noise ratio. Figure 7: Spatial distribution of the point spread function (PSF) across the SOLO field of view. The full field is divided into 500\times 500 pixel subregions, with each panel displaying a mean-stacked PSF image for its respective area. Given the large pixel scale of SOLO ( 2^{\prime\prime}.97 pixel -1 ), the majority of the PSF flux is confined within only a few pixels in all subregions, indicating a compact and spatially consistent PSF over the field of view. To evaluate the optical performance of SOLO, we first constructed a master sky flat to characterize the system’s vignetting (Figure 6 ). Because of the wide field of view, standard twilight flats were found to be affected by sky gradients and thus unsuitable for flat-fielding. In such a wide-field system as SOLO, the twilight sky brightness is not spatially uniform across the field, leading to systematic flat-field errors if used directly. Alternatively, we generated the master flat using 1,132 night-sky images taken between 2025 July 17 and August 30. These frames were selected to have exposure times longer than 60 seconds and elevations >50\degree to ensure a uniform background. The master flat was created by median-stacking these data after applying sigma clipping and masking bad pixels and stellar sources (identified using sep ; Bertin & Arnouts 1996 ; Barbary 2016 ). The resulting flat field shows a consistent response across most of the FoV. However, significant vignetting is present near the edges, where the relative flux decreases to <70\% of the central value. This strong vignetting arises because the telescope’s corrected image circle (43.3 mm) is smaller than the sensor’s diagonal (52 mm), which is consistent with the manufacturer’s specifications for the Celestron RASA-11 (Celestron Inc., 2025 10 10 10 https://www.celestron.com/products/11-rowe-ackermann-schmidt-astrograph-rasa-11-v2-optical-tube-assembly-cge-dovetail ). It should be noted that the region outside the manufacturer-defined optimized image circle does not represent a hard cutoff for data acquisition. Rather, the throughput gradually decreases toward the field edges, leading to reduced illumination rather than a loss of usable data. In this work, we therefore attempt to utilize photometric data beyond the telescope’s corrected image circle (43.3 mm) by applying an appropriate flat-field correction. While this vignetting pattern is evident in the raw image (Figure 5 a), it is largely corrected by the master flat (Figure 5 b), rendering the data at the field edges scientifically useful. The optical design could also cause distortion of the point spread function (PSF), especially near the edges. Figure 7 shows the spatial distribution of the average PSF, measured in a 500\times 500~\mathrm{pixel^{2}} grid on a 180 s calibrated exposure. In each grid, we selected up to 20 bright, non-saturated stars to construct an average PSF. The system’s large pixel scale ( 2^{\prime\prime}.97~\mathrm{pixel^{-1}} ) results in an undersampled PSF; consequently, >99\ \% of the source signal falls within a 3\times 3~\mathrm{pixel^{2}} ( 9\times 9~\mathrm{arcsec}^{2} ) area, largely independent of seeing conditions. We fitted the PSF with 1D Gaussian functions along the x and y directions after resampling it onto a 2\times 2 sub-pixel grid. The resulting PSF characteristics are consistent across the FoV, achieving a FWHM <2 pixels (6 arcsec) and an ellipticity b/a\sim 0.9 in all grids. Given the relatively large pixel scale of SOLO, aperture photometry is performed using a large aperture radius (typically \approx 4 pixels). As a result, any moderate degradation of the PSF toward the field edges in the RASA-11 optical system has a negligible impact on the photometric measurements. We conclude that SOLO maintains a consistent optical performance across its field of view for a 180 s exposure. 3.2 Photometric calibration Figure 8: Results of aperture photometry for 10,062 Gaia sources ( 1313 mag), and iv) were isolated (no other Gaia sources found within 30^{\prime\prime} ). This selection yielded 10,558 sources in the field. We find that m_{\mathrm{SOLO}} versus G magnitudes shows a linear correlation over the analyzed magnitude range ( 13