William Ryan, New Mexico Tech; Eileen Ryan, New Mexico Tech
https://doi.org/10.64861/WVGN6903
Keywords: Cislunar, Space Situational Awareness, Color index, Photometry, Lightcurves, Ground-based observation.
Abstract:
Broadband multicolor photometric data has proven a reliable tool for satellite characterization in the LEO, MEO, and GEO space environments [1-5]. However, extending this technique into the cislunar regime within ~10 degrees from the moon can prove challenging due to lunar contributions to the local sky brightness. In particular, the sky noise, which scales as the square root of the photometric signal of the surrounding sky, can result in large photometric errors and even overwhelm detection of the target at all. For this effort, we assess the feasibility of photometric characterization of meter-sized targets in cislunar space, since these small sizes present a challenging case. Our investigation is based on data collected during a 4-day campaign using the Magdalena Ridge Observatory 2.4-meter telescope to observe 12U CubeSat CAPSTONE in November 2025. Our analysis of the lightcurve data suggests that the signal to noise quality is not only effective for detection, but adequate for photometric characterization. Therefore, we anticipate that the future use of multicolor (two or more bands) photometry for meter-sized targets will be a useful technique in the cislunar region.
Our first test in this study was to discern whether our 2.4-meter telescope could even detect CAPSTONE in the bright sky of the near lunar environment given its small size. Then the data were examined for photometric quality. Observations were collected on November 3, 4, 5, and 6, 2025 UT, under clear, photometric conditions for all but the November 4 date (which had thin clouds). Seeing conditions were ideal at near arc-second to sub-arcsecond quality. CAPSTONE’s brightness varied from V~17.0 to a V~21-22 during this time period, and the CubeSat distance ranged from as close to 5 degrees from the lunar surface, to as far away as an 11-degree separation. Over the duration of the observing run, a total of 660 images were collected in the VR filter, except for November 6, where an additional 95 images were acquired in the R-band filter for color analysis. For this examination, we have focused on the lightcurve data from November 6, since the narrower R-band filter is more appropriate for modeling color difference photometry. On this night, the target was 4.9 degrees from the center of the moon, had an average V of 17.5, and was moving at 35-38 arc-seconds per minute with respect to sidereal. The seeing conditions on November 6 varied over the course of the night from less than 1 to ~1.5 arc-seconds. The telescope followed the nominal track obtained from the JPL Horizons website. In order to keep the field star trails less than 2 seeing widths, exposures were limited to 3 seconds. This permitted CAPSTONE to be detected in most single images acquired, with peak signal-to-noise values of 6 or better.
Aperture photometry was performed with typical Gaia red magnitudes of about 16.5-17.5. Under dark sky conditions, this would normally be considered an easily attainable target even with a modest sized telescope. However, in the bright sky of the near lunar environment, the sky noise resulted in large photometric errors even with a 2.4-meter telescope. Measured photometric errors were on the order of 0.2 mag or greater. Since, unlike LEOs, the solar phase angle and distance were very slowly varying, the target exhibited long periods of near constant brightness. Therefore, improvements gained by co-adding several frames were examined since, in this clearly sky-noise dominated case, signal to noise should increase by the square root of exposure time, or the square root of the number of images when working with equal exposure times. In light of this, the lightcurve we generated was examined for near constant or slowly varying bright regions where one could reliably co-add images without the results being affected by intrinsic brightness variations. Looking at one example region where the magnitude was 17.0, we measured the photometric error to be 0.23 magnitudes. However, photometry of 5 and 10 coadded images resulted in photometric errors of 0.11 and 0.08 magnitudes, respectively. This error reduction was typical of image groupings throughout the data set and brings down the magnitude uncertainties to where color photometry should be a valid characterization tool. Moreover, this temporal sampling permits higher precision lightcurves with 30 – 40 sec granularity, while still allowing detection of larger photometric features occurring on shorter timescales.
Analyzing the intensity of reflected light from a resident space object has long been proven valuable for satellite discrimination and identification. Our future work will seek to acquire additional data with multiple filters to quantify whether multicolor photometry can truly be realized to similar photometric precisions for CAPSTONE and other targets in the near lunar environment. Even though for two color photometry, the errors will add in quadrature, the magnitude of these combined errors should still allow color photometry to be employed for reliable satellite characterization. Testing and quantifying this assumption will be one subsequent objective of our continued characterization analysis.
Date of Conference: September 15-18, 2026
Track: Satellite Characterization