© The Authors 2024

1 Introduction

Trans-Neptunian objects (TNOs), along with the objects coming from the Oort cloud, are considered to be the most primordial bodies in our Solar System. The study of their physical and dynamical properties helps us learn about their origin and evolution, which in turn provides crucial information about the origin and history of the early Solar System (Nesvorný & Morbidelli 2012). Currently, the Minor Planet Center (MPC) has counted about 5315 TNOs¹. This includes the population of Centaurs, which are objects believed to be in a transition stage between TNOs and Jupiter-family comets (e.g., Horner et al. 2004; Sarid et al. 2019).

Because TNOs are located in the outer region of the Solar System, they are difficult to study. These objects typically exhibit low brightness, and with an average surface temperature of approximately 30–40 K, their thermal emission peak occurs in the far-infrared spectrum, a range obstructed by Earth’s atmosphere. To derive radiometric sizes, it is necessary to observe them with space telescopes, as was done for more than 120 TNOs and Centaurs within the ESA Herschel mission “TNOs are Cool” open-time key program (e.g., Müller et al. 2009; Lellouch et al. 2013; Farkas-Takács et al. 2020, and references therein). The Herschel mission was completed in 2013.

An alternative to the radiometric technique for deriving sizes and albedos is the use of stellar occultations. The observation of stellar occultations by small bodies (asteroids, comets, Centaurs, TNOs, and planetary moons) of the Solar System is an instrumental relatively simple, but powerful, technique for: directly measuring the size and shapes of these objects with (sub)kilometer accuracy, probing the environment around them with the possibility of revealing a binary nature (Leiva et al. 2020); discovering moons (e.g., Gault et al. 2022) and rings (Braga-Ribas et al. 2014; Ortiz et al. 2015, 2017); and detecting, measuring, or constraining an atmosphere down to the nanobar pressure level (e.g., Hubbard et al. 1988; Sicardy et al. 2003; Oliveira et al. 2022). In addition, the occultation observation provides an astrometric measurement of the occulting object with (sub) milliarcsecond accuracy within the Gaia reference system (Rommel et al. 2020; Ferreira et al. 2022; Kaminski et al. 2023), which can be used to improve the orbit and therefore also predictions of future occultation events. In contrast to occultations by asteroids, predicting and successfully observing stellar occultations by TNOs can be very challenging, mainly due to the very small angular sizes of the TNOs together with the relatively large ephemeris uncertainties.

Trans-Neptunian object (143707) 2003 UY₁₁₇ was discovered by the 0.9 m Spacewatch telescope at Steward Observatory (Kitt Peak, Arizona) on 2003 October 23 (MPEC 2003-V03), shortly before its perihelion in the year 2004. Prediscovery observations date back to the year 2001 (MPEC 2003-Y50). This TNO travels around the Sun in a highly eccentric orbit (e ∼ 0.4), with a perihelion distance of q ∼ 32 au and an aphelion distance of Q ∼ 78 au. The MPC classifies 2003 UY₁₁₇ as a scattered disc object (SDO), but not very much is known about its physical properties. Sheppard (2012) obtained an absolute magnitude of H_R = 5.35 ± 0.03 mag and a color index of V − R = 0.56 ± 0.01 mag, and therefore H_V = 5.91 ± 0.04 mag. Alvarez-Candal et al. (2019) reported H_V = 6.13 ± 0.04 mag, H_R = 5.60 ± 0.04 mag, and V − R = 0.53 ± 0.06 mag.

The geometric albedo, p_V, and diameter, D, of a small body are related via (e.g., Russell 1916; Harris 1998) $$D=\frac{D_0}{\sqrt{p_V}}{10}^{-H_V/5},$$(1)

where H_V is the absolute magnitude of the object, $D_0=2\text{\hspace{0.17em}au}\cdot {10}^{V_{\odot }/5}$ , and V_⊙ is the apparent visual magnitude of the Sun. Values for the apparent magnitude of the Sun are V_⊙ = −26.76 mag (Willmer 2018) and V_⊙ = −26.74 mag (Rieke et al. 2008), resulting in D₀ = 1330.2 km and D₀ = 1342.6 km, respectively. An earlier (commonly known and often used) value is D₀ = 1329 km. Applying Eq. (1), with D₀ = 1330.2 km and assuming a geometric albedo, p_V, of either 6.9% or 17% for 2003 UY₁₁₇ as proposed by Santos-Sanz et al. (2012) for SDOs, yields effective diameters of about 333 km and 212 km, respectively. Farkas-Takács et al. (2020) derived an effective diameter of $D={196}_{-54}^{+114}$ km for 2003 UY₁₁₇ from Herschel (PACS²) thermal observations (using an absolute magnitude of H_V = 5.91 mag in their work).

In this paper, we report the observation of a stellar occultation by 2003 UY₁₁₇ and the results we obtained from it (Sect. 2). We also obtained photometric observations in order to derive the absolute magnitude and the rotation period of 2003 UY₁₁₇ (Sect. 3). Finally, we combined these results in order to constrain the three-dimensional size of the body and to derive the geometric albedo (Sect. 4).

2 2020 October 23 occultation

2.1 Prediction

Within the Lucky Star collaboration³, we predicted a stellar occultation of a G = 14.5 mag star for 2020 October 23 using the Gaia DR2 star catalog and the NIMA⁴ ephemeris (Desmars et al. 2015). Table 1 summarizes the occultation parameter and the details of the occulted star. The prediction details in Table 1 were taken from the nominal NIMA (version 3) prediction⁵. The target star data from Gaia DR3 are also given for comparison. About a week before the occultation date, we updated and refined the prediction using high-precision astrometry (σ ∼ 15 mas) that we obtained with the 2 m Liverpool Telescope (LT) at Roque de Los Muchachos Observatory (ORM) on the island of La Palma, Spain. The update shifted the ground track farther to the north into a region with even better observability potential, especially for the European region, with a dense network of telescopes and observers (Fig. 1). We then organized an observation campaign to detect the occultation from as many sites as possible. We used the Occultation Portal (Kilic et al. 2022)⁷ for observation reporting and data storage.

2.2 Observations

The weather conditions were unfavorable during the event for large parts of the occultation path. However, we obtained four positive detections from three different locations in Spain. Additionally, we recorded three very close misses to the south of the body from Spain and ten misses from another nine observing sites (see Fig. 1 and Table A.1 for observation details). We utilized synthetic aperture photometry to obtain the occultation light curves from the observations. The four positive detections are shown in Fig. 2.

The star’s apparent diameter was calculated to be 0.0178 mas (V -mag) and 0.0172 mas (B-mag) using the formulae published by Kervella et al. (2004). This translates to a distance of 0.4 km at the projected distance of 2003 UY₁₁₇ (∆ = 33.47 au), or 0.02 s for the shadow velocity of 21.80km/s. The Fresnel scale $R_F=\sqrt{\lambda \cdot \text{Δ}/2}$ is 1.32 km, or 0.06 s for a wavelength of λ = 700 nm. As all positive detections were recorded with exposure times ≥2 s, any effects due to diffraction or the apparent stellar diameter are negligible.

The ingress (disappearance) and egress (reappearance) times were extracted from the fitted occultation light curves and were translated into chords on the sky plane. To model the light curves and to fit the profile, we utilized the SORA⁸ Python package (Gomes-Júnior et al. 2022), which also facilitates the extraction of ingress and egress times. The extracted times are listed in Table 2.

Fig. 1 Map of the ground track of our latest prediction update, based on astrometry obtained at the 2 m LT on La Palma. A spherical diameter of D = 285 km was used for 2003 UY117 for the predicted shadow path width (blue lines) plotted in this figure. The map also displays the sites where the occultation was observed, with blue markers indicating a positive detection and red ones indicating a negative detection (i.e., a “miss”). Negative observations reported from Belgium and England, which were located within the uncertainty of the original nominal prediction, are outside this map. Map credit: OpenStreetMap.

Fig. 2 Occultation light curves. The light curves (flux vs. time) are normalized and shifted from each other on the y-axis by an offset value of 2 for clarity. The site, telescope, and instrument details are given in Table A.1.

2.3 Profile fit

Assuming a spheroidal or an ellipsoidal object, the projected cross section on the sky plane is an ellipse. Therefore, we fit an ellipse to the extremities of the chords (derived from the ingress and egress times as described in Sect. 2.2), taking into account the near misses of the occultation as an additional constraint (Fig. 3). The five solve-for parameters were: the center of the ellipse ( f, g) with respect to the center of the fundamental plane defined by the geocentric star and TNO position for the event time; the semimajor axis, a′; the oblateness, є′ = (a′ − b′)/a′; and the position angle of the ellipse, φ′⁹. The prime (′) indicates that these parameters belong to the projected (“apparent”) profile ellipse of the object and distinguishes them from the axes of a physical body (triaxial ellipsoid with semiaxes a, b, and c). The parameters were estimated using the Levenberg- Marquardt optimization algorithm. The goodness of the fit was evaluated from the χ² per degree of freedom (pdf) value, defined as ${\chi }_{\text{pdf}}^2={\chi }^2/(N-M)$, where N = 8 is the number of data points and M = 5 is the number of adjustable parameters. Ideally, this value should be close to one. We obtain ${\chi }_{\text{pdf}}^2=0.43$ for our fit. The 1σ uncertainties in the retrieved parameters were obtained from a grid search in the parameter space, by varying one parameter from its nominal solution value while keeping the other parameters constant. Acceptable values were those that gave a χ² between ${\chi }_{\min }^2\text{ and }{\chi }_{\min }^2+1$. The results of our instantaneous best-fitting limb are summarized in Table 3.

3 Photometry

To interpret the occultation results with respect to the threedimensional shape and size of the physical body, we carried out photometric observations of 2003 UY₁₁₇ to determine its rotational light curve. We carried out observations with the 1.5 m telescope at Sierra Nevada Observatory (Spain) and with the 1.23 m telescope at Calar Alto Observatory (Spain) over six nights and with longer time coverage than the observations with the 2m LT on La Palma (Sect. 2.1); the latter were done with the IO:O instrument and a Sloan r′ filter, and were focused on astrometry with the aim of updating the occultation prediction. We also used sparse observations made at Calar Alto in 2019 by our group. Observations at the 1.5 m telescope were made with an Andor iKon-L CCD camera (model DZ936N-BEX2-DD)¹⁰ without filters in order to increase the signal-to-noise ratio. At the 1.23 m telescope, we used the DLR-MKIII instrument¹¹ (also without filters), except for two nights when we took images with V and R filters to determine the V − R color of the TNO. Image calibration and photometry were performed using the same algorithms and procedures as for the LT images (Sect. 2.1). The science images were calibrated in the usual manner, namely, bias and flat field image correction were applied to them.

Fig. 3 Elliptical fit to the 2020 October 23 occultation observations (chords). This fit describes the limb of 2003 UY117 for the moment of the occultation on the sky plane, defined by the (f, g) axes. As two chords were derived from the same site (OSN90 and OSN150; Sierra Nevada Observatory), they are not distinguishable in the plot. Also shown is the chord for the nearest site to the south (CAHA220: Calar Alto Observatory, 2.2m telescope) and to the north (Cannet), both of which had a negative detection (“miss”). The gray shaded area is the 1σ uncertainty region of the derived ellipse.

3.1 Absolute magnitude

From the calibrated CCD images, we derived magnitudes in the R band using our algorithms that use the Gaia DR2 field stars to determine photometric transformation equations, which take the color information into account (Morales et al. 2022). For the color of the TNO, we used a value of V − R = 0.46 ± 0.07 mag, which was derived from two nights of observations with the Calar Alto 1.23 m telescope. From a linear regression of the reduced R magnitude (the apparent magnitude that the TNO would have at 1 au from the Sun and Earth) at several phase angles, α, we obtained the absolute magnitude, H_R = m_R(1,1,0), and the phase slope, β (Fig. 4). From the fitted trend line, we derived an absolute magnitude of H_R = 5.45 ± 0.01 mag and a slope parameter of β = 0.22 ± 0.01 mag/°. Our observations covered the phase angle range α = [0.118°, 1.638°]. The scatter around the trend line indicates a significant rotational modulation.

Fig. 4 Reduced magnitude mR (1, 1, α) vs. the phase angle, α. In total, 431 observations obtained with the 2 m LT, with the 1.5 m telescope at Sierra Nevada Observatory, and with the 1.23 m telescope at Calar Alto Observatory were used. The brightness variation cluster is due to the rotation of 2003 UY117.

3.2 Rotational light curve

After de-trending¹² the reduced and light-travel-time-corrected photometry, we performed a search for the rotation period of 2003 UY₁₁₇ using different period-finding techniques. In total, we had more than 400 observations from 2019 November 30 to 2021 February 15. We used the Lomb-Scargle (LS; Lomb 1976; Scargle 1982) algorithm to estimate the most likely rotation period from our data. The algorithm works in the frequency domain, and the most prominent light curve frequencies (in units of 1/day as the timescale of the data is days) are shown in the LS periodogram (Fig. 5). The normalized spectral power reveals a frequency of about 4/day (the exact value is f = 3.878329) as the dominant frequency. Given that a small (ellipsoidal) body typically executes two light curve periods in a single rotation period, the most likely rotation frequency is f = 3.878329/2 (day⁻¹), which corresponds to a rotation period of P = 12.3765 h.

To further constrain and verify the rotation period, we folded the data with a period in the range 4–30 h in 0.00001 day steps. For each period value, we fit a second-order Fourier series to the folded data and calculated the root mean square error (RMSE; Fig. 6). Additionally, we verified the phased plots for the most prominent periods that were derived in both approaches, and we also evaluated split-halves plots. The period with the smallest RMSE is P = 12.3763 h, which we chose as our best estimate for the rotation period of 2003 UY₁₁₇. The second prominent minimum at 16.692 h (Fig. 6) corresponds to 2.8756 cycles/day, which is a 24h alias of the 1.9392 cycles/day frequency that corresponds to our preferred period of 12.376 h. The periodogram in cycles/day (Fig. 5) in addition to the periodogram in hours helps to identify 24 h aliases of the main peak, which are usually separated by ~ 1 cycle/day. From the results of the two approaches, we conclude that the rotation period is P = 12.376 ± 0.0033 h¹³. From the best Fourier fit (Fig. 7), we derived a peak-to-valley amplitude of ∆m = 0.36 ± 0.13 mag.

Fig. 5 LS periodogram. The rotational light curve frequency is given in 1/day. These values are light curve frequencies (periods); since a small body typically executes two light curve periods in a single rotation period, the best rotation period we obtain from the LS analysis is P = 12.3765 h.

Fig. 6 Fourier fit periodogram. We scanned the period range 4–30 hours in 0.00001 h steps, folded the photometric data with the selected period, and fit a second-order Fourier series to the data. The best period we obtained, P = 12.3763 h, is the value for which the RMS of the residuals (data minus fit) reaches a minimum.

4 Results

4.1 Size and shape

The instantaneous occultation ellipse limb fit to the projected profile of 2003 UY₁₁₇ (Fig. 3) yields a′ = 141 ± 9 km, b′ = 94 ± 16 km, and φ′ = −31^o ± 11^o with a ${\chi }_{\text{pdf}}^2=0.43$ value (Table 3). Given the large amplitude of the rotational light curve, we can expect that the intensity variations are due to shape effects, and this implies that 2003 UY₁₁₇ is not a spheroid. We assumed that a triaxial ellipsoid with semiaxes a > b > c (spin-axis c) is a good approximation for the physical body. In the following, we deduce the possible shape and size of this ellipsoid from the occultation observation combined with our light curve results. Using the Maclaurin sequence equations (Chandrasekhar 1969), a hydrostatic equilibrium body rotating at ~12 h down to a density of ρ ~ 0.2 g/cm³ would have taken on a Maclaurin spheroid shape, but 2003 UY₁₁₇ clearly does not have an (oblate) spheroid shape. Valid triaxial ellipsoids in hydrostatic equilibrium (Jacobi solutions) are possible only for densities ranging from 0.254 g/cm³ to 0.333 g/cm³ given the rotation period of 12.38 h. These densities are too low to be realistic for bodies in this size range.

The orthogonal projection of a triaxial ellipsoid (axes a > b > c, rotating around c) for a given spin state, expressed by the aspect angle, ψ (i.e., the angle between the rotation axis, c, and the line of sight) and the rotational phase, ϕ, is (e.g., Magnusson 1986) $$A=b^2{c}^2{\sin }^2\psi {\sin }^2\varphi +a^2{c}^2{\sin }^2\psi {\cos }^2\varphi +a^2{b}^2{\cos }^2\psi ,$$(2) $$\begin{array}{l}-B=a^2\left({\cos }^2\psi {\sin }^2\varphi +{\cos }^2\varphi \right)+b^2\left({\cos }^2\psi {\cos }^2\varphi +{\sin }^2\varphi \right)\hfill \\ +c^2{\sin }^2\psi ,\hfill \end{array}$$(3) $$a^{\prime }={\left(\frac{2A}{-B-{\left(B^2-4A\right)}^{1/2}}\right)}^{1/2},$$(4) $$b^{\prime }={\left(\frac{2A}{-B+{\left(B^2-4A\right)}^{1/2}}\right)}^{1/2},$$(5)

where (a′, b′) are the projected semiaxes that correspond to the apparent semiaxes of the projected cross section of an ellipsoidal object during a stellar occultation. The rotational light curve amplitude for such an ellipsoid can be calculated as (e.g., Binzel et al. 1989, p. 426) $$\text{Δ}m=2.5\text{\hspace{0.17em}log}\left(\frac{a}{b}\right)-1.25\text{\hspace{0.17em}log}\left(\frac{a^2{\cos }^2(\psi )+c^2{\sin }^2(\psi )}{b^2{\cos }^2(\psi )+c^2{\sin }^2(\psi )}\right).$$(6)

By performing a grid search for the three body semiaxes, a, b, and c, and the polar aspect angle, ψ, using the rotational phase angle, ϕ, for the observed occultation time, we can find the best fit to the projected shape derived from the occultation while simultaneously fitting the rotational light curve amplitude (∆m). The rotational phase at the time of the observed 2020 October 23 occultation was ~0.63, measured from the absolute maximum of brightness (Fig. 7). The observed peak-to-valley amplitude is ∆m = 0.36 ± 0.13 mag. We defined the cost function to be minimized as χ² = (0.36 − ∆m_c)²/0.13² + (1.5 − a′/b′)²/0.35² + (141 − a′)²/9², with the modeled light-curve amplitude, ∆m_c, derived by Eq. (6), and the apparent semiaxes a′, b′ as obtained from Eqs. (4)–(5) for each triaxial ellipsoid “clone” (defined by a, b, c, and ψ; ϕ = 0.63 ⋅ 2π) created during the grid search. The scanned parameter space was c = [60,120] km, b = [c, 160] km, and a = [b, 200] km, with a grid spacing of 2 km. The aspect angle, ψ, was scanned between 0 and 90 degrees in 1° steps. From this search, we obtained a family of possible solutions of triaxial ellipsoids and aspect angles. The model that minimizes χ² has axes a = 166 ± 12 km, b = 108 ± 12 km, and $c={90}_{-12}^{+14} \text{km}$, with an aspect angle of $\psi ={70}_{-12}^{+20}\text{\hspace{0.17em}deg}$. The diameter (of an equal-volume sphere) for this solution is D_eq = 235 ± 25 km. The 1σ uncertainties in the retrieved parameters were obtained by varying one parameter from its nominal solution value with corresponding ${\chi }^2={\chi }_{\min }^2\text{ up to }{\chi }^2={\chi }_{\min }^2+1$, while keeping the other parameters constant.

The spin axis orientation for this object is currently unknown. But if the pole orientation (α_p , δ_p) is known, the aspect angle can be computed as $$\cos \psi =\sin \delta \sin {\delta }_p+\cos \delta \cos {\delta }_0\cos \left(\alpha -{\alpha }_p\right),$$(7)

where α, δ are the coordinates of the object for the date of interest (e.g., the occultation date). By means of a grid search over the whole parameter space in 1° steps, we obtained a pole orientation of (α_p, δ_p) = (337° ± 10°, 62° ± 5°).

Fig. 7 Phased rotational light curve for 2003 UY117 using all the photometric data folded with a period of P = 12.376 h (upper panel) and P = 16.692 h (lower panel). The P = 12.376 h is our preferred solution. This double-peaked light curve with an amplitude of 0.36 mag indicates a highly nonspherical body with a presumable triaxial shape. Phase 0.0 corresponds to the moment of mid-occultation for the observed occultation chords.

4.2 Absolute magnitude and albedo

From our photometric observations, we obtained an absolute magnitude of H_R = 5.45 ± 0.01 mag and a phase slope coefficient of β = 0.22 ± 0.01 mag/° for 2003 UY₁₁₇ (Fig. 4). Taking into account the brightness contribution due to the rotational phase at the occultation time (about 0.06 mag), and using a color value V − R = 0.46 ± 0.07 mag, which we got from our observations as well, this yields H_R = 5.51 ± 0.01 mag and H_V = 5.97 ± 0.07 mag. Our V − R value is slightly smaller than the V − R value of 0.56 ± 0.01 mag reported by Sheppard (2012) and the V − R = 0.59 ± 0.01 mag reported by Tegler et al. (2016). Sheppard and Tegler derived absolute magnitudes of H_R = 5.35 ± 0.03 mag and H_V = 5.91 mag, respectively. Taking into account the large light curve amplitude (which was not considered in the estimates derived by these authors), their values are compatible with ours within the error bars. Our result also agrees with the values of H_R = 5.60 ± 0.04 mag, H_V = 6.13 ± 0.04 mag, and V − R = 0.53 ± 0.06 mag obtained by Alvarez-Candal et al. (2019), where a half rotational light curve amplitude value of ∆m/2 = 0.06 mag was also considered in the parameter estimation. By applying Eq. (1) with the values D₀ = 1330.2 km, D_eq,A = 228 ± 21 km as area-equivalent diameter, and H_V = 5.97 ± 0.07 mag, we get a geometric albedo of p_V = 0.139 ± 0.027 for 2003 UY₁₁₇.

4.3 Astrometry

The ellipse center coordinates given in Table 3, (119 ± 7, −10 ± 8) km, are two of the five solve-for parameters of the ellipse fit and represent the offset (O-C) between the observed and the predicted position (defined by the object ephemeris and the star position). This information was used to calculate the object position. We derived an astrometric place (ICRS¹⁴) for 2003 UY₁₁₇ at 22:19:21.8 UT on 2020 October 23 with equatorial coordinates $$\begin{array}{c}\alpha (\text{hms})=\mathrm{03\; 23\; 26.21404}\pm 0.3\text{\hspace{0.17em}mas},\\ \delta (\text{dms})=+\mathrm{22\; 47\; 19.3184}\pm 0.4\text{\hspace{0.17em}mas}.\end{array}$$

This high-precision astrometry will be used in our orbit determination of the TNO and will also improve the accuracy of future occultation predictions.

5 Conclusions

• A stellar occultation by TNO (143707) 2003 UY₁₁₇ has been predicted and successfully observed for the first time. From four occultation chords observed at three different sites, we derived an instantaneous projected elliptical size of the object with dimensions (282 ± 18) × (184 ± 32) km. The area-equivalent diameter is D_eq,A = 228 ± 21 km.

• We also obtained the absolute magnitude, phase slope, V − R color, and rotation period for this TNO from our photometric observation campaign. Our preferred rotation period, P, is 12.376 ± 0.0033 hours. The light curve is double-peaked with a peak-to-valley amplitude of ∆m = 0.36 ± 0.13 mag. We obtained an absolute magnitude of H_R = 5.51 ± 0.01 mag and H_V = 5.97 ± 0.07 mag using a V − R = 0.46 ± 0.07 mag value, which was also derived from our observations. From the area-equivalent diameter of D_eq,A = 228 ± 21 km and the absolute magnitude, H_V, given above, we derived a geometric albedo of p_V = 0.139 ± 0.027.

• By combining the occultation with the rotation light curve results, we derived tight constraints on the three-dimensional size and shape of 2003 UY₁₁₇. Our best solution is 2a = 332 ± 24 km, 2b = 216 ± 24 km, and $2c={180}_{-24}^{+28}\text{\hspace{0.17em}km}$ for a triaxial body, which yields an equivalent spherical diameter of D_eq = 235 ± 25 km. This value is slightly larger than the radiometric result $D_{\text{eq}}={196}_{-54}^{+114}\text{\hspace{0.17em}km}$ (Farkas-Takács et al. 2020), but well within the error margins. The aspect angle we derived for the occultation epoch is $\psi ={70}_{-12}^{+20}\text{\hspace{0.17em}deg}$. A pole solution that is compatible with the findings above is (α_p, δ_p) = (337° ± 10°,62° ± 5°).

• We derived an occultation-based astrometric position (ICRS) for 2003 UY₁₁₇.

Acknowledgements

Part of this work was supported by the Spanish projects PID2020-112789GB-I00 from AEI and Proyecto de Excelencia de la Junta de Andalucía PY20-01309. Authors J.L.O., P.S-.S., N.M., M.V.-L., and R.D. acknowledge financial support from the Severo Ochoa grant CEX2021-001131-S funded by MCIN/AEI/ 10.13039/501100011033. PS-S also acknowledges financial support from the Spanish I+D+i project PID2022-139555NB-I00 (TNO- JWST) funded by MCIN/AEI/10.13039/501100011033. A.A-.C. acknowledges financial support from the Severo Ochoa grant CEX2021-001131-S funded by MCIN/AEI/10.13039/501100011033. J.I.B.C. acknowledges grants 305917/20196, 306691/2022-1 (CNPq) and 201.681/2019 (Rio de Janeiro State Research Support Foundation, FAPERJ). This work is partly based on observations collected at the Centro Astronómico Hispano en Andalucía (CAHA) at Calar Alto, operated jointly by Junta de Andalucía and Consejo Superior de Investigaciones Científicas (CSIC). This research is also partially based on observations carried out at the Observatorio de Sierra Nevada (OSN) operated by Instituto de Astrofísica de Andalucía (IAA-CSIC) and on observations made with the Liverpool Telescope (LT) at the Roque de los Muchachos Observatory (ORM) on the island of La Palma as part of the Observatorios de Canarias (OCAN) operated by the Instituto de Astrofísica de Canarias (IAC). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001 and National Institute of Science and Technology of the e-Universe project (INCT do e-Universo, CNPq grant 465376/2014-2). FBR acknowledges CNPq grant 316604/2023-2. ^˙İST60 and ^˙İST40 are the observational facilities of the Istanbul University Observatory which were funded by the Scientific Research Projects Coordination Unit of Istanbul University with project numbers BAP-3685 and FBG-2017-23943 and the Presidency of Strategy and Budget of Republic of Turkey with the project 2016K12137. We acknowledge all observers which might have tried to observe this occultation event and which are not mentioned explicitly in Table A.1.This work has made use of data from the European Space Agency (ESA) mission Gaia ¹⁵, processed by the Gaia Data Processing and Analysis Consortium (DPAC)¹⁶. Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement. We thank the referee and editor for their valuable comments, which helped us to improve the quality and readability of the manuscript. This research has made use of the CORA web portal¹⁷.

Appendix A Additional table

References

All Tables

All Figures

Fig. 1 Map of the ground track of our latest prediction update, based on astrometry obtained at the 2 m LT on La Palma. A spherical diameter of D = 285 km was used for 2003 UY117 for the predicted shadow path width (blue lines) plotted in this figure. The map also displays the sites where the occultation was observed, with blue markers indicating a positive detection and red ones indicating a negative detection (i.e., a “miss”). Negative observations reported from Belgium and England, which were located within the uncertainty of the original nominal prediction, are outside this map. Map credit: OpenStreetMap.

In the text

	Fig. 2 Occultation light curves. The light curves (flux vs. time) are normalized and shifted from each other on the y-axis by an offset value of 2 for clarity. The site, telescope, and instrument details are given in Table A.1.
In the text

Fig. 3 Elliptical fit to the 2020 October 23 occultation observations (chords). This fit describes the limb of 2003 UY117 for the moment of the occultation on the sky plane, defined by the (f, g) axes. As two chords were derived from the same site (OSN90 and OSN150; Sierra Nevada Observatory), they are not distinguishable in the plot. Also shown is the chord for the nearest site to the south (CAHA220: Calar Alto Observatory, 2.2m telescope) and to the north (Cannet), both of which had a negative detection (“miss”). The gray shaded area is the 1σ uncertainty region of the derived ellipse.

In the text

	Fig. 4 Reduced magnitude mR (1, 1, α) vs. the phase angle, α. In total, 431 observations obtained with the 2 m LT, with the 1.5 m telescope at Sierra Nevada Observatory, and with the 1.23 m telescope at Calar Alto Observatory were used. The brightness variation cluster is due to the rotation of 2003 UY117.
In the text

	Fig. 5 LS periodogram. The rotational light curve frequency is given in 1/day. These values are light curve frequencies (periods); since a small body typically executes two light curve periods in a single rotation period, the best rotation period we obtain from the LS analysis is P = 12.3765 h.
In the text

	Fig. 6 Fourier fit periodogram. We scanned the period range 4–30 hours in 0.00001 h steps, folded the photometric data with the selected period, and fit a second-order Fourier series to the data. The best period we obtained, P = 12.3763 h, is the value for which the RMS of the residuals (data minus fit) reaches a minimum.
In the text

Fig. 7 Phased rotational light curve for 2003 UY117 using all the photometric data folded with a period of P = 12.376 h (upper panel) and P = 16.692 h (lower panel). The P = 12.376 h is our preferred solution. This double-peaked light curve with an amplitude of 0.36 mag indicates a highly nonspherical body with a presumable triaxial shape. Phase 0.0 corresponds to the moment of mid-occultation for the observed occultation chords.

In the text