Irregular flashes from tumbling asteroids reveal their surface compositions

Sparse survey snapshots can uncover the spin rates and chemical makeups of rocky minor planets across the Solar System before full decadal mapping begins.

Across the wide expanse between Mars and Jupiter, millions of rocky fragments tumble continuously through interplanetary space, reflecting shifting amounts of sunlight toward Earth. Because these ancient bodies are rarely spherical, their rotating cross-sections alternately expand and shrink relative to the sun. This geometric cycle causes their brightness to rise and fall in a rhythmic pattern known as an astronomical light curve. Tracing these subtle optical pulses reveals how quickly an asteroid spins, how elongated its physical shape is, and what minerals cover its battered surface.

Measuring these physical traits allows astronomers to understand the violent history of the early Solar System. Asteroids represent the rocky debris left behind by planet formation billions of years ago, preserving vital records of primordial chemistry, mutual collisions, and orbital drift. Surface composition reveals the temperature conditions of the early solar nebula where the rocks condensed, while rotational velocities indicate whether an object is a solid monolith or a fragile heap of rubble held together only by weak gravity.

Extracting these properties requires a telescope to record several distinct stages of an asteroid's journey. First, sunlight illuminates the rotating rock from a specific angle, casting shadows across uneven craters and ridges. As the elongated body spins, an observer on Earth sees two distinct brightness peaks and two troughs during each complete rotation cycle. Meanwhile, surface minerals absorb and reflect particular wavelengths of light differently, creating a characteristic colour signature. When observers subtract the rhythmic brightening and dimming caused purely by the asteroid's spinning geometry, the remaining colour differences isolate the true chemical composition of the surface.

In a preprint posted to arXiv on September 25, 2026, lead author Valerio Carruba of São Paulo State University and nine colleagues across Brazil, the United States, Chile, and Peru established an automated computational framework designed to resolve this difficulty using initial Science Validation survey data from the Vera C. Rubin Observatory. The authors showed that coupling high-order Fourier series fitting with multi-band Lomb-Scargle period analysis allows researchers to extract rotation periods, light-curve amplitudes, and taxonomic classifications from uneven and sparse observations.¹

How do wide-field sky surveys detect rotating asteroids?

Large-scale sky surveys encounter moving asteroids at unpredictable intervals rather than tracking individual targets continuously over hours of unbroken observation. As the Vera C. Rubin Observatory surveys the southern sky using its wide-field optical camera, moving asteroids cross the focal plane during routine sky scans, generating fleeting snapshots separated by hours, days, or weeks. The authors noted that this observational cadence creates irregular sampling, where an asteroid may be imaged only a handful of times in different colour filters under fluctuating viewing geometries.¹

These operational conditions contrast sharply with early demonstration datasets. An earlier study led by Sarah Greenstreet and published in Astrophysical Journal Letters on January 10, 2026, analysed Rubin First Look data collected over nine nights between April 21 and May 5, 2025.¹ That earlier analysis examined approximately 340,000 observations across roughly 2,000 objects, identifying reliable rotation periods for 75 main-belt asteroids and one near-Earth object with several hundred individual observations available per target.¹ The authors of the new preprint pointed out that later Science Validation observations obtained across 2025 and 2026 produced far more heterogeneous and fragmented coverage across the observed asteroid population.¹

Submission records submitted to the Minor Planet Center and archived by the Asteroid Institute illustrate this observational spread. In the February 2026 Minor Planet Center submission batch analysing observations from June through September 2025, the dataset included 697,934 observations of 119,916 known objects dedicated to astrometric validation, 99,221 observations of 5,872 candidate discoveries linked using heliocentric linking software, 25,452 observations of 648 new discoveries including near-Earth objects, and 9,516 observations of 636 known or candidate trans-Neptunian objects. However, Carruba and colleagues reported that among the thousands of bodies in the February 2026 release, only 30 main-belt objects and two near-Earth objects were observed more than 150 times.¹ An April 10, 2026 submission added 27,293 observations of 105 known main-belt asteroids, but only 57 main-belt objects within that set had at least 150 recorded measurements.

Addressing the empirical basis for this selection threshold, Carruba told Primary that "the value of 150 observations was used as a practical selection criterion for identifying the best-sampled objects, rather than as a universal threshold established by a formal statistical test."^contributed Carruba added that "our sparse-data experiments indicated that reliability depends strongly on temporal and phase coverage, particularly on having several well-separated observing clusters. Thus, an object with fewer than 150 observations can still yield a reliable period if those observations provide adequate phase coverage, while a larger number of poorly distributed observations does not guarantee a robust solution."^contributed

How does the new pipeline separate asteroid spin from surface colour?

Two complementary mathematical techniques allow researchers to identify true periodic asteroid rotations within unevenly spaced brightness measurements. Carruba and colleagues employed a dual-engine architecture combining high-order Fourier light-curve modeling with multi-band Lomb-Scargle period analysis.¹ By applying two distinct mathematical formulations to the same photometric time series, the researchers can confirm whether a detected periodic signal reflects a genuine physical rotation period rather than an artificial alias created by the telescope's regular observing schedule.

Before searching for rotational periodicity, the pipeline standardises raw apparent magnitudes through geometrical preprocessing. As an asteroid traverses interplanetary space, its apparent brightness shifts continuously due to changing distances from the sun and Earth as well as varying solar phase angles, which represent the angle formed between the sun, the asteroid, and the terrestrial observer. In the high-order Fourier analysis, the authors corrected observed magnitudes by subtracting five times the common logarithm of the product of the heliocentric distance and geocentric distance, while adjusting observation timestamps for light-travel time.¹ In the multi-band Lomb-Scargle implementation, the team derived continuous ephemerides from the Jet Propulsion Laboratory Horizons service, subtracting the predicted apparent visual magnitude from observed magnitudes before clipping statistical outliers beyond three standard deviations.

To model rotational modulation, the high-order Fourier technique fits a Fourier series ranging from second order to sixth order alongside a quadratic phase function and band-dependent absolute magnitude offsets. Carruba and colleagues explained that Fourier orders below two fail to reproduce the double-peaked morphology typical of rotating elongated asteroids, while orders above six introduce superfluous free parameters that cause severe overfitting on sparsely sampled datasets.¹ The pipeline evaluates candidate frequencies through weighted linear least squares, applying pairwise statistical variance-ratio tests known as F-tests to select the simplest model order that satisfactorily explains the photometric variance.

Explaining this numerical boundary, Carruba said in response to questions from Primary that "when we allowed higher Fourier orders, the main limitation was overfitting rather than a specific numerical instability."^contributed Carruba noted that "for sparsely and irregularly sampled light curves, additional Fourier terms can begin fitting individual measurements and gaps in the observing cadence rather than the underlying rotational signal. This produces artificial oscillations, extra maxima and minima, and sometimes unrealistically sharp light-curve features between observed points. Higher-order models can also make alias periods appear artificially competitive because their greater flexibility allows them to reproduce the sparse measurements. We therefore restricted the Fourier order to 2–6 as a compromise: sufficiently flexible to reproduce the non-sinusoidal shapes expected for asteroid light curves, but constrained enough to reduce spurious structure when the temporal coverage is limited."^contributed

Once a unified light-curve model identifies the best-fitting rotation period and peak-to-peak amplitude, the framework derives rotation-corrected colour indices by calculating differences between the fitted absolute magnitudes in individual filters. In an illustrative demonstration on the asteroid 2026 DO14, the team resolved a pronounced periodogram dispersion minimum at a frequency of 1.05 cycles per hour, establishing a well-constrained rotation period of 1.8993 hours that accurately folded multi-band measurements into a clean double-peaked light curve. The authors reported that incorporating the infrared z-band filter alongside standard green, red, and near-infrared g, r, and i filters substantially improves compositional discrimination, resolving diagnostic silicate absorption features that remain ambiguous in traditional three-filter colour surveys.¹

Explaining the physical mechanism behind the filter choice, Carruba told Primary that "the key advantage of the z band is that it extends Rubin photometry into the wavelength region where the broad silicate absorption feature near 1 μm begins to become important."^contributed Carruba explained that "asteroids with surfaces dominated by minerals such as olivine and pyroxene do not simply have different overall spectral slopes; the depth and shape of this absorption feature also vary with mineralogy. The g, r, and i filters mainly constrain the visible spectral slope, whereas adding z provides a measurement on the red side of the visible spectrum, close to this diagnostic silicate feature. Consequently, the i−z color can distinguish objects that may look rather similar at shorter wavelengths but have different behavior around the 1-μm band. This is especially useful for identifying basaltic, V-type-like surfaces and for reducing degeneracies among silicate-rich taxonomic classes. Broad-band photometry cannot replace spectroscopy, but the z band adds an important compositional constraint that is absent when only gri colors are available."^contributed

What are the limits of recovering properties from early survey scans?

Early Science Validation datasets do not provide sufficient temporal sampling to extract definitive rotation periods for every observed asteroid. Carruba and colleagues emphasised that while their methodology reliably extracts preliminary estimates of physical properties from sparse and irregular coverage, robust and unambiguous rotation period determinations require sustained temporal sampling over multiple complete rotational cycles.¹ When an asteroid is detected only a dozen times across weeks of survey operations, large gaps between successive data points create severe mathematical aliasing, where multiple unrelated rotational frequencies appear to fit the sparse observations equally well.

Clarifying when candidate periodicities fail, Carruba told Primary that the team did not define a universal sampling gap above which a period becomes an alias.^contributed "The more relevant quantity is phase coverage," Carruba said.^contributed "In our sparse-data tests, period recovery became unreliable when the observations were concentrated into too few temporal clusters or repeatedly sampled only a limited fraction of the rotational phase. In particular, the high-order Fourier method generally required at least three well-separated observing clusters to recover a stable solution. A candidate period would therefore be treated with caution if folding the data at that period left large portions of the rotational cycle unsampled, if individual clusters covered only nearly identical phases, or if small changes in the sampling caused the preferred period to jump to an alias. We also required consistency between the Fourier and multi-band Lomb–Scargle solutions. Thus, we rejected periods based on incomplete and unstable phase coverage rather than imposing a single maximum allowable time gap."^contributed

Simulated tests on artificially thinned data showed that high-order Fourier modeling is especially vulnerable to sparse sampling, requiring a denser concentration of observation epochs to converge reliably on the correct Fourier harmonic.¹ In contrast, the multi-band Lomb-Scargle approach proved significantly more resilient when dealing with fragmented time series, though it remains prone to period-doubling ambiguities when asymmetric light curves exhibit unequal peaks. Detailing this failure mode, Carruba told Primary that "in our artificially thinned datasets, the high-order Fourier solutions became unreliable when the observations were reduced to fewer than about three well-separated observing clusters or when those clusters sampled too little of the rotational phase. At that point, the additional Fourier degrees of freedom could fit the sparse measurements in several different ways, producing unstable periods and artificial light-curve structure."^contributed Carruba noted that the Lomb-Scargle method proved more tolerant of this reduced sampling, justifying the paired implementation.

The authors also noted that their ephemeris corrections do not explicitly model changing viewing aspect angles over extended observation baselines, which can gradually distort light-curve amplitudes over multi-month timescales.¹ Carruba told Primary that rather than enforcing a fixed calendar limit, the practical rejection criterion is evidence for epoch-dependent light-curve morphology.^contributed When an asteroid moves along its orbit, shifts in viewing geometry alter apparent amplitudes and light-curve shapes, meaning widely separated observations cannot be represented by a single static periodic Fourier model.

The study was posted as a preprint that may not have been peer reviewed on the arXiv server on September 25, 2026, and its validation remains confined to a small subset of well-observed asteroids available in early commissioning batches. The authors stated that the small number of objects analysed in their study reflects an intrinsic characteristic of early commissioning data rather than a systemic failure of the analytical framework. However, they acknowledged that demonstrating the performance and automated reliability of the pipeline across millions of minor bodies will require future survey-scale data releases.¹

What do these automated pipelines mean for future Solar System discoveries?

Establishing reliable automated algorithms for sparse light curves ensures that upcoming survey operations can rapidly transform raw transient alerts into detailed planetary catalogs. The Vera C. Rubin Observatory's Legacy Survey of Space and Time is scheduled to monitor the southern sky systematically for a decade, delivering multi-band time-series photometry for millions of previously uncatalogued Solar System bodies. Without automated frameworks capable of disentangling rotational brightness variations from true surface colours, planetary scientists would struggle to classify the mineral compositions of newly discovered objects.

By confirming that reliable rotational periods and taxonomies can be extracted using combinations of high-order Fourier analysis, Lomb-Scargle fitting, and four-band griz photometry, the framework developed by Carruba and colleagues provides an operational baseline for future data releases.¹ As the public alert stream expands and commissioning surveys transition into full ten-year baseline operations, these automated techniques will allow researchers to identify fast-spinning near-Earth asteroids, map compositional gradients across the main asteroid belt, and reconstruct the ancient fragmentation events that shaped our planetary system.

The record

39 sentences trace to 1 source and 1 contributor.

All 2Documents 1Records 0Reporting 0People 1

1. 1 VC V. Carruba, S. Greenstreet, M. Mastropietro, S. Aljbaae, G. Caritá, R. C. Domingos, M. Huaman, M. M. Bala, R. D. Z. Pedroso, E. M. D. S. Delfino Paper · arxiv.org · 25 Sep 2026 Extracting Asteroid Physical Properties from Vera C. Rubin Observatory Science Validation Survey Data: Rotational Periods, Colors, and Taxonomies Preprint · may not have been peer reviewed See the source
2. 2 VC Valerio Carruba Contributor · interview 29 Sep 2026 · in the article since 9 Oct 2026 Contribution — Valerio Carruba 17 statements added to this article

Also consulted: 1 page across 1 site, none cited.

Article history

1. Published 9 Oct 2026, 18:18
   Assembled by the Primary desk from 1 source · 1 contributor · 39 cited sentences
2. VC Valerio Carruba, lead author 17 statements added 29 Sep 2026 11 days before publication

   1. Go to the sentence: However, Carruba and colleagues reported that among the thousands of bodies in the February 2026 release, only 30 main-belt objects and two near-Earth objects were observed more than 150 times.↑
   2. Go to the sentence: Addressing the empirical basis for this selection threshold, Carruba told Primary that "the value of 150 observations was used as a practical selection criterion for identifying the best-sampled objects, rather than as a universal threshold established by a formal statistical test."↑
   3. Go to the sentence: Carruba added that "our sparse-data experiments indicated that reliability depends strongly on temporal and phase coverage, particularly on having several well-separated observing clusters. Thus, an object with fewer than 150 observations can still yield a reliable period if those observations provide adequate phase coverage, while a larger number of poorly distributed observations does not guarantee a robust solution."↑
   4. Go to the sentence: Carruba and colleagues explained that Fourier orders below two fail to reproduce the double-peaked morphology typical of rotating elongated asteroids, while orders above six introduce superfluous free parameters that cause severe overfitting on sparsely sampled datasets.↑
   5. Go to the sentence: Explaining this numerical boundary, Carruba said in response to questions from Primary that "when we allowed higher Fourier orders, the main limitation was overfitting rather than a specific numerical instability."↑
   6. Go to the sentence: Carruba noted that "for sparsely and irregularly sampled light curves, additional Fourier terms can begin fitting individual measurements and gaps in the observing cadence rather than the underlying rotational signal. This produces artificial oscillations, extra maxima and minima, and sometimes unrealistically sharp light-curve features between observed points. Higher-order models can also make alias periods appear artificially competitive because their greater flexibility allows them to reproduce the sparse measurements. We therefore restricted the Fourier order to 2–6 as a compromise: sufficiently flexible to reproduce the non-sinusoidal shapes expected for asteroid light curves, but constrained enough to reduce spurious structure when the temporal coverage is limited."↑
   7. Go to the sentence: The authors reported that incorporating the infrared z-band filter alongside standard green, red, and near-infrared g, r, and i filters substantially improves compositional discrimination, resolving diagnostic silicate absorption features that remain ambiguous in traditional three-filter colour surveys.↑
   8. Go to the sentence: Explaining the physical mechanism behind the filter choice, Carruba told Primary that "the key advantage of the z band is that it extends Rubin photometry into the wavelength region where the broad silicate absorption feature near 1 μm begins to become important."↑
   9. Go to the sentence: Carruba explained that "asteroids with surfaces dominated by minerals such as olivine and pyroxene do not simply have different overall spectral slopes; the depth and shape of this absorption feature also vary with mineralogy. The g, r, and i filters mainly constrain the visible spectral slope, whereas adding z provides a measurement on the red side of the visible spectrum, close to this diagnostic silicate feature. Consequently, the i−z color can distinguish objects that may look rather similar at shorter wavelengths but have different behavior around the 1-μm band. This is especially useful for identifying basaltic, V-type-like surfaces and for reducing degeneracies among silicate-rich taxonomic classes. Broad-band photometry cannot replace spectroscopy, but the z band adds an important compositional constraint that is absent when only gri colors are available."↑
   10. Go to the sentence: Carruba and colleagues emphasised that while their methodology reliably extracts preliminary estimates of physical properties from sparse and irregular coverage, robust and unambiguous rotation period determinations require sustained temporal sampling over multiple complete rotational cycles.↑
   11. Go to the sentence: Clarifying when candidate periodicities fail, Carruba told Primary that the team did not define a universal sampling gap above which a period becomes an alias.↑
   12. Go to the sentence: "The more relevant quantity is phase coverage," Carruba said.↑
   13. Go to the sentence: "In our sparse-data tests, period recovery became unreliable when the observations were concentrated into too few temporal clusters or repeatedly sampled only a limited fraction of the rotational phase. In particular, the high-order Fourier method generally required at least three well-separated observing clusters to recover a stable solution. A candidate period would therefore be treated with caution if folding the data at that period left large portions of the rotational cycle unsampled, if individual clusters covered only nearly identical phases, or if small changes in the sampling caused the preferred period to jump to an alias. We also required consistency between the Fourier and multi-band Lomb–Scargle solutions. Thus, we rejected periods based on incomplete and unstable phase coverage rather than imposing a single maximum allowable time gap."↑
   14. Go to the sentence: Simulated tests on artificially thinned data showed that high-order Fourier modeling is especially vulnerable to sparse sampling, requiring a denser concentration of observation epochs to converge reliably on the correct Fourier harmonic.↑
   15. Go to the sentence: Detailing this failure mode, Carruba told Primary that "in our artificially thinned datasets, the high-order Fourier solutions became unreliable when the observations were reduced to fewer than about three well-separated observing clusters or when those clusters sampled too little of the rotational phase. At that point, the additional Fourier degrees of freedom could fit the sparse measurements in several different ways, producing unstable periods and artificial light-curve structure."↑
   16. Go to the sentence: The authors also noted that their ephemeris corrections do not explicitly model changing viewing aspect angles over extended observation baselines, which can gradually distort light-curve amplitudes over multi-month timescales.↑
   17. Go to the sentence: Carruba told Primary that rather than enforcing a fixed calendar limit, the practical rejection criterion is evidence for epoch-dependent light-curve morphology.↑