100% found this document useful (1 vote)
10 views28 pages

Measurement of The Photosphere Oblateness of Cassiopeiae Via Stellar Intensity Interferometry With The VERITAS Observatory

The document presents the first measurement of the photosphere oblateness of the star γ Cassiopeiae using stellar intensity interferometry with the VERITAS Observatory. The study reports a minor-axis angular diameter of 0.43 ± 0.02 mas and a major-to-minor-radius ratio of 1.28 ± 0.04, consistent with previous spectroscopic measurements. This research demonstrates the effectiveness of intensity interferometry in measuring stellar shapes and sizes at optical wavelengths, providing new insights into rapidly rotating stars.

Uploaded by

Jing Ma
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
10 views28 pages

Measurement of The Photosphere Oblateness of Cassiopeiae Via Stellar Intensity Interferometry With The VERITAS Observatory

The document presents the first measurement of the photosphere oblateness of the star γ Cassiopeiae using stellar intensity interferometry with the VERITAS Observatory. The study reports a minor-axis angular diameter of 0.43 ± 0.02 mas and a major-to-minor-radius ratio of 1.28 ± 0.04, consistent with previous spectroscopic measurements. This research demonstrates the effectiveness of intensity interferometry in measuring stellar shapes and sizes at optical wavelengths, providing new insights into rapidly rotating stars.

Uploaded by

Jing Ma
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 28

Draft version June 19, 2025

Typeset using LATEX preprint style in AASTeX7.0.1

Measurement of the photosphere oblateness of γ Cassiopeiae via Stellar Intensity


Interferometry with the VERITAS Observatory
A. Archer ,1 J. P. Aufdenberg ,2 P. Bangale ,3 J. T. Bartkoske ,4 W. Benbow ,5
J. H. Buckley ,6 Y. Chen ,7 N. B. Y. Chin ,8 J. L. Christiansen ,9 A. J. Chromey,5
A. Duerr ,4 M. Escobar Godoy,10 S. Feldman ,7 Q. Feng ,4 S. Filbert ,4 L. Fortson ,11
A. Furniss ,10 W. Hanlon ,5 O. Hervet ,10 C. E. Hinrichs ,5, 12 J. Holder ,3 Z. Hughes,6
T. B. Humensky ,13, 14 W. Jin ,7 M. N. Johnson ,10 M. Kertzman,1 M. Kherlakian ,15
arXiv:2506.15027v1 [astro-ph.SR] 18 Jun 2025

D. Kieda ,8 N. Korzoun ,3 T. LeBohec ,8 M. A. Lisa ,16 M. Lundy ,17 G. Maier ,18


N. Matthews ,19, 20 P. Moriarty ,21 R. Mukherjee ,22 W. Ning ,7 R. A. Ong ,7
A. Pandey ,4 M. Pohl ,23, 18 E. Pueschel ,15 J. Quinn ,24 P. L. Rabinowitz ,6
K. Ragan ,17 P. T. Reynolds,25 D. Ribeiro ,11 E. Roache,5 J. G. Rose ,16 I. Sadeh ,18
L. Saha ,5 M. Santander ,26 J. Scott ,16 G. H. Sembroski,27 R. Shang ,22 D. Tak ,28
J. V. Tucci,29 J. Valverde ,30 V. V. Vassiliev,7 D. A. Williams 10 And S. L. Wong 17
The VERITAS Collaboration

1
Department of Physics and Astronomy, DePauw University, Greencastle, IN 46135-0037, USA
2
Embry-Riddle Aeronautical University, Physical Sciences Department, 1 Aerospace Blvd, Daytona Beach, FL 32114,
USA
3
Department of Physics and Astronomy and the Bartol Research Institute, University of Delaware, Newark, DE
19716, USA
4
Department of Physics and Astronomy, University of Utah, Salt Lake City, UT 84112, USA
5
Center for Astrophysics | Harvard & Smithsonian, Cambridge, MA 02138, USA
6
Department of Physics, Washington University, St. Louis, MO 63130, USA
7
Department of Physics and Astronomy, University of California, Los Angeles, CA 90095, USA
8
University of Utah, Department of Physics and Astronomy, 115 South 1400 East 201, Salt Lake City, Utah, USA
9
Physics Department, California Polytechnic State University, San Luis Obispo, CA 94307, USA
10
Santa Cruz Institute for Particle Physics and Department of Physics, University of California, Santa Cruz, CA
95064, USA
11
School of Physics and Astronomy, University of Minnesota, Minneapolis, MN 55455, USA
12
Department of Physics and Astronomy, Dartmouth College, 6127 Wilder Laboratory, Hanover, NH 03755 USA
13
Department of Physics, University of Maryland, College Park, MD, USA
14
NASA GSFC, Greenbelt, MD 20771, USA
15
Fakultät für Physik & Astronomie, Ruhr-Universität Bochum, D-44780 Bochum, Germany
16
The Ohio State University, Department of Physics, 191 W Woodruff Ave, Columbus, Ohio, USA
17
Physics Department, McGill University, Montreal, QC H3A 2T8, Canada
18
DESY, Platanenallee 6, 15738 Zeuthen, Germany
19
Space Dynamics Laboratory, Utah State University, Logan, UT, USA
20
Université Côte d’Azur, CNRS, Institut de Physique de Nice, France
21
School of Natural Sciences, University of Galway, University Road, Galway, H91 TK33, Ireland
22
Department of Physics and Astronomy, Barnard College, Columbia University, NY 10027, USA
23
Institute of Physics and Astronomy, University of Potsdam, 14476 Potsdam-Golm, Germany

Corresponding author: Josephine Rose


Email: josieg.rose.86@gmail.com
2
24
School of Physics, University College Dublin, Belfield, Dublin 4, Ireland
25
Department of Physical Sciences, Munster Technological University, Bishopstown, Cork, T12 P928, Ireland
26
Department of Physics and Astronomy, University of Alabama, Tuscaloosa, AL 35487, USA
27
Department of Physics and Astronomy, Purdue University, West Lafayette, IN 47907, USA
28
SNU Astronomy Research Center, Seoul National University, Seoul 08826, Republic of Korea.
29
Department of Physics, Indiana University Indianapolis, Indianapolis, Indiana 46202, USA
30
Department of Physics, University of Maryland, Baltimore County, Baltimore MD 21250, USA and NASA GSFC,
Greenbelt, MD 20771, USA

ABSTRACT
We use the stellar intensity interferometry system implemented with the Very Ener-
getic Radiation Imaging Telescope Array System (VERITAS) at Fred Lawrence Whip-
ple Observatory (FLWO) as a light collector to obtain measurements of the rapid rotator
star γ Cassiopeiae, at a wavelength of 416 nm. Using data from baselines sampling dif-
ferent position angles, we extract the size, oblateness, and projected orientation of the
photosphere. Fitting the data with a uniform ellipse model yields a minor-axis angular
diameter of 0.43 ± 0.02 mas, a major-to-minor-radius ratio of 1.28 ± 0.04, and a position
angle of 116◦ ± 5◦ for the axis of rotation. A rapidly-rotating stellar atmosphere model
that includes limb and gravity darkening describes the data well with a fitted angu-
lar diameter of 0.604+0.041 +0.8
−0.034 mas corresponding to an equatorial radius of 10.9−0.6 R⊙ ,
a rotational velocity with a 1 σ lower limit at 97.7% that of breakup velocity, and a
position angle of 114.7+6.4
−5.7 degrees. These parameters are consistent with Hα line spec-
troscopy and infrared-wavelength Michelson interferometric measurements of the star’s
decretion disk. This is the first measurement of an oblate photosphere using intensity
interferometry.

Keywords: Astronomy data modeling (1859) — High angular resolution (2167) — Long
baseline interferometry (932) — Nonclassical interferometry (1120) — Stel-
lar radii (1626) — Stellar rotation (1629)

1. INTRODUCTION
Spectroscopic measurements of the Hα line of the subgiant γ Cassiopeiae (γ Cas) indicate that it is
a Be-type star rotating at nearly its breakup velocity about an axis inclined by 60◦ to our line of sight.
Thus, it is expected to have a pronounced equatorial bulge. The history of Michelson interferometric
measurements of γ Cas spans many decades (see §3 of Stee, Ph. et al. 2012, for a brief overview).
These measurements reveal an extended decretion disk that suggests a 122◦ position angle of the
rotational axis, measured east of celestial north (Sigut et al. 2020). However, these measurements,
performed at infrared wavelengths, are unable to measure the photosphere itself.
Stellar intensity interferometry (SII) at optical wavelengths has the potential to measure the pho-
tosphere itself at sub-milliarcsecond scale. Originally developed in the 1950’s and used through the
early 1970’s (Hanbury Brown & Twiss 1954; Hanbury Brown 1974), SII has very recently seen a
renaissance with the availability of large light collectors in imaging atmospheric Cherenkov telescope
(IACT) arrays, high-speed electronics and modern computing at several observatories (Abeysekara
3

Figure 1. Baseline coverage of the six telescope pairs in the u − v plane for the data tabulated in Table B1.
Points reflect the weighted-mean position of each ∼hour-long run. The curves are point-reflected about the
origin.

et al. 2020; Abe et al. 2024; Zmija et al. 2023; Acharyya et al. 2024). By correlating the intensity
of light at separated telescopes, rather than the wave amplitude (as done with Michelson interfer-
ometry), SII is robust against effects of optical imperfections, including atmospheric turbulence, and
works well at optical wavelengths, providing sub-milliarcsecond resolutions with 100 m baselines.
Relative to Michelson interferometry, this comes at a cost of reduced signal-to-noise for the same
observing duration.
In this paper, we present the first SII measurement sensitive to the shape, as well as the size, of a
star. The work is organized as follows. Section 2 gives a brief description of the VERITAS stellar
intensity interferometer (VSII) and the datasets discussed here. Section 3 presents the data analysis
and extracted squared visibilities. Section 4 uses simple geometric models of the star to extract size,
shape, and orientation information from the measurements. The dependence of the squared visibility
on u and v is fitted with uniform disk and uniform ellipse models. Section 5 discusses a complete
stellar model of a rapid rotator with limb and gravitational darkening. Comparison of our data with
this model provides tight constraints on the size, rotational speed, and orientation of the star. In
Section 6, we summarize and provide an outlook for future work.

2. INSTRUMENT AND DATASET


VERITAS is composed of four 12-m diameter IACTs, usually used for very high-energy (E >
100 GeV) gamma-ray astronomy (Holder et al. 2006). During bright-moon periods, the system is used
for SII measurements. The four telescopes are individually referred to as T1, T2, T3, and T4, and
are unequally spaced with separations ranging between about 80 and 170 meters, and interferometric
projected baselines ranging between about 35 and 170 meters. We have presented further details of
the VSII stellar intensity interferometer in previous publications (Abeysekara et al. 2020; Acharyya
4
Table 1. VSII Observations of γ Cassiopeiae

UTC Date Telescope Pairs Used Observation Time Target - Moon Angle
(yyyy-mm-dd) (TA ,TB ) (hours) (degrees)
2023-01-08 (1,3), (2,3), (2,4), (3,4) 4.5 79
2023-02-02 (1,2), (1,3), (2,3), (2,4), (3,4) 3.25 62
2023-12-25 (2,4) 2.0 51
2023-12-26 (1,2), (1,3), (2,3), (2,4), (3,4) 7.28 56
2023-12-27 (1,2), (1,3), (2,3), (2,4), (3,4) 6.0 63
2024-02-19 (1,2), (1,3), (1,4), (2,3), (2,4), (3,4) 2.5 59
2024-02-20 (1,2), (1,3), (2,3), (2,4) 2.0 66
2024-02-21 (1,3), 2,3) 1.0 74
2024-02-22 (1,3), (2,3), (2,4), (3,4) 1.75 82
2024-05-18 (1,2), (2,3), (3,4) 1.0 MBH
2024-05-19 (1,2), (2,3), (3,4) 1.0 MBH
2024-12-13 (1,2), (1,3), (2,3), (2,4), (3,4) 2.0 46
2024-12-15 (1,2), (1,3), (2,3), (3,4) 1.0 55
2024-12-17 (1,3), (2,3), (2,4), (3,4) 1.0 70
Note—MBH denotes moon below horizon.
Dates and durations of VSII observations of γ Cas are listed, together with the Moon angle and the
telescope pairs used. The four VERITAS telescopes are identified by numbers from one to four. The Moon
angle is the angle between the Moon and the observed star. It correlates with the night sky background
level affecting the data. Only a subset of telescope pairs from each observation were used in the analysis
because some telescopes were not working, they had too large of a baseline, or the data was too noisy.

et al. 2024). The optical light from the observed star is detected by a single Hamamatsu super-
bialkali R10560 photomultiplier tube (PMT) located at the telescope focal plane, and continuously
digitized at a 250 MHz rate with 8 bit resolution. We use a narrow-band filter (λ = 416) with a
bandpass of 10 nm to focus on the photosphere of the star. An upgrade associated with the present
study is the use of a CPU-based server that simultaneously correlates all six telescope pairs with
a frequency-domain algorithm. This correlator produces correlations identical to the FPGA-based
correlator using a time-based algorithm that was used in our previous measurements (Scott 2023).
For some observations in the present paper, a 10-MHz clock improved the start-time synchronization.
Table 1 lists the γ Cas observations for which data were collected, over 6 bright moon periods
between January 2023 and December 2024. Intensity interferometry is generally robust against
ambient light for bright stars; VSII operations are reliable for target-to-moon angles from 30◦ to 95◦
(Kieda et al. 2021). Target-to-moon angles are listed in table 1.
Over 160 pair-hours of data were used, after quality cuts discussed below. The baseline coverage in
the star image Fourier reciprocal (u − v) plane is shown in figure 1. The relative telescope separation
lies in the u − v plane, the celestial east-north plane, following general interferometric convention.
5

(See § 4 of Thompson et al. (2001) and Dravins et al. (2013) for a further description of the u − v
plane.)
3. DATA ANALYSIS
3.1. Overview
The size and shape of the photosphere are determined by measuring the dependence of the intensity
correlation on the length and orientation of the projected baseline (i.e. separation between the
telescopes perpendicular to the direction of the starlight) and then fitting this data to a model.
In this section, we discuss the steps taken to measure the intensity correlation between pairs of
telescopes; they are similar to those we have presented previously (Abeysekara et al. 2020; Acharyya
et al. 2024).
The relative-time intensity correlation is measured as a function of time during a run as described
in section 3.2. After removing effects of night-sky background light (section 3.3) radio interference,
and tracking corrections (section 3.4), the magnitude of the correlation is quantified by a Gaussian
fit to the intensity correlation projected over a run, as discussed in section 3.5. Finally, in section 3.6
we identify some runs for which the noise is too large to find a correlation peak.
3.2. The correlogram
The current I(t) from the photomultiplier tube in each telescope is digitized and recorded at
250 MHz, i.e. in 4-ns samples. A correlogram is formed by convoluting the digitized waveforms from
two telescopes A and B and normalizing by the product of the average values of the currents:

∗ ⟨IA (t)IB (t − τ − τ0 )⟩t (T )


CAB (τ, T ) ≡ , (1)
⟨IA (t)⟩t (T ) × ⟨IB (t)⟩t (T )
where ⟨· · · ⟩t indicates an average over time and τ is the relative time (“time lag”) between samples.
The superscript (∗ ) denotes the correlation before correcting for effects of stray light, as discussed
in section 3.3. The constant τ0 is set equal to the optical path delay between the telescopes at
the beginning of the observation run. The normalized correlation is constructed every 2 seconds of
time-in-run T . We refer to these two-second intervals as “frames” below.
A correlogram for a 2-hour observation run with telescopes T2 and T4 is shown in figure 2. A clear
correlation signal is visible close to τ = 0 at the beginning of the run, due to proper adjustment
of τ0 (which we shall drop from further discussion for simplicity) in the correlator. The position of
the correlation in relative time changes over the course of the run, as the star moves across the sky,
which changes the optical path delay between the fixed telescope-pair positions.
3.3. Correction for residual light from other sources
During observation runs, the current in the phototubes is dominated by the light from the star.
Some of the light incident on the phototubes does not come from the star of interest, but rather
from the night sky background. This uncorrelated contribution leads to a reduction in the raw
measured correlation signal (Rou et al. 2013). Neglecting this contribution would generate a run-by-
run variation in the proportionality constant between the measured integral of the Gaussian fit of
the correlation peaks and the actual squared visibility.
We correct for this effect as we have done previously (Abeysekara et al. 2020; Acharyya et al. 2024).
Immediately before and after every observation run, the telescopes are pointed 0.5◦ north of the star
6

Figure 2. The correlation function as a function of relative time and time-in-run, for a typical 2 hour run.
While the correlogram is constructed every 2 seconds, the vertical axis has been binned more coarsely in
this figure, to reduce statistical noise and make the correlation more visible. The diagonal feature is the
correlation peak proportional to the squared visibility that changes during the run due to the variation in
the optical path delay between the two telescopes. This correlation function is from pair T2T4, taken at
03:37 on UTC 2023-12-26.

and the average currents are recorded. Depending on viewing conditions, stellar position, and moon
angle and phase for a given observation run, the off-star current is 1 − 15% of the on-star current.
The corrected correlogram C(τ, T ) is related to the raw correlogram C ∗ (τ, T ) (equation 1) as

∗ ⟨IA (t)⟩t (T ) ⟨IB (t)⟩t (T )


CAB (τ, T ) − 1 = (CAB (τ, t) − 1) · , (2)
(⟨IA (t)⟩t (T ) − ⟨IA∗ (t)⟩t (T )) (⟨IB (t)⟩t (T ) − ⟨IB∗ (t)⟩t (T ))

where ⟨IA∗ (t)⟩t (T ) represents time-averaged stray-light current at time-in-run T . We assume that
during the observations on the star itself, the ambient contribution varies linearly with time between
what we measure before and after the on-star run. Neglecting the stray light contamination would
lead to a ∼ 10% inaccuracy on the angular diameter estimate.

3.4. Removal of artifacts from tracking corrections and radio-frequency contamination


Some of the frames of the correlogram must be discarded before further processing, due to two
effects.
The first comes from telescope tracking corrections. Occasionally (typically once an hour), the
image of the star moves slightly off of the face of the photomultiplier tube. When this happens, the
operator slightly adjusts the tracking parameters. During this procedure, the telescope briefly stops
tracking altogether, and the current drops sharply, as starlight is no longer being registered. An
example is seen in figure 3.
The second problem arises from an on-site radio repeater at FLWO, which transmits a 50 W signal
at about 171 MHz in short (∼ 5 sec), sporadic (roughly every 10-30 seconds) bursts. This results
in a strong aliased 79.4 MHz sinusoidal correlation– more than 50 times larger than the physical
correlation of interest– in all telescope pairs (Scott 2023). In addition to producing the correlogram,
the correlator server calculates the frequency power spectrum of each telescope for each frame, and
the 79.4 MHz spikes are easily identified, as demonstrated in figure 4. Roughly one third of the frames
are contaminated by this signal and are removed from the correlogram. Starting in late December,
7

Figure 3. An example of a tracking correction in T2. A large negative ADC (analog-to-digital converter)
value corresponds to intense light incident on the PMT. Due to tracking limitations of the array, at about
200 sec into the run, the star’s image becomes slightly less centered on the PMT. At 400 sec into the run,
the operator adjusted the tracking, causing the telescope to briefly stop tracking the star altogether, and
the current drops sharply. When the tracking is corrected, the current is stronger. The frames around the
spike are removed from the correlogram before processing further.

Figure 4. A portion of the power spectrum for T1, as a function of frequency and time in run. Short,
sporadic bursts from the on-site radio repeater produce strong signals at 79.4 MHz (an alias of the 171 MHz
radio signal). Frames containing these bursts are removed from the correlogram before further processing.

2024, we received permission to disable the repeater during VSII running, resulting in cleaner data
and effectively about 40% greater data collection efficiency.
Finally, for many runs and telescopes, there are much smaller frequency spikes at 93.7, 99.6, and
107.5 MHz present in every frame. The presence and strength of these signals depends on telescope
orientation and vary from night to night. They are likely due to local radio stations. When observed,
the resulting sinusoidal correlations are subtracted from the correlogram, frame-by-frame, as discussed
in Acharyya et al. (2024).
8

Figure 5. The OPD-corrected projection of the correlation function as a function of relative time, defined
in equation 3. The correlation is quantified by the integral of a Gaussian peak fitted near relative time ≈ 0.
This correlation function is from pair T2T3, taken at 02:31 on UTC 2024-02-19. Before further analysis, the
uncertainty on the area of the Gaussian fit (A) is replaced as discussed in section 3.5.

3.5. Obtaining squared visibilities and their uncertainties


Once the bad frames are removed and radio-station signals subtracted, a one-dimensional correlation
function as a function of relative time is produced by shifting each frame by the appropriate optical
path delay (OP D) and averaging over frames, according to

g (2) (τ ) ≡ ⟨C(τ − OP D(T ), T )⟩T . (3)

An example is shown in figure 5. A clear signal is observed close to relative time τ = 0. Due to
uncertainties on the clock cycle and exact start time, the peak position may vary on order 10 ns,
from one run to another (Acharyya et al. 2024).
A Gaussian fit is performed on g (2) (τ ) − 1 within a window |τ | < 14 ns. An actual correlation
signal is expected to have a width around 4 ns, due to the finite sampling and response time of the
photomultiplier tube and preamp. If the peak-finder fails to find a peak in the window, or if its
Gaussian width is less than 2 ns or greater than 6 ns, then the data is discarded. Variation of the
|τ | window between |τ | < 10 ns and |τ | < 20 ns has a negligibly small effect on the measurements.
The peak width cut was varied between 1.5 and 7 ns, and the effect of these variations is included
in the systematic uncertainty estimate below. The integral of the Gaussian, measured in units of
time and denoted A in Figure 5, is proportional to the squared visibility (Acharyya et al. 2024). The
proportionality constant, determined by the bandwidth of the optical filter, electronic response, and
telescope optics, is expected (Matthews 2020) to be on order 10−6 ns. We treat this normalization
constant as a parameter in our fits below.
The statistical uncertainty of this integral could be derived from the fit parameters, but this ignores
the residual correlations present in the background. To arrive at a better estimate of the uncertainty
that accounts for the residual background, we randomly add false peaks, with the same area as the
fitted peak, on top of the background far from the signal region (|τ | > 50 ns). We then perform the
same Gaussian fitting and selection procedure on these false peaks as we do for the true peak. The
standard deviation of the distribution of false-peak Gaussian integrals is taken as the uncertainty
of the fitted peak near τ ≈ 0. The uncertainty extracted in this way is typically about 40% larger
9

Figure 6. Distribution of areas from fits to background noise in the region of relative time far from the
peak, for all runs. The gray band on fig. 7 corresponds to the red lines at ±3.58 fs, which are ±2σ from the
mean of this distribution. This measures the spread of areas that may be found when fitting to background
noise alone, and includes 95% of the data in the distribution.

than what would be obtained if we ignored the residual correlated background. This procedure is
illustrated in more detail in Appendix A.
We have reanalyzed the data of our previous measurement (Acharyya et al. 2024) of the slow
rotator β UMa, using our improved method of uncertainty estimations and removing tracking correc-
tions. The resulting angular diameter for a uniform-disk fit changes by only 1% (well within stated
uncertainties), but the χ2 /ndf is reduced from 2.30 to 1.04.

3.6. Exclusion of sub-threshold peaks


In the presence of statistical and residual correlated noise, any peak-finding algorithm finds spurious
peaks when the actual signal is too small. To identify the threshold at which this occurs, we used
the peak-finder in random regions on the background far from the actual peak region (|τ | > 50 ns),
again applying the same Gaussian fitting and selection criteria as for the actual signal peaks.
The resulting distribution of Gaussian integrals is shown in figure 6. As we have discussed previ-
ously (Acharyya et al. 2024), this results in a bimodal distribution with a strong dip at zero. We take
twice the standard deviation of this distribution as a threshold, which includes 95% of the fits (Rose
2025). Fitted peaks with integrals less than this threshold are indistinguishable from noise and are
discarded when extracting the size and shape of the star, as discussed in sections 4 and 5.
Figure 7 shows our measured visibilities as a function of the projected baseline. The gray points
represent those that pass the peak selections discussed in section 3.5 but which fall below the noise
threshold, shown as a gray band in the figure. These come from runs in which the noise is too large,
or the signal too small, to distinguish actual signal peaks from the noise. These gray points are not
used in further analysis. We have varied the threshold between 3.0 and 4.1 fs to obtain a systematic
uncertainty on fit parameters discussed in section 4.
10

Figure 7. The area of the fitted correlation peak, proportional to the squared visibility, as a function of
the magnitude of the baseline. The curve represents the best fit with a uniform disk model. Points in gray,
excluded from the fit, are below the threshold where the signal/noise is too low to reliably detect peaks, as
discussed in section 3.6. This threshold is marked with a gray shaded box.

4. EXTRACTING THE SIZE AND SHAPE OF THE PHOTOSPHERE


In this section, we fit the visibilities with simple models, to extract an effective size, shape, and
orientation of the star. It is common to model a star as a uniformly illuminated disk of angular
radius θU D . In this case, the squared visibility for a projected baseline b is
 2
2J1 (πb θUD /λ)
A(b) = CUD × , (4)
πb θUD /λ
where λ = 416 nm and J1 is the first-order Bessel function. CUD is the proportionality constant
between actual squared visibilities and peak integrals measured from correlation functions, which we
treat as a fit parameter. The red curve in figure 7 is the best fit of equation 4 to the data. The
effective uniform disk angular radius is 0.44 ± 0.01(stat) ± 0.02(syst) mas, consistent with the result
reported by the MAGIC collaboration (Abe et al. 2024) of 0.515 ± 0.038(stat) ± 0.023(syst). Our
systematic uncertainties were calculated by varying the peak quality cuts and threshold and not
removing tracking corrections or accounting for the background light contributions. As discussed
below, this star is not expected to be circular, so some tension between the values of θU D measured
by different observatories may be expected.
A rapid rotator like γ Cas is expected have an equatorial bulge and anisotropic gravitational
darkening, which would lead to an oblate source with spatial extension perpendicular to the axis
of rotation. Therefore, fitting the data with a uniform circular disk may not be appropriate, as the
length scale may depend on the orientation of the baseline. An intuitive way to explore this possibility
11

o
Slice 6: 22.5 < φ < 45o Slice 5: 0o < φ < 22.5o o
Slice 4: -22.5 < φ < 0o
o
Slice 3: -45 < φ < -22.5o
b b b b
A (fs)

A (fs)

A (fs)

A (fs)
18 χ2 / ndf 19.75 / 14 18 χ2 / ndf 11.22 / 12 18 χ2 / ndf 18.49 / 15 18 χ2 / ndf 32.26 / 9
16 θUD (mas) 0.57 ± 0.02 16 θUD (mas) 0.57 ± 0.02 16 θUD (mas) 0.47 ± 0.01 16 θUD (mas) 0.47 ± 0.03
14 14 14 14
12 12 12 12
10 10 10 10
8 8 8 8
6 6 6 6
4 4 4 4
2 2 2 2
0 0 0 0
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
|b| (m) |b| (m) |b| (m) |b| (m)
Slice 7: 45 < φ < 67.5o
o
Slice 2: -67.5 < φ < -45o
o
b b
A (fs)

A (fs)
18 χ2 / ndf 35.35 / 21 18 χ2 / ndf 9.32 / 21
16 θUD (mas) 0.50 ± 0.01 16 θUD (mas) 0.43 ± 0.01
14 14
12 12
10 10
8 8
150
vλ (m)

6 6
4 4
5 4
2 2
0 100 6 3 0
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
|b| (m) |b| (m)
Slice 8: 67.5 < φ < 90
o
Slice 1: -90 < φ < -67.5o
o
b
o
7 2 b
A (fs)

A (fs)
18 χ2 / ndf 18 χ2 / ndf
15.72 / 11 50 15.24 / 5
16 θUD (mas) 0.48 ± 0.02 16 θUD (mas) 0.44 ± 0.02
14 8 1 14
12 12
10 0 10
8 8
6
150 100 50 0 −50 −100 −150 6
4
2
uλ (m) 4
2
0 0
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
|b| (m) |b| (m)

Figure 8. The center panel shows the u − v midpoint of all runs and telescope pairs, with different color
shaded regions indicating angular slices in the u − v plane. For data within each of the individual slices,
the surrounding panels of corresponding colors show the squared visibility as a function of the length of the
baseline. Each slice is fit with a uniform disk model, to extract the effective 1-dimensional radius of a given
angular projection. The zero-baseline squared visibility is constrained to be the same for all slices, with a
value from the uniform ellipse fit of figure 10. Identical light shaded regions in each panel show the range
of fits between the most extreme values of the 8 individual slices; these are drawn to guide the eye, making
small differences between panels more apparent. A datapoint at large baseline in slice 5 is drawn in red and
discussed in the text.

is to measure the uniform disk effective angular size as a function of angle of the projected baseline
in the u − v plane, ϕb . The angle with the smallest effective angular diameter would correspond to
the position angle of the star rotation axis.
In figure 8, the visibility measurements are sorted according to ranges in ϕb . In the center panel,
the baseline for each measurement is shown in the u − v plane, and 22.5◦ -wide slices are defined
and shown as differently colored regions. Measurements that fall within each slice are shown in
the surrounding panels and fitted with the uniform disk model of equation 4 to extract an effective
one-dimensional angular diameter for a small range in ϕb . While the visibility measurement from
T1T4 with baseline of 145 m passes our threshold cut (see figures 7 and 10), we suspect it may be
12
χ2 / ndf 9.69 / 5

θUD (mas)
θUD (mas) 0.48 ± 0.01
0.65
δ 0.13 ± 0.02
φ* (o) 117.21 ± 4.13
0.6

0.55

0.5

0.45

0.4
−100 −50 0 50 100
φ (o)
b

Figure 9. The uniform-disk radius from fitting A versus baseline to datapoints in different slices in the
u − v plane (see figure 8), as a function of the average angle of the slice.

a fluctuation, like those discussed in section 3.6. Therefore, we report fit parameters obtained both
including and excluding this point, and it is drawn as a red square in Figures 8, 10, and 11.
Figure 9 shows the effective angular diameter versus the average position angle of the slice. The
data are well fit with the function

θU D (ϕb ) = θU D (1 − δ cos(2(ϕb − ϕ∗ ))). (5)

The average effective diameter is θU D = 0.48 ± 0.01 mas. The minimum diameter, the minor axis, is
at ϕb = ϕ∗ = 117◦ ± 4◦ , which we associate with the position angle of the stellar rotation axis and
is consistent with previous measurements of the accretion disk of this star (Sigut et al. 2020). The
extracted magnitude of the modulation, δ = 0.13 ± 0.02 corresponds31 to a maximum-to-minimum
ratio of 1+δ
1−δ
= 1.30 ± 0.05.
While figures 8 and 9 provide an intuitive way to estimate the anisotropy and orientation of the
rapid rotator, it is more accurate to perform a simultaneous fit of all visibilities as a function of u
and v. The hour angle, H, for a given source and observatory, uniquely defines the u − v coordinates
for each telescope pair, so we fit the data as a function of hour angle (Dravins et al. 2013). The data
is fitted with a uniformly illuminated ellipse model (Tycner et al. 2006)
 2
2 2J1 (πθmin s)
A(H) = A(u(H), v(H)) = Cnorm |V | = Cnorm , (6)
πθmin s
where p
s≡ r2 (u cos ϕ∗ − v sin ϕ∗ )2 + (u sin ϕ∗ + v cos ϕ∗ )2 . (7)

31
If an oscillation A sin(2ϕ) is uniformly integrated over bins of width ∆, the resulting histogram will exhibit oscil-
lations of magnitude A sin∆∆ . In our case, this is small effect: the finite-binning-corrected oscillation magnitude is
π/8
0.13 sin(π/8) = 0.133.
13

V2

V2
1.4
T1T2 1.4

T1T3
1.4

T1T4
1.2 1.2 1.2

1 1 1

0.8
V2

0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

V2

V2
0 0
−8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8
1.4 T2T4
T2T3 1.4 (h)
Hour angle 1.4 (h)
Hour angle
T3T4 Hour angle (h)

1.2 1.2 1.2

1 1 1

0.8 0.8 0.8


V2

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0
−8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8
Hour angle (h) Hour angle (h) Hour angle (h)

Figure 10. For each telescope pair, the normalized visibility is shown as a function of local hour angle.
Curves represent the best fit with a uniform ellipse model. The singular T1T4 data point is shown as a red
square, because it passes all quality cuts yet we suspect it may be a fit to random background fluctuations,
like those discussed in section 3.6. It is included in the fit shown. Fits are reported with and without this
point in table 2.

Here, r is the ratio of the major to minor axes (r ≥ 1), θmin is angular diameter of the minor axis,
and ϕ∗ is the position angle of the minor axis of the ellipse.32 For a rapid rotator, then, ϕ∗ is the
position angle of the axis of rotation.
The proportionality between the actual squared visibility and the peak integrals is Cnorm , which
we treat as a fit parameter. Figure 10 shows the peak integrals (divided by Cnorm ) for each telescope
pair, as a function of local hour angle.
This more complete model describes the data better than the uniform disk model discussed above;
see table 2. The angular size of the minor axis is θmin = 0.43±0.02(stat)±0.02(syst) mas, and the axis
ratio is r = 1.28 ± 0.04 ± 0.02. Fitting an elliptical source with a uniform circular model (equation 4)
would result in a uniform disk radius (θD ) somewhere between 0.43 mas and 0.55 mas, depending on
the orientation of the baseline. Therefore, the value of θU D extracted from any given measurement
will depend on the orientation of the baselines used during observation. This complicates comparisons
between, say VSII and MAGIC (Abe et al. 2024), when using a circular source model on a rapid
rotator.
It also means that θU D from equation 5 and θU D from equation 4 represent slightly different quan-
tities. While figures 8 and 9 are an intuitive way to display the non-circular nature of the star, the

32
Note that Tycner et al. (2006) describes the model for r ≤ 1.
14

characteristics of the elliptical appearance of the star are most accurately extracted by a fit to an
elliptical image model.
The extracted position angle of the rapid rotator star axis is estimated to be ϕ∗ = 116◦ ± 5◦ (stat) ±
7◦ (syst), consistent with measurements of the decretion disk at longer wavelengths (Sigut et al. 2020),
indicating that the decretion disc is generally aligned along the equatorial bulge of the photosphere
itself.

Table 2. Parameters of simple models fits.

T1T4 point Model ϕ∗ (º) θmin (mas) θmaj /θmin Cnorm (fs) χ2 /ndf
Uniform Disk - 0.44 ± 0.01 ± 0.03 (1) 11.2 ± 0.2 ± 0.5 213/114
Excluded uv slices 117 ±4 ±6 0.42 ± 0.02 ± 0.02 1.30 ± 0.05 ± 0.03 12.4* 9.69/5
Uniform Ellipse 116 ±5 ±7 0.43 ± 0.02 ± 0.02 1.28 ± 0.04 ± 0.02 12.4 ± 0.5 ± 0.5 170/112
Uniform Disk - 0.43 ± 0.01 ± 0.02 (1) 11.0 ± 0.2 ± 0.5 215/115
Included uv slices 118 ±4 ±6 0.41 ± 0.01 ± 0.02 1.27 ± 0.05 ± 0.04 12.2* 7.39/5
Uniform Ellipse 117 ±5 ±8 0.42 ± 0.02 ± 0.02 1.27 ± 0.05 ± 0.03 12.2 ± 0.5 ± 0.5 178/113
Note—note that χ2 /ndf for the uv slices analysis is based on sinusoidal fit to round model fits of the 8
“pizza slice” sections.
*Cnorm for the uv slice model is fixed to Cnorm from the Uniform Ellipse model fit.
Errorbars are reported as ±(stat) ± (syst).

5. CONSTRAINING A STELLAR MODEL OF A RAPID ROTATOR


We have computed synthetic squared visibilities, high-resolution spectra and spectral energy distri-
butions for the photosphere of γ Cas using PHOENIX (Hauschildt & Baron 1999, version 20.01.02B),
a code for nova, supernova, stellar and planetary model atmospheres. We use a Roche–von Zeipel
model for a rapidly rotating star (Aufdenberg et al. 2006; Sackrider & Aufdenberg 2023), modeled
as an infinitely concentrated central mass under uniform angular rotation.
Each axially symmetric model is parameterized by a trigonometric parallax, ϖ, equatorial angular
diameter, θeq , polar effective temperature, Tpole , polar gravity, log(g)pole , fraction of the critical
(breakup) angular rotation rate, Ω/Ωc , and the von Zeipel gravity darkening parameter, β. As seen
from Earth, the model is further parameterized by the inclination i, and the position angle, ϕ∗ , of
the visible stellar rotation axis.
The intensity of each point on the surface of the model star, as viewed from Earth, is interpolated
from a grid of 106 PHOENIX stellar atmosphere radiation fields: specific intensities at 78 direction
cosines between surface normal and the direction of an observer and 106 (Teff , log(g)) pairs (with
step sizes of 500 K in Teff and 0.1 in log(g)) between the hot, high gravity pole and the cooler, lower
gravity equator. A model rotating very close to the critical limit (Ω/Ωc = 0.9999) has a surface
temperature difference of ∆Teff ≃ 12000 K and a gravity difference of ∆ log(g) ≃ 1.2.
For our model of γ Cas we fixed ϖ = 5.94 mas (van Leeuwen 2007, 5.94 ± 0.12 mas), i = 60◦ as
constrained by Lailey & Sigut (2024) (see also Sigut et al. (2020) and Tycner et al. (2006)), and
15

Figure 11. Using the same data shown in figure 10, the best fit using a Roche-von Zepiel models for a
rapidly rotating star.

β = 0.20, as empirically determined for rapid rotating stars imaged with CHARA/MIRC (Che et al.
2011).
We computed 7750 models for γ Cas with 31 θeq values from 0.40 mas to 0.70 mas, 25 ϕ∗ values
from 90◦ to 138◦ and 10 Ω/Ωc values from 0.9000 to 0.9999. With six baselines and 118 hour angles,
this corresponds to a table of 5,487,000 synthetic visibility values.
Figure 11 shows the VERITAS data (the same as in Figure 10) with a Roche-von Zeipel stellar
model. Figure 12 shows the distributions of the best-fit parameters. The χ2 /ndf fit to the data
confronts the upper physical bound of unity, therefore we can only define a 1σ lower limit on the
rotation rate at 97.7%
Figure 13 is a rendering of the calculated light intensity distribution corresponding to the Roche-von
Zeipel model that fits the data best.
For the best-fitting median values, θeq = 0.60 mas and Ω/Ωc = 0.99, Tpole = 26500 K provides
a reasonable match to available absolute archival spectrophotometry in the UV optical range (see
Figure 14). A log(g)pole = 3.82 yields a mass of 15.1 M⊙ , at the top of the range 13-15 M⊙ adopted
by Smith (2019). Table 3 shows the best fit and the derived parameters.
Our adopted polar gravity and inclination yield a projected rotation velocity of v sin i = 389 ±
20 km s−1 . Chauville et al. (2001) found v sin i values 420-425 km s−1 by comparing models to the
He I λ 4471 line. Figure 15 shows a comparison between the best-fit model to the squared visibility
data and a high-resolution spectrum in the VSII band around 416 nm. Although the Hδ line is filled
with emission from the disk, the model roughly matches the wings of the line profile. The other
recognizable absorption line is He I λ 4026, which is stronger than the model with roughly the same
width.
16

eq (mas) = 0.604+0.041
0.034

/ c = 0.990+0.007
0.013
2 4 6 8
0.9 0.9 0.9 0.9
/ c

* (degrees) = 114.741+6.362
5.721
8
12
* (degrees)

0
12
2
11
4
10

CRvZ × 106 (ns) = 12.324+0.888


0.769
CRvZ × 106 (ns)
.5
13
.0
12
.5
10

2
4
6
8

.5

.0

.5
0.5

0.5

0.6

0.6

0.7

0.9
0.9
0.9
0.9

10

11

12

12

10

12

13

eq (mas) / c * (degrees) CRvZ × 106 (ns)

Figure 12. Best fit parameter distributions for the Roche-von Zeipel stellar model from fitting 1000
bootstrap samples with replacement from Table B1. The dashed lines show the 1σ lower bound error and
the 1σ upper bound of each parameter: θeq , the equatorial angular diameter; Ω/Ωc , fraction of the critical
angular rotation rate; ϕ∗ position angle of the visible rotation axis; and CRvZ , the proportionality constant
between the model squared visibilities and measured correlation peak integrals.
17

Figure 13. A synthetic photosphere for γ Cas with θeq =0.60 mas, Ω/Ωcrit =0.9888 and ϕ∗ = 114◦ consistent
with the best fit values to the VERITAS interferometry.

6. SUMMARY AND OUTLOOK


We have used the VERITAS telescope array to perform intensity interferometry measurements
on the rapid rotator γ Cassiopeiae, combining data from more than 160 pair-hours of observation.
While the size and elongation of the decretion disk of this star have been well measured at infrared
wavelengths, measurements of the photosphere at visible wavelengths have previously only been
sensitive to its size. The six telescope pairs of the VSII observatory provide good coverage of the
Fourier u − v plane. By varying the orientation as well as the magnitude of the baseline between
telescopes, we were able to measure the size, the equatorial bulge, and the orientation of the axis
of rotation of the photosphere. This is the first time that intensity interferometry has been used to
measure the oblateness of a star.
A uniform ellipse fit of our data indicates that the angular size of the minor axis of the photosphere
is 0.43 ± 0.02 ± 0.02 mas, with a major-to-minor axis ratio of 1.28 ± 0.04 ± 0.02. The position angle
of the minor axis, which corresponds to the rotator axis, is 116◦ ± 5◦ ± 7◦ , in agreement with earlier
measurements of the accretion disk.
We have also compared our data to a Roche-von Zeipel model of a rapid rotator, which includes
effects of limb and gravitational darkening as well as temperature gradients. The temperature profile
was adjusted to match published spectrophotometry measurements. Our data is well fit within this
model with an equatorial diameter of 0.604+0.041 +6.4
−0.034 mas, position angle of 114.7−5.7 degrees, and a lower
limit on the angular velocity very near the critical value, Ω/Ωc = 0.977. The equatorial radius is
+0.8
10.9−0.6 R⊙ and the mass of the star in the model is 15 ± 2 M⊙ .
18

Figure 14. A comparison of two Roche-von Zeipel synthetic spectral energy distributions (SEDs) to archival
absolute spectrophotometry of γ Cas between 1200 Å and 8170 Å. Data for wavelengths beyond 3200 Å are
from Burnashev (1985). International Ultraviolet Explorer (IUE) data have been rebinned from high
dispersion to low dispersion (González-Riestra et al. 2000) and extracted from the IUE Newly-Extracted
Spectra (INES) data archive. The faster spinning model (Ω/Ωc = 0.9999) is more gravity darkened and
requires a larger polar temperature at the same equatorial angular diameter to match the spectrophotometry.

Previous stellar intensity interferometry measurements have been limited to extracting the overall
size of the photosphere and resolving the extended decretion disk in its Hα light (Matthews et al.
2023). In order to take advantage of our coverage in the u − v plane and gain sensitivity to the stellar
shape, we found it important to remove artifacts from external radio interference and manual tracking
corrections. Our remaining systematic uncertainties are dominated by uncertainties in the clock start
synchronization, the peak-finding algorithm, and by residual correlated noise in the background. The
overall systematic error is on the order of the statistical error, suggesting that only adding hours at
previously unobserved hour angles would increase the precision of our measurement.
Having established the capability of stellar intensity interferometry to measure not only the size, but
also the shape and orientation of a photosphere, we are continuing analysis of other rapid rotators, to
expand the scope of data in the stellar catalog of bright stars. This will provide valuable input and
constraints to stellar models that must match both the spectral and the geometrical consequences of
extreme rotation.
19

Figure 15. A normalized, high-resolution spectrum of γ Cas in the VSII bandpass from the ELODIE archive
(Moultaka et al. 2004, observation 20030816/0030, in black) and a normalized model spectrum (in blue) from
the same model Roche-von Zeipel that best fits the VERITAS data in Table B1. The rotationally broadened
Hδ line at 4100 Å is filled in the emission from the disk surrounding γ Cas. The disk is not included in our
model.

Table 3. Stellar Parameters for best fitting Roche-von Zeipel model for γ Cas

Parameter Value Reference


With a fixed inclination of the rotation axis, i = 60◦ and gravity darkening β = 0.2
Equatorial angular diameter, θeq (mas) 0.605+0.041
−0.034 This paper
+0.007
Fraction of critical rotation rate, ω = Ω/Ωc 0.990−0.013 This paper
Parallax, ϖ (mas) 5.94 ± 0.12 van Leeuwen (2007)
θ
Equatorial radius, Req (R⊙ ) 10.9+0.8
−0.6 derived, Req = 107.48 ϖeq
Polar Radius, Rpole (R⊙ ) 7.9 ± 0.4 derived, Rpole = ωRequ /η

With a fixed log(g)polar = 3.82, i = 60 , gravity darkening β = 0.2
Mass, M (M⊙ ) 15 ± 2 2 /G
M = gpole Rpole
Equatorial velocity, veq (km s−1 ), 450 ± 20 veq = Req Ω
−1
Projected equatorial velocity, v sin i (km s ) 389 ± 20 derived from veq and i = 60◦ , see Figure 15
Polar effective temperature, Tpole (K) ≃ 26500 − 28000 spectrophotometry, see  Figure 14 
2
Pole-equator difference, ∆Teff (K) ≃ 9200 − 12000 derived, ∆Teff = Tpole 1 − ( ωη2 − 27 8
ηω)β
−1
depend on η = 3 cos π+cos3 (ω) , while the equatorial velocity, veq ,
 
Note—The parameters Rpole and ∆Teff
q
8 GM
depends on Ω = ω 27 R3 pole

ACKNOWLEDGMENTS
This research is supported by grants from the U.S. Department of Energy Office of Science, the
U.S. National Science Foundation and the Smithsonian Institution, by NSERC in Canada, and by the
20

Helmholtz Association in Germany. We acknowledge the excellent work of the technical support staff
at the Fred Lawrence Whipple Observatory and the collaborating institutions in the construction and
operation of the VERITAS and VSII instruments. The authors gratefully acknowledge support under
NSF grants #AST 1806262, #PHY 2117641, and #PHY 2111531 for the construction and operation
of the VSII instrumentation. We acknowledge the support of the Ohio Supercomputer Center and
of an Ohio State University College of Arts and Sciences Exploration Grant. We acknowledge the
financial support of the French National Research Agency (project I2C, ANR-20-CE31-0003). This
work made use of the High Performance Cluster “Vega” at Embry-Riddle Aeronautical University.
This research has made use of the VizieR catalogue access tool, CDS, Strasbourg, France (DOI :
10.26093/cds/vizier). The original description of the VizieR service was published in 2000, A&AS
143, 23. Part of this research is based on INES data from the IUE satellite.
Software: PHOENIX (version 20.01.02B) (Hauschildt & Baron 1999), Astropy (Astropy Collabo-
ration et al. 2018), specutils (Earl et al. 2022), Minuit (James & Roos 1975), root (Brun & Rademakers
1997), Cobaya (Torrado & Lewis 2021), SciPy (Virtanen et al. 2020), Corner (Foreman-Mackey 2016).
21

APPENDIX

A. CALCULATING UNCERTAINTIES ON PEAK AREAS WITH CORRELATED


BACKGROUND CONTRIBUTION
This appendix illustrates in more detail the calculation of an uncertainty on the area of the Gaussian
peak, which we discuss in section 3.5. Figure A1 shows an example of a fit to a simulated peak located
at about -70 ns in relative time, with the same area as the original correlation peak located at 0 ns
in relative time. The simulated peak is inserted into the background noise away from the peak,
fitted with the same procedure as the original peak, and subjected to the same quality cuts. This
is repeated about 5000 times for each correlation function. The differences between the areas found
from these fits and the true area of the simulated peak are added to a distribution like the one shown
in figure A2. The uncertainty on the peak area is the RMS of this distribution, which describes how
much the background noise can distort the area.

Figure A1. An example of one simulated peak inserted into the background noise and fitted with a Gaussian
function to extract its area.

B. UNNORMALIZED VISIBILITIES ABOVE THRESHOLD


Table B1 lists all Gaussian integrals for visibilities above thresholds discussed in section 3.
22

Figure A2. A distribution of the difference between the original known area of the simulated peaks and
the areas found by the fitting algorithm once they are inserted into the background noise, for many fits. The
RMS of this distribution is the uncertainty on the peak.

Table B1. γ Cassiopeiae squared visibilities

UTC Date (yyyy-mm-dd hh:mm) Pair Area (fs) Baseline (m) uλ (m) vλ (m) Hour angle (h)
2023-12-25 01:57 T2T4 7.2 ± 1.0 95.5 -82.5 47.6 0.28
2023-12-25 03:04 T2T4 6.8 ± 1.3 90.0 -86.2 25.0 1.44
2023-12-26 01:34 T1T2 7.4 ± 1.3 96.2 -90.4 -32.4 0.00
2023-12-26 01:34 T1T2 4.6 ± 0.9 98.5 -83.5 -51.9 0.98
2023-12-26 01:34 T1T3 5.1 ± 0.5 120.0 -108.9 48.2 0.50
2023-12-26 01:34 T2T3 6.2 ± 0.8 94.3 -15.5 92.9 0.00
2023-12-26 01:34 T2T3 7.0 ± 0.9 92.5 -28.4 88.0 0.98
2023-12-26 01:34 T2T4 6.3 ± 1.0 96.5 -80.5 52.8 0.00
2023-12-26 01:34 T2T4 5.7 ± 1.1 92.5 -85.7 34.3 0.97
2023-12-26 01:34 T3T4 6.6 ± 1.1 76.5 -65.0 -40.1 0.01
2023-12-26 01:34 T3T4 5.6 ± 1.2 78.8 -57.5 -53.7 0.97
2023-12-26 03:37 T2T3 5.8 ± 1.5 89.2 -40.1 79.6 2.04
2023-12-26 03:37 T2T3 6.5 ± 1.6 84.7 -48.5 69.3 3.05
2023-12-26 03:37 T2T4 7.5 ± 1.1 86.6 -85.3 14.3 1.99
2023-12-26 03:37 T2T4 8.5 ± 1.5 78.4 -78.0 -6.6 3.10
2023-12-26 03:37 T3T4 7.0 ± 1.3 80.2 -45.7 -65.8 1.99
2023-12-26 03:37 T3T4 3.8 ± 0.9 80.9 -29.7 -75.1 3.06
2023-12-26 05:38 T1T2 4.0 ± 1.4 99.4 -30.3 -94.4 4.07
2023-12-26 05:38 T1T2 4.4 ± 1.7 99.2 -7.5 -98.6 5.04
23

UTC Date (yyyy-mm-dd hh:mm) Pair Area (fs) Baseline (m) uλ (m) vλ (m) Hour angle (h)
2023-12-26 05:38 T1T3 5.7 ± 0.8 92.0 -83.9 -37.1 4.07
2023-12-26 05:38 T2T3 5.4 ± 1.1 78.6 -53.6 57.3 4.07
2023-12-26 05:38 T2T3 7.9 ± 1.4 71.2 -54.9 45.2 5.04
2023-12-26 05:38 T2T4 9.1 ± 1.0 70.1 -66.2 -22.6 4.07
2023-12-26 05:38 T2T4 8.7 ± 1.1 61.7 -50.2 -35.5 5.03
2023-12-26 05:38 T3T4 5.4 ± 0.9 81.1 -12.6 -80.0 4.07
2023-12-26 05:38 T3T4 8.6 ± 1.0 81.1 4.9 -80.8 5.04
2023-12-26 07:39 T1T2 4.5 ± 1.9 99.3 20.3 -96.7 6.21
2023-12-26 07:39 T1T3 4.4 ± 1.1 73.8 -31.3 -65.8 6.21
2023-12-26 07:39 T2T3 12.2 ± 2.2 60.3 -51.7 31.0 6.20
2023-12-26 07:39 T2T4 8.6 ± 1.5 53.6 -26.6 -45.7 6.21
2023-12-26 07:39 T3T4 6.3 ± 0.9 81.0 25.1 -76.6 6.21
2023-12-27 01:43 T1T2 5.0 ± 1.2 96.8 -89.4 -36.5 0.20
2023-12-27 01:43 T1T3 3.6 ± 0.8 119.2 -109.8 44.2 0.66
2023-12-27 01:43 T2T3 6.4 ± 0.9 91.9 -31.0 86.5 1.20
2023-12-27 01:43 T2T4 7.5 ± 1.5 95.8 -82.0 49.1 0.20
2023-12-27 01:43 T2T4 10.8 ± 1.4 91.4 -86.1 30.1 1.18
2023-12-27 01:43 T3T4 6.0 ± 1.2 77.1 -63.8 -43.1 0.21
2023-12-27 03:44 T1T2 4.6 ± 0.9 99.7 -66.7 -73.7 2.24
2023-12-27 03:44 T1T3 6.2 ± 1.1 105.4 -104.3 -5.8 2.65
2023-12-27 03:44 T2T3 8.0 ± 1.2 88.5 -42.0 77.7 2.24
2023-12-27 03:44 T2T3 7.9 ± 1.7 83.7 -49.7 67.2 3.24
2023-12-27 03:44 T2T4 11.0 ± 1.5 84.9 -84.2 9.4 2.24
2023-12-27 03:44 T2T4 8.4 ± 2.1 77.3 -76.6 -8.8 3.22
2023-12-27 03:44 T3T4 7.0 ± 1.4 80.4 -42.2 -68.3 2.24
2023-12-27 05:47 T1T2 3.7 ± 1.6 99.3 -25.8 -95.6 4.26
2023-12-27 05:47 T1T3 6.9 ± 1.0 90.0 -79.9 -40.7 4.27
2023-12-27 05:47 T1T3 9.6 ± 1.7 81.6 -60.2 -54.8 5.14
24

UTC Date (yyyy-mm-dd hh:mm) Pair Area (fs) Baseline (m) uλ (m) vλ (m) Hour angle (h)
2023-12-27 05:47 T2T3 11.1 ± 1.5 70.5 -54.9 44.1 5.13
2023-12-27 05:47 T2T4 7.4 ± 1.3 68.4 -63.2 -25.5 4.27
2023-12-27 05:47 T2T4 8.5 ± 1.8 60.7 -47.9 -37.0 5.16
2023-12-27 05:47 T3T4 7.1 ± 1.2 81.1 6.3 -80.8 5.12
2023-01-08 05:23 T1T3 7.8 ± 1.3 86.1 -71.5 -47.5 4.66
2023-01-08 05:23 T1T3 7.7 ± 1.1 76.9 -45.4 -61.6 5.70
2023-01-08 05:23 T2T3 8.8 ± 1.3 74.4 -54.8 50.3 4.64
2023-01-08 05:23 T2T3 9.3 ± 1.3 65.7 -53.9 37.5 5.66
2023-01-08 05:23 T2T4 7.5 ± 1.2 64.9 -56.7 -31.0 4.67
2023-01-08 05:23 T2T4 9.3 ± 1.8 56.7 -37.7 -42.0 5.68
2023-01-08 05:23 T3T4 6.4 ± 1.2 81.1 -1.9 -81.0 4.66
2023-02-02 03:10 T1T2 5.1 ± 1.3 99.4 -29.8 -94.6 4.09
2023-02-02 03:10 T1T3 8.8 ± 1.5 91.7 -83.4 -37.6 4.10
2023-02-02 03:10 T1T3 7.5 ± 1.1 82.2 -61.7 -53.8 5.07
2023-02-02 03:10 T2T3 9.1 ± 0.9 78.4 -53.7 57.1 4.09
2023-02-02 03:10 T2T3 7.6 ± 1.7 71.0 -54.9 44.9 5.07
2023-02-02 03:10 T2T4 6.0 ± 1.6 70.0 -65.9 -22.9 4.08
2023-02-02 03:10 T2T4 11.7 ± 1.7 61.4 -49.4 -36.0 5.07
2023-02-02 03:10 T3T4 7.5 ± 1.7 81.1 -12.3 -80.0 4.08
2023-02-02 03:10 T3T4 7.3 ± 1.1 81.1 5.5 -80.8 5.08
2023-02-02 05:12 T1T3 4.0 ± 1.2 73.6 -31.0 -66.0 6.22
2023-02-02 05:12 T2T3 11.2 ± 1.6 60.2 -51.7 30.8 6.22
2023-02-02 05:12 T2T4 7.9 ± 1.4 53.5 -26.5 -45.8 6.21
2024-02-19 02:31 T1T3 6.1 ± 1.0 87.2 -73.9 -45.7 4.55
2024-02-19 02:31 T2T3 9.9 ± 1.1 75.1 -54.7 51.3 4.55
2024-02-19 02:31 T3T4 4.7 ± 1.3 81.1 -4.1 -80.9 4.54
2024-02-19 03:38 T1T2 4.4 ± 1.7 99.2 7.9 -98.6 5.68
2024-02-19 03:38 T1T3 5.7 ± 1.6 77.4 -46.9 -61.0 5.65
25

UTC Date (yyyy-mm-dd hh:mm) Pair Area (fs) Baseline (m) uλ (m) vλ (m) Hour angle (h)
2024-02-19 03:38 T1T4 4.1 ± 1.1 144.4 -31.1 -140.4 5.65
2024-02-19 03:38 T2T3 9.9 ± 1.5 65.5 -53.8 37.3 5.68
2024-02-19 03:38 T2T4 10.6 ± 2.1 56.7 -37.5 -42.1 5.68
2024-02-20 02:07 T1T2 3.7 ± 1.9 99.4 -29.9 -94.6 4.09
2024-02-20 02:07 T1T3 9.0 ± 2.3 91.8 -83.6 -37.5 4.09
2024-02-20 02:07 T2T3 9.4 ± 1.9 78.6 -53.7 57.4 4.07
2024-02-20 02:07 T2T4 4.4 ± 1.6 70.1 -66.1 -22.7 4.07
2024-02-20 03:14 T1T3 6.1 ± 1.7 81.3 -59.4 -55.3 5.17
2024-02-20 03:14 T2T3 7.4 ± 2.8 70.2 -54.9 43.7 5.16
2024-02-21 02:22 T1T3 10.3 ± 1.1 88.1 -75.8 -44.2 4.46
2024-02-21 02:22 T2T3 9.6 ± 2.9 75.8 -54.6 52.5 4.46
2024-02-22 03:26 T1T3 5.8 ± 2.0 78.9 -52.4 -58.9 5.44
2024-02-22 03:26 T2T4 7.4 ± 1.6 58.5 -42.7 -39.9 5.43
2024-02-22 03:26 T3T4 4.5 ± 1.1 81.1 11.9 -80.2 5.44
2024-02-22 03:59 T2T3 10.2 ± 2.3 62.7 -52.9 33.5 5.98
2024-02-22 03:59 T3T4 7.9 ± 2.3 81.0 21.4 -78.1 5.98
2024-02-22 04:32 T1T3 5.6 ± 2.1 71.5 -18.8 -68.7 6.65
2024-02-22 04:32 T2T3 9.0 ± 1.8 55.7 -49.4 25.7 6.65
2024-02-22 04:32 T2T4 8.8 ± 1.4 51.4 -17.2 -48.2 6.64
2024-05-18 10:00 T1T2 6.7 ± 1.7 62.4 -14.9 60.2 -6.02
2024-05-18 10:00 T2T3 8.2 ± 1.3 80.1 52.8 60.1 -6.02
2024-05-18 10:00 T3T4 16.8 ± 1.3 42.0 -21.4 35.8 -6.02
2024-05-19 10:07 T1T2 7.3 ± 1.4 62.8 -18.4 59.6 -5.87
2024-05-19 10:07 T2T3 4.7 ± 1.4 81.0 52.2 61.9 -5.87
2024-05-19 10:07 T3T4 11.9 ± 1.1 42.6 -23.7 35.1 -5.88
2024-12-13 01:34 T1T2 4.8 ± 1.2 93.3 -91.6 -16.6 -0.75
2024-12-13 01:34 T2T3 6.8 ± 0.9 94.9 -4.1 94.7 -0.80
2024-12-13 01:34 T2T4 6.5 ± 1.3 98.6 -72.3 66.8 -0.80
2024-12-13 01:34 T3T4 8.5 ± 1.0 73.7 -68.1 -27.8 -0.80
26

UTC Date (yyyy-mm-dd hh:mm) Pair Area (fs) Baseline (m) uλ (m) vλ (m) Hour angle (h)
2024-12-13 02:40 T1T2 4.1 ± 1.0 97.0 -88.9 -38.3 0.29
2024-12-13 02:40 T1T3 4.8 ± 1.1 121.1 -108.4 53.4 0.29
2024-12-13 02:40 T2T3 6.7 ± 0.9 93.8 -19.5 91.7 0.30
2024-12-13 02:40 T2T4 6.3 ± 1.3 95.4 -82.6 47.4 0.29
2024-12-13 02:40 T3T4 5.8 ± 1.3 77.3 -63.2 -44.4 0.29
2024-12-15 05:43 T1T2 4.0 ± 1.0 99.6 -44.7 -88.7 3.41
2024-12-15 05:43 T1T3 6.8 ± 1.3 98.6 -95.6 -23.4 3.40
2024-12-15 05:43 T2T3 5.4 ± 1.1 82.7 -50.8 65.2 3.41
2024-12-15 05:43 T3T4 7.3 ± 1.0 81.0 -23.8 -77.2 3.41
2024-12-17 02:28 T1T3 3.8 ± 0.9 120.8 -108.8 51.9 0.35
2024-12-17 02:28 T2T3 7.1 ± 0.6 93.7 -20.3 91.4 0.36
2024-12-17 02:28 T2T4 6.9 ± 0.7 95.2 -83.0 46.2 0.36
2024-12-17 02:28 T3T4 7.4 ± 1.0 77.5 -62.7 -45.2 0.36
27

REFERENCES
Abe, S., Abhir, J., Acciari, V. A., et al. 2024, James, F., & Roos, M. 1975, Comput. Phys.
Monthly Notices of the Royal Astronomical Commun., 10, 343,
Society, 529, 4387, doi: 10.1093/mnras/stae697 doi: 10.1016/0010-4655(75)90039-9
Abeysekara, A. U., Benbow, W., Brill, A., et al. Kieda, D., et al. 2021, PoS, ICRC2021, 803,
2020, Nature Astronomy, 4, 1164–1169, doi: 10.22323/1.395.0803
doi: 10.1038/s41550-020-1143-y Lailey, B. D., & Sigut, T. A. A. 2024, MNRAS,
Acharyya, A., et al. 2024, Astrophys. J., 966, 28, 527, 2585, doi: 10.1093/mnras/stad3321
doi: 10.3847/1538-4357/ad2b68 Matthews, N. 2020, PhD thesis, University of
Astropy Collaboration, Price-Whelan, A. M., Utah
Sipőcz, B. M., et al. 2018, AJ, 156, 123, Matthews, N., Rivet, J.-P., Vernet, D., et al. 2023,
doi: 10.3847/1538-3881/aabc4f AJ, 165, 117, doi: 10.3847/1538-3881/acb142
Aufdenberg, J. P., Mérand, A., Coudé du Foresto, Moultaka, J., Ilovaisky, S. A., Prugniel, P., &
V., et al. 2006, ApJ, 645, 664, Soubiran, C. 2004, PASP, 116, 693,
doi: 10.1086/504149 doi: 10.1086/422177
Brun, R., & Rademakers, F. 1997, Nucl. Instrum. Rose, J. 2025, Bachelor’s honors thesis, The Ohio
Meth. A, 389, 81, State University.
doi: 10.1016/S0168-9002(97)00048-X https://kb.osu.edu/handle/1811/105962
Chauville, J., Zorec, J., Ballereau, D., et al. 2001, Rou, J., Nuñez, P. D., Kieda, D., & LeBohec, S.
A&A, 378, 861, 2013, Monthly Notices of the Royal
doi: 10.1051/0004-6361:20011202 Astronomical Society, 430, 3187,
doi: 10.1093/mnras/stt123
Che, X., Monnier, J. D., Zhao, M., et al. 2011,
Sackrider, J. L., & Aufdenberg, J. P. 2023,
ApJ, 732, 68, doi: 10.1088/0004-637X/732/2/68
Research Notes of the American Astronomical
Dravins, D., LeBohec, S., Jensen, H., & Nuñez,
Society, 7, 216, doi: 10.3847/2515-5172/ad023b
P. D. 2013, Astroparticle Physics, 43, 331,
Scott, M. 2023, Bachelor’s honors thesis, The
doi: https:
Ohio State University.
//doi.org/10.1016/j.astropartphys.2012.04.017
http://hdl.handle.net/1811/102963
Earl, N., Tollerud, E., Jones, C., et al. 2022,
Sigut, T. A. A., Mahjour, A. K., & Tycner, C.
astropy/specutils: V1.7.0, v1.7.0, Zenodo,
2020, ApJ, 894, 18,
doi: 10.5281/zenodo.6207491 doi: 10.3847/1538-4357/ab8386
Foreman-Mackey, D. 2016, The Journal of Open Smith, M. A. 2019, PASP, 131, 044201,
Source Software, 1, 24, doi: 10.21105/joss.00024 doi: 10.1088/1538-3873/aaf70b
González-Riestra, R., Cassatella, A., Solano, E., Stee, Ph., Delaa, O., Monnier, J. D., et al. 2012,
Altamore, A., & Wamsteker, W. 2000, A&AS, A&A, 545, A59,
141, 343, doi: 10.1051/aas:2000317 doi: 10.1051/0004-6361/201219234
Hanbury Brown, R. 1974, The intensity Thompson, A. R., Moran, J. M., & Swenson Jr.,
interferometer. Its applications to astronomy G. W. 2001, Interferometry and Synthesis in
(London: Taylor & Francis Ltd.) Radio Astronomy (WILEY-VCH Verlag GmbH
Hanbury Brown, R., & Twiss, R. 1954, The & Co. KGaA)
London, Edinburgh, and Dublin Philosophical Torrado, J., & Lewis, A. 2021, JCAP, 2021, 057,
Magazine and Journal of Science, 45, 663, doi: 10.1088/1475-7516/2021/05/057
doi: 10.1080/14786440708520475 Tycner, C., Gilbreath, G. C., Zavala, R. T., et al.
Hauschildt, P. H., & Baron, E. 1999, Journal of 2006, Astron. J., 131, 2710, doi: 10.1086/502679
Computational and Applied Mathematics, 109, van Leeuwen, F. 2007, A&A, 474, 653,
41, doi: 10.48550/arXiv.astro-ph/9808182 doi: 10.1051/0004-6361:20078357
Holder, J., Atkins, R., Badran, H., et al. 2006, Virtanen, P., Gommers, R., Oliphant, T. E., et al.
Astroparticle Physics, 25, 391, doi: https: 2020, Nature Methods, 17, 261,
//doi.org/10.1016/j.astropartphys.2006.04.002 doi: 10.1038/s41592-019-0686-2
28

Zmija, A., Vogel, N., Wohlleben, F., et al. 2023,


Monthly Notices of the Royal Astronomical
Society, 527, 12243,
doi: 10.1093/mnras/stad3676

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy