MGAB-V1134 Light Curves by Bruce Gary
Webmaster: Bruce Gary, 2020.05.25 23 UT
Current
Status
Five LCs show dips occurring at times consistent with an
ephemeris that is derived from ZTF/PanSTARRS/NMPT data
(hereafter referred to as ZTF+). Dip depth has
increased during the past 6 days from 30 to 60 %. The most
recent depth is almost as great as for most of the ZTF+
data: 84 %. Transit length is 4.5 ± 0.5 minutes,
which can be produced by an object size of 18 ± 4 R_earth.
This size is consistent with what would be produced by a low-mass
brown dwarf (10 to 20 × R_earth). However, it
would be premature to rule-out a massive gas giant
planet (10 × R_earth). Transit dip depth has a pattern
of being lower than normal (30 % vs. 84 %) for maybe a week,
and these episodes of low depth repeat every 240 days (for the
past 3 years). We are now exiting such a low depth episode and
the next one is predicted for 2020 Dec 30. It is important to
monitor dip depth at representative times (such as weekly) to
be sure that low depth episodes do not occur until late 2020
December. It will be important for modeling to characterize
the dip shape accurately, which can require averaging many dip
events. One goal is to understand why transit depths change,
and derive a model to account for this (such as precession
produced by a third massive planet in an outer orbit). The
latest ephemeris for the fades is 2456409.8605 + E × (96.31015
/ 1440).
Links Internal to this Web
Page
General
information
List
of observation dates
Light curves for
each observation date
Phase-folded LC
analysis
Miscellaneous
analyses of ZTF and PanSTARRS data
Predicted
dip depth behavior
Finder image and
reference stars
Why
I like this star system
References
General Information
MGAB-V1134 is a 18.4-magnitude white dwarf that
has been observed by the ZTF and PanSTARRS to exhibit deep fades
at intervals of of ~ 96 minutes. Occasionally fades are absent
when expected, which is the most interesting aspect of the WD
system. Spectral type is DA (hydrogen lines only).
RA/DE = 11:00:45.2 +52:10:44, g'-mag = 18.34,
observing season centered on March 3.
SDSS image.
Suggested by analyses on this web page:
WD Teff = 25,000 K or greater
Secondary orbit period = 96.310150 minutes (BJDo
= 2456409.8605)
Secondary density is greater than 50 g/cc (36 to
62 g/cc)
Secondary's mass is within the range 9,500 to
48,000 M_earth
Secondary size = 10 x R_earth (if massive gas
giant planet) or 10 to 20 x R_earth if BD (based on density and mass
relationships)
Secondary size = 18 ± 4 x R_earth (based on
transit length of 4.5 ± 0.5 minutes)
Length of dips (ingress to egress) = 4.5 ± 0.5
minutes (when dip depth > 50 %)
Secondary is most likely a BD (based on size,
derived from length of transit)
Orbit of secondary changes (precesses?) with
period of 240 days (minimum dip depth occurs at BJD = 2458973 + E ×
240)
List of Observation
Dates
2020.05.17 - Gary
2020.05.15 Vanmunster
2020.05.14 Gary
2020.05.13 Gary
2020.05.12 Gary
2020.05.12 Vanmunster
2020.05.08 Gary
2020.05.07 Gary
2020.05.06 Gary
Light Curves
for Each Observation Date
2020.05.17 - Gary
Photometry signal aperture radii of 3 & 4 gave same NF (30
%), radius 4 gave slightly higher (32 %), so I averaged the 3
& 4 LC data sets for subsequent analysis.
Data
2020.05.14 Vanmunster
2020.05.14 Gary
Data
2020.05.13 Gary
Data
2020.05.12 Gary
NF = 30, 32 & 37 % for photometry aperture radii = 3, 4 &
5 pixels.
Data
2020.05.12
Vanmunster
JD range = 2458161.40 to 2458161.58 (May 12 UT range = -2.4 to 1.9).
2020.05.08
Full moon and some wind caused loss of precision.
Here are LCs that combine data for May 06, 07 and 08:
2020.05.07
2020.05.06
Phase-Folded LC
Analysis
Here's a phase-folded normalized flux LC for the ZTF/PS1/NMPT data
(hereafter abbreviated as ZTF+):
Figure P01. Phase folded normalized flux LC for the ZTF+
data.
Notice the 6 % peak-to-peak "illumination effect." This is either
produced by a planet reflecting WD light or a BD in a synchronous
orbit being heated by the WD on the WD-facing side.
The normalized flux at mid-transit is ~ 16 %, seen more clearly in
the next graph.
Figure P02. Phase folded normalized flux LC for the ZTF+
data.
Notice the 3 measurements in the middle of an expected transit that
are at the OOT level. It's important to know if transits were in
fact missing on those dates. If they were missing, then we would
have to invoke temporal variation of either the presence of the
transiting material (dust cloud) or orbital orientation of the
transiting object.
Let's see if transit depth varies with date using recent
ground-based observations. My first three LCs show a mid-transit
normalized flux of ~ 60 %.
Figure P03. Phase-folded normalized flux LC for May 06,
07 & 08.
Figure P04. Phase-folded normalized flux LC for May 12.
The May 12 observations by both Gary and Vanmunster are in
agreement that mid-transit normalized flux = ~25 %. If the trend
from 60 % for ~ May 7 to ~25 % for May 12 is true, then there's a
trend for increasing dip depth that suggests a return to ZTF+
level of 16 % (with a median date of 2018 May).
Saul suggests that a third massive object could be in an outer
orbit that causes the secondary's orbit to precess. Thus, over
time the transits could vary from non-existent to a nearly
complete eclipse. The evidence for this is explored in the next
section.
Miscellaneous
Analyses of ZTF & PanSTARRS Data
Consider the ZTF+ measurements for the transit phase region.
Figure M01. Phase folded normalized flux LC for
the ZTF/PS1/NMPT data.
There are 16 observations withing the phase interval where dips
are expected, and for 3 of these dips were not present. That's a 19
+/- 11 % non-dip fraction.
The dates for the 3 non-dip observations have separations of 727,
1665 and 2392 days. The largest interval that has integer
multiples in agreement with these 3 separations is 239 days. This
is a candidate for the interval between non-dip situations. If the
non-dip situation is 19 ± 11 % of the 239 day periodicity then
non-dipping would last 45 ± 26 days. An ephemeris for the
non-dipping situations is:
Non-dipping BJD = 2456338 + E x 239
This ephemeris predicts a non-dipping situation for BJD = 2458967
(corresponding to E = 11). This corresponds to 2020.04.27, with a
45-day range that extends from 2020.04.04 to 2020.05.20. Note that
this date range includes the observations shown above, in the
"Light Curves for Each Observation" section, where no dips are
present. Since the 45-day interval is uncertain by 26 days, the
non-dip date range is longer than these dates. Nevertheless, this
is a prediction that dipping should return sometime in late May or
early June of 2020.
The orbital period of ~ 96 minutes, combined with the requirement
that the secondary must exist no closer to the primary than the
Roche distance, means that the secondary must have a density
exceeding ~ 50 g/cc. Depending on the internal mass distribution
of the secondary, the bulk density can have a range from 36 to 62
g/cc. The following chart is useful for determining these density
values.
Figure M02. Secondary density minimum vs. orbit period
that assures secondary is outside the WD's Roche distance (based
on Rappaport et al., 2013). The 3 traces are for different
internal mass distributions (blue corresponds to ratio of
density at center to mean = 6, gray ratio = 1).
What secondary objects have these densities?
Figure M03. Empirical mass-density relation for
planets and stars (Hatzes & Rauer, 2015).
According to this figure the previously
determined minimum density for the secondary means that the
secondary should be either a "massive gas giant planet" or a
"low-mass brown dwarf." The graph also requires that the
secondary's mass is between 30 and 150 M_jup (or 9,500 to 48,000 M_earth).
Here's a graph showing the empirical relationship
between size and mass for planets and stars.
Figure M04. Empirical relation between mass and size for
planets and stars.
Given that the secondary's mass is within the range 9,500 to 48,000
M_earth, the above graph says that if the secondary is a gas giant
it will have R_secondary = 10 R_earth whereas if it's a BD
R_secondary will be within the range 10 to 20 R_earth.
What can we learn from the transit length (cf. Fig. M02)? The
ingress to egress interval is 0.047 ± 0.005 phase units, or 4.5 ±
0.5 minutes (when depth > 50 %). Let's first determine if the
mass of the secondary is significant enough to affect the total
system mass for orbit calculations? The maximum allowed BD mass of
150 M_jup = 0.14 M_sun (since M_sun = 1047 M_jup). So for the BD
case the mass of the secondary is ~ 1/4 of a typical WD mass.
Therefore, the secondary's mass will have a small effect on total
system mass. Let's consider a mass for the WD plus secondary = 0.63
(same as WD1145). Solving the Kepler equation yields an orbital
radius of 8.2e5 km, which corresponds to an orbit circumference of
5.2e6 km. 0.047 ± 0.005 phase units corresponds to 2.43e5 (±
10 %) km, which is 38 ± 4 R_earth. Adopting R_wd = 1.3 R_earth
yields R_secondary = 18 ± 4 R_earth.
Therefore, the length of the transit is similar to that of a BD,
but it would be premature to rule out planet. If the obscuring
object was much larger than either a planet or BD then we would have
needed to invoke a dust cloud to explain the transit length, but
this is not necessary. Indeed, a reasonable dust cloud would be too
large for the observed transit length.
Saul points out that a BD can have a flux comparable to a WD due to
its larger size (in spite of its lower temperature). If the WD and
BD have similar fluxes at optical wavelengths then the transit depth
for a full edge-on eclipse would be about 50 %. We don't know the
temperature of the WD, yet (working on that), so using the fact that
ZTF/PanSTARRS data show a 16 % normalized flux for mid-transit
constitutes a constraint on the WD/BD model. Of course, if the
secondary is a planet then we would observe a normalized flux of 0 %
for an eclipse.
Can we constrain anything about the BD using the V1134 SED?
SDSS magnitudes
Figure M05. SED for a WD with Teff = 25,000 K and a BD
with Teff = 3000 K and size that 12.7 / 1.32 = 9.6 times larger
than the WD.
If the secondary is a BD then it can't have a Teff much greater than
~ 3000 K. At higher Teff the SED would be incompatible with the
measured z'-mag. I suspect that a BD at the low end of the mass
region should have Teff << 3000 K. The only value of the SED,
therefore, is to establish a temperature minimum for the WD.
By the way, there are no 2MASS measurements of this star.
Predicted Dip Depth
Behavior
Based on the previous sections it is possible to predict dip depth
behavior using an empirical model. First, here's the data and model
for mid-transit normalized flux for the recent past:
Figure D01. Measured and predicted mid-transit normalized
flux vs. date (2020 April 20 to June 10).
All data sources, ZTF+, Gary and Vanmunster, are compatible with a
model for mid-transit normalized flux.
Here's the same comparison for the pst 3 years and the rest of this
year:
Figure D02. Measured and predicted mid-transit normalized
flux vs. date for a 3-year interval (2018 March to 2021 March)
Again, all data sources are compatible with the model.
So far the measurements are supportive of an empirical model
describing mid-transit normalized flux. Observations centered on
2020 December 30 will serve to validate the model.
Finder
Image and Reference Stars
Figure F01. Finder image, 15.6 x 10.5 'arc, north up,
east left, showing V1134 and 8 stars used for reference.
My unfiltered telescope system has a passband centroid that closely
resembles the g' band passband centroid, so I process unfiltered
images using a set of g'-mags.
Star ID
|
g'-mag
|
V1134
|
(18.400)
|
1
|
12.506
|
2
|
14.226
|
3
|
16.506
|
4
|
13.860
|
5
|
15.199
|
6
|
15.001
|
7
|
15.856
|
8
|
15.456
|
Why I Like This Star
System
A longstanding unanswered question in astronomy has been "What
fraction of stars retain a planet during their transition from main
sequence to WD?" The discovery of WD 1145+017 (Vanderburg et al.,
2015) established that asteroids can be retained (or possibly they
are planet cores), but the fraction for asteroid (or planet cores)
retention is still uncertain. The first suggestion of a planet
orbiting a WD was published by Manser et al., 2019. However, the
evidence for this is indirect (just periodic absorption by a gas
thought to come from a gas giant planet). Rumor has it that a more
direct detection of a giant planet may be in the works. V1134 is
most certainly either a "massive gas giant planet" or a "low mass
BD." In either case this means that at least something close to
planet mass is able to survive the transition from main sequence to
WD, and this bolsters hope that smaller sized planets are also able
to survive. That's why I like this star system.
References
Rappaport, Saul, Roberto
Sanchis-Ojeda, Leslier A. Rogers, Alan Devine and Joshua N.
Winn, 2013, arXiv
Manser, Christopher J., Boris T.
Gansicke, + 30 others, 2019, "A Planetesimal Orbiting the
Debris Disk around a White Dwarf Star," arXiv
Vanderburg et al, 2015, "A
Disintegrating Minor Planet Transiting a White Dwarf," Nature,
2015 Oct 22, arXiv:1510.063387
________________________________________________________________
This site opened: 2020.05.10