We know exomoons should exist but so far none are confirmed. This is one the hottest search areas for astronomy. But how can we find them and what can we expect from the future discoveries? Finding out in this interview.
00:00 Intro 01:21 Current state of exomoons 06:67 Interferometers and the GRAIVTY instrument 10:07 Which exomoons can we expect? 22:38 Why didn't we find exomoons yet? 25:05 Interferometers 32:38 What's next? 34:16 Current obsessions 38:58 Final thoughts
📰 GUIDE TO SPACE NEWSLETTER Read by 70,000 people every Friday. Written by Fraser. No ads. Subscribe for Free: https://universetoday.com/newsletter
⚖️ LICENSE Creative Commons Attribution 4.0 International (CC BY 4.0)
You are free to use my work for any purpose you like, just mention me as the source and link back to this video.
We know of over 6,000 exoplanets out
there and they've been discovered
through a range of methods. The radio
velocity method, astrometry, the transit
method, even direct imaging, but we know
of zero confirmed exomoons. There have
been some candidates, but still we don't
have a good enough observation that the
astronomical community can agree that
yes indeed an exomoon has been found.
And this is because it's already
difficult enough to find an exoplanet.
To actually find a moon orbiting around
an exoplanet is just like orders of
magnitude more difficult. But there have
been some really good ideas on how you
might find an exomoon. And so my guest
today has been thinking a lot about
this. So today I talked with Dr. Thomas
Winterhalder. He's a post-doal
researcher at Leiden University and
released a new paper where he talked
about how one of the most powerful
interferometers in the world, the very
large telescope and specifically the
gravity plus instrument on that
telescope can make measurements of
exoplanets with such precision that you
should be able to detect the presence of
an exomoon orbiting around it. So, if
this is interesting to you, here's my
interview with Dr. Thomas Winterholder.
Thomas, we're talking about this in uh
late 2025 right now. How many exomoons
does humanity know about?
>> Well, um unamiguously none candidates.
>> Yeah,
>> sorry.
>> Confirmed. None.
>> Yeah, none. So, um there are some
candidates, but um yeah, there's been
papers kind of, uh talking about these
detections and um calling them into
question. Um but maybe very soon we'll
we'll find one. I guess it's kind of
just on on the edge, just on the verge
of finding one. It could be imminent,
but then again, it might take another
few years, you know.
>> I mean, we're here in the solar system
and there are dozens if not hundreds of
moons around Jupiter actually. You know,
when you think about the whole solar
system, Saturn, right? Like, like there
are hundreds and hundreds of moons in
the solar system, it feels inevitable
that we're going to find them. They're
out there,
>> right? Yeah. That's basically the
argument, right? That's that makes this
whole search so compelling because in a
way, we kind of expect, we kind of know
that they're there. Um but yeah um it's
basically a race uh to find the first
one and then probably they'll be um
tumbling out um one after the other.
>> Yeah. In the same way with exoplanets.
>> Exactly.
>> But but describe for like give people a
sense of why finding a moon around an
exoplanet is like an orders of magnitude
more difficult challenge.
>> Yeah. So basically um you're looking at
much smaller signals. Um we're all
talking about signal to noise of course
in astronomy and um whatever method
you're using be it phototric or
astrometric what we are looking at um
the signals are just way smaller and um
we really have to build the
instrumentation first and now we're kind
of at a point where it might work soon
but um we still have to put it off um so
that's probably why
>> you know my audience is very familiar I
think with Dr. Dr. David Kipping and his
Cool Worlds lab and he just came out
with a big video and talked about how
they tried to use the transit method
using James Web to be able to find an
exomoon around one fairly nice
candidate. There's another paper that we
reported on uh people used the very
large telescope the gravity instrument
on that to try and do an astrometric
finding of an exomoon watching the
wobble of the planet and the moon the
wobble of the moon on the on the star
and made a candidate announcement but
you know obviously this is going to be
met with a certain amount of of
skepticism. Now your paper was more
theoretical. What are you proposing?
So basically um the second paper you
mentioned um that paper basically uses
the method we look at in our paper. Um
so what we're trying to do or what we're
looking into this is basically a
feasibility study um is to detect it
astrometrically. So by the movement or
the the position uh of the planet we
want to kind of tease out the wobbles
that you would expect from um a moon
orbiting a plan or a planet that is
orbited by a moon. So basically if you
um if you take the example of a star
that's orbited by a planet um of course
we we've been doing this um a lot with
um stars and planets. So basically
looking at the orbital reflex motion of
the star as it is kind of orbited by um
the planet and by that movement which is
very small actually um you can then
infer the presence of the planet without
ever seeing the planet. Right? And here
this uh technique basically tries to to
apply this method to like one step
deeper to the planet itself. So
monitoring the position of a planet
relative to its star over long periods
of time or or a certain period of time.
Um taking measurements um again and
again and then trying to see if there is
some non-capillarian or some residual
motion there that can cannot be
explained by two um two objects in orbit
around each other. And that would then
hint at um the presence of a moon. And
this would then kind of man manifest as
this wobble motion or this epicyclic
motion um on top of the actual planetary
orbit around the star. But what you need
to do for this of course is um you need
two things. You need first of all you
need to be able to actually directly
detect the planet. So actually collect
the moon uh the light that's coming from
the planet directly uh so as to make a
measurement of its position relative to
the star. And then also what you need is
you need extreme precision um in that
measurement. So you need to be really
sure, you need to be really precise in
your positional your astrometric
measurement because otherwise um these
these reflex motions uh they're so small
that you would not see them. So I guess
one of the kind of the the twists in
this paper is that we want or we're
looking at the possibility of doing this
not with um a normal imaging instrument
but with um optical interrometry because
this is something this specific field of
astrometry is something um where optical
really excels and where we can use it to
the best advantage um because it's
orders of magnitude better than what we
would get with a normal imaging
telescope imaging instrument.
>> And so I guess you know you just hit one
of my audience's favorite terms which is
intererometry. So explain the setup then
telescope wise on on what it's going to
take to be able to make that kind of an
observation because you're not just
measuring the star, you're actually
measuring the planet. And the downside I
know of interferometry is is if the
object is faint then that makes it very
hard. You get the resolution, but you
don't necessarily get the be able to to
resolve the faint objects.
>> Yes. So, um yeah, I saw that Nick Scott
was on your show.
>> Yeah. The terror array. Yeah,
>> exactly. So, um Charara is one of the
big uh facilities doing um optic
interrometry. The other one um is the
VTI, the very large telescope
interrometer. And this is the one uh we
were looking at in this study. Um so, at
the VTI, we've got an instrument called
gravity. And gravity is built for this.
You can basically say it's built for for
doing this kind of stuff. What it was
actually built for is to look at the
galactic center. And um actually the
group that built it won the Nobel Prize
for the science they did with gravity.
Um that's how cutting edge it is. And um
what we're using it for is actually
doing doing this kind of stuff looking
at planets. And um with gravity we can
achieve insane astrometric precision. So
we're looking at like um on good nights
uh 50 micro arcseconds of of astrometric
precision and that's because of the
baseline we get right we have um four
big telescopes for unit telescopes uh in
paranal so at the very large telescope
um and they are separated something
between um 50 to 130 m in baseline and
uh that's really where this awesome
power comes from because our precision
scale scales with the length of the
baseline and not the size of the
telescope. Um and actually Charara has
longer baselines. Um so in principle you
could do this even better with Charara
but um it is also much more sensitivity
limited and that's the other thing
you've mentioned just now right. Um so
because the telescopes in the array are
smaller um we would not be able to to
see these planets that we're interested
in to actually look for moons. But with
VTI we can because we have four 8 m
class telescopes and that gives us the
sensitivity and the contrast um
capabilities that we need to do this
kind of study. Um yeah so basically
that's the two things we have
sensitivity and we have baseline
>> right you you you have the the
separation which gives you that
resolution and then you also happen to
have four of the biggest telescopes in
the world connected together into this
network and so that gets you to be able
to the ability to look at those at those
faint objects
>> and and I think you know when we think
about discovering planets with
astrometry we're looking at the wobbles
of the star and then the influence of
the planet on the star And in this case,
you're saying no, no, let's look at the
planet and look at the wobbles on the
planet from the influence of the moon,
>> which is which is just next next level.
And so, you know, according to the study
that you did, the calculations that you
made, what are the kinds of moons that
you think are within the grasp of
gravity?
>> So, yeah, that's a good question. Um, I
get that question a lot when I present
this at conferences, actually. Um to be
honest, these would be like with any
other moon detection method at the
moment, these would be rather extreme
cases. Um so we're looking at what we
call irregular moons. Um these are moons
that are thought to form as a
consequence of a collision for example
or um as a consequence of capture. Um
both of which we see in the solar system
as well. So um probably this is also
happening elsewhere. for example, um our
own moon. So, Earth's moon um formed
after a collision event and you've got
Triton as an example of a moon that was
captured and you can see it in its weird
orbit around um around Uranus. Uh no,
>> Neptune. Yes, like it go it goes
backwards from the rest of the moons in
the solar system. Um so both of this we
expect to happen and this means um this
gives us an opportunity basically
because um from a in situ formation so
in the circumlanetary disc u which is
what you would expect to have happened
with the galian moons um that has kind
of an upper limit uh in mass ratio um
between the planetary host and the moon
and these regular moons they are very
hard to find because they are compared
to the planet they are tiny um But with
these irregular moons, we can kind of
it's a kind of trick for us to um to be
able to to at least expect that we
should see um higher mass ratios and
therefore we should be able to to see
them nowadays today with our current
instrumentation and that's the ones um
we are currently sensitive to with
currently sensitive to with gravity. So
these would be moons with math mass
ratios um relative to the planet of um
10 the minus 2 to 10us 3. Um
it depends a bit on the how you apply
the method actually because this is very
similar to what you do with radio
velocities to detect planets. What you
do there is you um sometimes you find
yourself in a position where your signal
to noise of a single epoch so a single
measurement is rather low. But if you
just pile up measurements, if you pile
up epoch, so let's say 20, let's say 30
um measurements um monitoring this
system, you can build your significance.
And um this is the same with the
astrometric method and with our method
of of moon detection. So um depending on
the number of epochs um you can vary
that kind of estimate and um that's also
something we investigate in the paper
but we are looking at these irregular
moons for now.
>> Mhm. And so is it kind of like the
equivalent of of like hot Jupiters were
what exoplanet
scientists found first because they are
just extreme. You've got this giant
planet that's very close to the star. it
is putting a lot of sort of
gravitational tug back and forth and it
just pops out from the radial velocity
method and then it took longer to find
smaller planets, mdorf planets, things
that are and we still haven't even found
what's very similar to the Earth yet.
And so are these irregular moons kind of
the hot Jupiters of exomoons?
>> I think that's a very valid comparison.
It's it's basically um the lowhanging
fruit. So these are the most extreme
cases and the ones we expect to see
first. Um but whether or not they exist
or whether or not they exist um in this
parameter space uh we don't know yet and
that's why we have to look right. Um
we'll only find out if we if we look and
we do have models that that that model
this but we um yeah this can also be uh
these can be data points for these
models right. So um it can in a way
inform the theorists um by eventual non
detections or eventual detections. We
we'll have to see but there aren't a lot
of exoplanets that can be resolved
directly. I mean that in the beginning
you need a very specific type of planet
at this point even with the VT.
>> Yes, for sure. Um so we've got a few
dozen maybe that we can actually look at
with gravity. Um and actually um just as
a side note um even without the kind of
moon signs case looking at these planets
with gravity is is is really special
because um by that astrometry we can
then uh constrain the orbital solution
of the planet around the star and get
masses and um so even if we do not find
these moons having many data points will
give us very precise orbits very precise
masses and that's something we're also
lacking in in expplanet science. But
you're right, um there's not too many of
these planets that that are kind of
viable for this method. But um there's
something big coming and that's the ELT.
So just half an hour away from from
Paranal um the biggest telescope in the
world is being built right now. And with
that we expect um the number of directly
imaged um planets to to rise steeply.
And um then of course if they're at the
right separation we will also be a and
and the right contrast we will also be
able to follow them up with gravity. Um
so it's kind of gravity and the VTI they
will not be obsolete once the ELT comes
around. Um this precision measurement
this this crazy accuracy in in
astrometry will not be um will not be
given by the ELT for example. Um but
there are kind of um symbiosis there is
a symbiosis between the two facilities
for sure
>> right um and do you get any constraints
on the orientation because like the big
downside of the transit method is that
you need to have this alignment of the
star the planet and I guess the moon and
then our perspective from earth. Do you
get a lot more leeway when you're using
this method with astrometry?
>> I guess so. Yes. because first of all it
kind of um there is no condition on the
uh orbital geometry of the planet um
star of the star planet system. So
that's one thing that will increase um
our um well it in a way it will increase
our sample size but then again of course
there are way more planets found um
transiting because of because of the the
facilities we already have and how
prolific they were in in finding these
planets right um but yeah so um in a way
you could say that we are less
constrained but um the truth is I think
the fact that we don't know so many
directly imagable planets right now um
kind of act as a stronger constraint
than than what we loosen by by not
having to go for the aligned ones. Um
but then again um the moon the planet
moon system there's also no constraint
there right so um basically if there is
a moon around one of these directly
detected planets and it's in a regime
where we can in a mass regime where we
and the separation regime where we can
find them with this method then we will
find them. It doesn't matter how
inclined they are because the the wobble
we see the projection of the wobble um
it doesn't matter um if it's face on or
edge on we will see that wobble and the
the astrometric uh signal the amplitude
will be the same. Um, it's true that if
it's very inclined, so if the wobble
would be in one line, basically the
star, the planet would move back and
forth along one line, then we would
probably need more epochs, more
measurements to find it. Um, but it will
be detectable just as it would be if it
were face on.
>> Right. Right. Right. So you're either
getting the planet going back and forth
because of the influence of the moon in
your observations or the planet is
making little little circles
>> as it's as it's going around the the
star and you get the influence of of the
moon either way.
>> Yes. And there is one additional thing
that is probably that that might be
interesting. Um if we have a very
inclined orbit, if it's along a line,
let's say, um we do get these moments in
time um when the planet is exactly where
where you would expect it to be even
without a moon. And that's dangerous
because if we take um if we take an
epoch at exactly that moment um and we
take another epoch um one period after,
then we would always see it at the
position where we would expect it to be.
Um and then we wouldn't be able to to
detect the moon or we would say there is
no moon there even though there is
because we we we sampled our time
series. So our kind of data set is
sampled in the wrong way. And that's
kind of a pitfall that we need to avoid
and it's actually not too tricky to
avoid it. You just have to um vary your
cadence and if you do that then you will
avoid that pitfall. But with the face on
moons um you will always have um a
certain astrometric signal because um
the the planet will always be removed
from the position um you would expect it
to be if there were no moon. So if you
do the model comparison you would always
have some sort of uh of signal within
those epochs.
And I mean obviously
you're getting you know planets are
going to have multiple moons and so you
know do you or or in many cases I mean
you know we have the moon but but
Jupiter and stuff as I said hundreds um
uh do you and I mean right now you're
finding these irregular ones and so do
you think that it's you know you're
going to get one irregular moon around a
planet. Are you going to get multiple
regular moons? Are you going to see
weird variations in multiple moon
systems?
>> I think the expectation is that you
would um see one at least. I mean, this
goes back to this lowhanging fruit
thing. Um you probably do have if you
have one, you probably also have some
minor ones. Uh if you if you're looking
at um a gas giant planet, but the
problem is that those minor ones, they
leave almost no signal. So you would be
um the first systems we would find would
be ones with um one moon that has a
rather high mass ratio. And then also
you get run into problems with
degeneracies when you have several moons
and um they might kind of cancel each
other out in terms of the the signal
they produce. Um and especially they are
drowned out by by the major moon. So I
would expect that the first ones we find
um we see one signal and then some noise
on top of that probably and that noise
might be additional moons but it might
also just be noise. So
>> an exciting idea would be a binary
planet. Would that pop out of the data?
>> Right. So that's actually probably the
lowest hanging fruit. So um because this
is in a sense it's the same system
configuration. You have um a star, you
have a planet, and you have a companion
to that planet. Whether or not that
companion has a mass ratio of 1% or
tenth of a percent or 25% which you
would probably then call kind of a
binary planet system. Um it doesn't
matter for the method. um the method
will detect um these these
configurations if they have the right um
the right mass ratio, the right
separation, if we are sensitive to the
configuration and and then it's really a
matter of definition, right? Um, so you
could talk of a binary planet. Um, but
there's mass ratios in in the solar
system as well between um Pluto and
Charon, if I'm not mistaken, have um, so
it's not a planet anymore, but um, still
dwarf planet and its moon of 12%. Um, so
is that actually a binary dwarf planet
or is it a dwarf planet with the moon?
Um, also the earth and the moon have 1%.
So it's also kind of massive. Um, and
these definitions I think they h I mean
we haven't really come up with a proper
definition of what is a binary planet
and what is a moon but it doesn't matter
to the method. Uh, and both would be
interesting to find.
>> Yeah. Yeah, that would be pretty
amazing. Um, and so, you know, is it
right now that there aren't enough
candidates or not enough time has been
dedicated examining each of the
potential planets to really make a
confirmation?
Do you mean um for moons in general or
for moons?
>> Yeah. Yeah. Like you know, you mentioned
you we talked earlier on that you know,
you said maybe there's like 25ish
planets that have been directly observed
that are sort of within the capabilities
of that. When we know of 6,000 plus
exoplanets, mostly found through the
transit method, 10,000 plus candidates.
Like there's a lot of planets out there,
but you need that separation. this the
you need a fairly large planet going
around a star that's not too bright.
Like there's you put a lot of
constraints and we only know of a of a
handful of directly observed exoplanets.
>> Well, I think in a way that's true. Um
in a way having more planets to go on is
obviously better, but there is also an
argument to be made that some planets
planets are better than others. uh in
the sense that you would expect the
survival of moons around planets to be
different um depending on the orbital
configuration of those planets. So if
you're close to the star, your hill
radius, so this is basically your um
sphere of influence around the planet
will shrink because the the the
influence of the star will increase. And
therefore planets closer to stars, so um
separation wise closer to stars, they um
they would be expected to not be able to
hold on to their planets so well because
they would have to be bound closer. And
if they're bound closer, it's also more
likely for them to be torn apart. Um, so
this is why you would expect planets
with short separations to host fewer
moons or at least host fewer moons that
survive until um we can actually detect
them. And um this is why yes, we do have
fewer planets or directly imaged planets
that might be viable um kind of planet
moon hosts, let's say. um but they might
be more favorable in the sense that um
here we actually expect moons to be
still present because they haven't been
ejected and they also haven't been torn
apart. So um in a way targeting the
directly imaged population might be a
better way to go.
>> Yeah. Um well I'd like to shift gears
then and chat about just interferometers
in general and your experience with the
very large telescope. So give us a
sense. You sort of talked about this
idea that you know these telescopes are
separated up to what you say 140 m.
What's it what's it like using an
interferometer like like that as sort of
an experience as an astronomer?
>> So um the experience of using it is
probably not that different um to the
experience of using um another
telescope, a single telescope for
example um because mostly it's online
anyway. So you would be remote observing
um if you're lucky enough to to get time
on on the telescope. Um but yeah I think
the main difference is um comes
afterwards. So basically um the data
analysis you have to do the pipelines
you run the this data through um are
very different. So um your interfometric
setup or your this technique this this
mode of observation is way more
intricate. I'd say it's um you would
have to go through way more many more
motions to actually pull out the signs
from the data. Um you also you're
looking at different observables. So
you're not looking at um like well
you're not looking at um your
traditional kind of um astronomical
observables. What we're looking at is
closure phase um fringe phases, fringe
amplitudes. So it's a very kind of niche
uh field. It's a very specialized mode
of of of observation, but it obviously
comes with with its own kind of benefits
in the sense that we can access
parameter ranges um that are not
accessible to other telescopes. And we
also, as we already mentioned, um have
precisions. We're working with
precisions, angular resolutions, and
astrometric precisions that are all of
them are actually better. But but when I
think about say a radio interferometer
like with the event horizon telescope
like they were gathering their data
separately at each telescope they had
atomic clocks and then they brought it
in together into the central place and
then and then compiled the data
together. But you can't do that with an
infrared interrometer or with a visible
light. It's got to be done
while the data are being captured by the
telescopes. And so you it's not like you
get data from four different telescopes
and then you have to merge them and come
up with your with your answers. That had
to be done correctly by the telescopes
and the setup. And so is it I'm sort of
imagining like what it's like to be
there and you're like did we get it like
did it did it come together in that way
that we need
>> in the way that you can't just fix it
afterwards and you know oh we messed up
our process okay run the analysis again.
It's it's baked in because it had to be
done while it was captured.
>> Yes. So, um that's exactly right. So, um
compared to radio interrometry, this is
more intricate. It's more like um
hands-on. It's a physical um mode of of
of actually producing interference
fringes, right? So, um yeah, we we
cannot basically the problem is that we
cannot really measure the the phase um
of the light when it's coming in because
it's too quick. Um and you can do that
with with radio telescopes and you can
do that with um arrays of radio
telescopes and this is why you can
record your data and then later on do
this kind of um virtual um interference
and we can't do that. We have to do it
physically. So we have delay lines, we
have optical fibers that transport the
beam um and yeah so it's the
infrastructure is is way more intricate
and it costs more money. Um and it also
maintenance is is a big thing. Um but
yeah um that's the main difference and
um during the observation um you would
have metrics that tell you uh okay this
worked well this didn't work well we we
lost fringes for example um you would
get notified that that happens and then
you would have to go through the motions
again and that does happen actually so
it's not like this is uh rare it's
actually a common thing to happen during
the night and um yeah
>> yeah I I sort of imagine like with a lot
of telescopes you just you just provide
the coordinates, you provide the times,
you provide the wavelengths, you you
know talk about the filters that you
want and then you come back later and
the data are there waiting for you in a
dropbox and you pick it up and and then
start to process it. But it is, you
know, when you're doing your
observations, it's much more of a kind
of a live event and there's got to be
people there who are t turning knobs and
and watching it and making sure that
that your observations are coming
together as you intended because it's a
special mode of these telescopes and
then other times they're just they're on
their own individually looking at
different targets,
>> right? Um there's one more thing to be
said here though. Um most of us um
observe in in service mode. So we
basically program we write little
programs um little bit pieces of code
that actually we send to the observatory
and then that gets executed while most
uh astronomers in Europe at least if
you're observing in Chile um are already
asleep or actually no while they're
already um up again. So the hands-on
actually experience probably doesn't
happen because that's actually done by
the telescope operator. So, um, if
you're very lucky, you get to go down
there to Chile and then you will see all
this what you've just described. Um, and
that's really you have to be really
lucky to to see that. Um, yeah, but
normally it's in service mode because
for the observatory it saves a lot of
time and money to not fly everyone down
there and time in the sense that they
would schedule
>> but yeah, I mean I sort of even imagine
it though with the telescope operator,
right? Like there's a it's just it feels
like it's a different event when you're
just pointing this giant telescope at
one target or when you're trying to
>> coordinate all four observatories and
you're trying to bring the the the lines
together and keep them all in phase like
it just feels like it's a it's a totally
different experience and and nowadays
it's actually one more thing to add here
is nowadays it's way more um of an event
as well because we recently had the
gravity plus upgrade. So this basically
installed um laser guide stars on all
four UT's and wavefront sens sensors on
all four UTS so we can get do AO corre
uh correction um on all the beams and
this first of all increases our signal
noise a lot but what it also does is
when you're there you actually get to
see the four telescopes with um four
laser beams pointing at the same target
and that's kind of impressive. So, um,
you're very right. It's a very
different, um, mode of form of
observation, and if you're lucky enough
to be there, that's amazing.
>> Yeah. And have you seen the mock-ups of
the ELT with with It's got I forget how
many lasers it's got. Like
>> I think it's got eight like Yeah. Yeah.
It really looks like some kind of Death
Star firing into space.
because the the more of those lasers you
can put on the on in the area, then the
more you can tune your adaptive optics
to to capture the the area. And so
instead of just one star
>> or in your case, I guess four stars, but
I guess each telescope is is producing
its own star,
>> you're getting much more. So it's so
sort of taking this technology to the
next level.
>> So what do you think the future holds? I
mean, we're very early on in in this
process of even knowing where exoplanets
are.
Nobody has, you know, the the scientific
community has not agreed that an exomoon
has been found yet. But, you know, as
this plays out into the future, do you
think astrometry is going to be doing
like the lion's share of of exomoon
finding?
>> That's a good question. Um I'm not so
sure about that to be honest because um
this method like the RV method for
example we made that um comparison
earlier as well um it requires
um gathering data over um extended
period of time and it requires 12 data
points 18 data points 24 data points. Um
so um if we want to do that for a lot of
targets then that will just require a
lot of telescope time and um as with
exoplanets we've seen this with
exoplanets most of the exoplanets come
out of survey missions right so um this
will change as well in the I mean it
will still be servey missions but it
will there will be a revolution very
soon when when the Gaia exoplanets come
out and then we're looking at huge
amounts I think astrometry
applied by groundbased um um
observatories on single targets only
will not um provide the line share, but
it might just provide um the first few
ones. We never know. It's um we'll see.
>> Yeah. Almost like how radio velocity
gave us those first planets and then
>> yeah,
>> transit took over.
>> Exactly.
>> Thomas, what are you obsessed with right
now?
>> Um I'm still working on this. Um I'm
preparing what's called a white paper
for um the ESO expanding horizons call.
So basically this is where ISO looks at
what's coming after the ELT and um you
know it would be amazing if we could
build an interrometer with huge
baselines because then that would would
do wonders with our astrometric
precision. Well already orders of
magnitude better than than what you
would get with an imaging instrument.
>> Yeah. Their ideas of building much
bigger interferometers in that facility.
Right.
>> Yes. So, um we're looking at different
ideas. We have um we could add a fifth
UT for example and um and increase our
baseline length um or longer space
length to 200 m more or less. Um that
would be amazing already. Um kind of
change the the um configuration of the
of the of the array as well a bit. So if
you add a fifth one, you could do like a
um one baseline at right angles with the
longest one. um that we have at the
moment and uh that would be very good
for for what we do. But you could also
look at um putting telescopes on other
um hills around in the location. Um of
course you'd run into problems with the
beam transport. You'd have to think of
some kind of fiber system for that. Um
but these are ideas that are explored
right now and right now is the time for
doing that. So
>> So what's the coolest one? Like you're
looking at and go, "Yes, please. That
would be amazing if we could if we could
actually summon the will and the
finances to make this happen. What would
that telescope look like?
>> So that's a that's a good question. I
guess um having um huge baselines with
uh the 8 m telescope class um apertures
along those baselines. Um so you this
would be called I think at the moment
it's kind of dubbed as a kilometric
baseline facility in in in paranal
um extended VTI some people call it um
that would be the coolest one I think.
>> So so you're saying like they could put
another one of those 8 meter telescopes
a kilometer away.
>> Yeah. So the the fifth UT idea that
would be another one of these 8 m
telescopes. Um and that would be kind of
on a on a on a found on a separate
foundation. So um because the the
platform is pretty much full. So on the
hilltop on the mountain top um it's been
flattened of course in the '9s when the
VT was built and then the four UTS got
put there. But now you're kind of
running out of space a bit because you
also have the 80s. So the auxiliary
telescopes 1.5 m. Um so if you wanted to
add a fifth uh UT you would have to
build some kind of uh foundation for it.
Um but yeah that would be eight meters
and um putting that even further away or
adding several of these that would be
cool but um these ideas they have to
contend with um a lot of other um um
ideas like um there's teams um that have
I would say that are a bit further in
the process of developing their ideas.
um the WST team for example at last is
kind of a radio dish and these will all
be contenders for the next facility that
ISO builds um after the ELT. So I think
um yeah it's a contest but um I think we
should definitely put some something
forward in terms of optical
interrometry.
>> Yeah. Yeah. It would be amazing to you
know if someone could put one of these
on another mountain nearby and you could
be separated connect them.
>> Yeah. And be separated by K. I mean,
it's the same challenge that they're
working with the char array. Can you
>> how far can you separate these
telescopes and still feasibly bring
those signals together? And I mean,
infrared is a little more forgiving than
visible light,
>> I guess. Yeah. Yeah. Um so, um they're
looking at they they built this movable
telescope um where they just um drive it
off to um to achieve a longer baseline.
Um but at some point you will have to
look at um so I think they're doing um
vacuum um tubes at the moment to
transport to transport the beam. Um but
at some point you would have to look at
fibers and then you have a lot more uh
other difficulties. You run into other
problems like um um yeah that you lose a
lot of light in those fibers and stuff
like this. Um
>> very cool. Awesome.
>> That would be very cool.
>> Yeah. All right. Well, Thomas, thank you
so much for taking the time to chat and
and good luck with your research and I
can't wait. Let let us know when you
when you find a moon.
>> Oh, I will for sure.
>> Thank you. I hope you enjoyed that
conversation with Thomas Winterhalder.
Uh it's funny. I should let you know
about the timing of this. I actually
originally scheduled this interview with
Thomas back in September and then there
were issues with his schedule and my
schedule and then I was away for a
month. And so during that time, another
group made an announcement using the
methodology that Thomas was talking
about in this interview to announce the
discovery of an exomoon. Now, so it
shows that the the method probably
works. It's still too early to have a
direct confirmation, but uh it won't be
long now, I think, is is where this is
going. And then, of course, this is sort
of separate from the work that's being
done from Dr. David Kipping and and his
group. Now, I'm going to give you some
final thoughts, but first I'd like to
thank our patrons. Thanks to Abe
Kingson, Brayley Griffin, Brian Bod,
Carwin, Chuck Hawkins, Commander Box,
Hin Nielson, Darkfinger, Dave Verb,
David Gilton, David Matz, Evan Pro, Greg
Phy, James Clark, Jeremy Madden, Jim
Burke, Jordan Young, Josh Schultz,
Marcel Smiths, Michael Purcell,
Nordspace, One for Animals.org, Porro,
Ring Kaidu, Richard Williams, Sean
Sergeant, Steven Fer, Munley, Team49,
Telescopes Canada, and Wolf Gang Clots
who support us at the master of the
universe level and all our patrons. All
your support means the universe to us.
If you sort of enjoy this topic of
intererometers and sort of what they can
do for astronomy, I think you'll really
enjoy an earlier interview. And in fact,
Thomas referenced that interview uh in
our conversation here. And this is the
interview that I did on the Charara
array, which is an optical
interferometer that is built in the
United States at the Mount Wilson
Observatory. and just the level of
engineering that goes to be able to
bring the light from all these separate
telescopes together in real time.
They've get a room with carts that are
moving around on tracks so that they can
align the beams coming from these
different telescopes as the object that
they're looking at is moving through the
sky. Like it changes the distance. And
so they actually have to change the
length of the that the light travels to
be able to bring it together and have
this interferometer actually come
together. And it's a very similar
technique that's used at the very large
telescope as well that when you are
trying to align these telescopes to do
their work together like one big
instrument, it is a gigantic challenge
and some of the greatest engineering in
astronomy comes from these
interferometers. So definitely go and
check out that interview if you haven't
already seen it. I think you'll really
enjoy it.