Clarissa Do O?
Constraining the Formation and Orbital Architecture of Directly Imaged Exoplanets

Hi everyone.
I'm really excited to be here today.
I'm Clarissa.
I'm a final year PhD student at UC San Diego, working with Professor in Quin Konopaki. So today you're going to get to hear about my thesis. Almost all of it. On constraining the formation and orbital architecture of directly image dexoplanets.
So of course I have to start with this plot that we've seen a few times today and probably will see more of.
Throughout this conference, so for my thesis, I focused on understanding the population of directly image dexoplanets, which you can see here in this slide is towards the longer period range and also the higher mass range. And this is mainly due to observational constraints. We can only detect to.
Their telescopes, young and self luminous and widely separated gas giants.
At least for now.
So because we are looking at very.
Systems, one thing that we can try to understand is how exoplanet systems form. So as William mentioned, most theory suggests that planets form from protoplanetary disks, and as time evolves, we eventually see the fully formed planet system.
But the question is, how does this process actually occur and how can we use the current data that we have to understand the process?
So for my work, I've mostly focused on understanding formation using the orbits of exoplanets, so the orbital parameters of these planets can really.
Tell us about how they formed and dynamically interacted in the past, and one particular parameter of interest is the eccentricity, which is the parameter that defines the shape of the orbit.
And the reason why is because essentially, if a planet forms from core accretion, it would slowly form from the slow build up of dust in and of gas in the disc.
And you could expect orbit to be near circular, so we've lower eccentricities, while if the planet forms through gravitational stability, it could have a more possibly a brown dwarf like formation, and the gas giant would formally form really rapidly in the disc and could potentially form planets with.
More elliptical orbits.
So.
Mind, throughout my thesis I was very lucky to take advantage of several different tools to answer this question about planet formation. I have done some simulations about planet evolution in a disc.
I'm not talking about this today, but if you're interested I will be talking about it on Monday at 2:00 PM during my dissertation talk.
And then I also got to work with some current observations, mainly with the Keck telescope. And I've also been part of some instrumentation efforts.
So I'm right here in the middle, and I'm gonna start with my observation. Work 'cause. That's what I've mostly worked on for my thesis.
So one thing that I got to do that was very exciting was to work on the population level eccentricities of exoplanets and brown dwarfs using observable priors. So to kind of breakdown what all of this means, essentially we're trying to answer the question if exoplanets, so objects that.
We consider are below 13 times the mass of Jupiter and brown dwarfs.
So the population between planets and stars has similar or different formation processes, and we essentially need to look at their eccentricities at the population level. So looking at several objects to understand this. However, most orbit fits in direct imaging.
Use what we call relative astronomy.
So that's just the relative position of the planet to the star over time. And because of the really long periods.
So remember on the right hand side of that exoplanet plot.
We essentially have a very short orbital arc.
These planets, compared to their true orbit, so they're often hundreds if not thousands of years in period.
And we have about a couple of decades of observations and O'Neill at all 2019 essentially found that if you apply uniform priors to your orbitfits combined with this under sampling of data, you could potentially see some biases in your orbital parameters, with the eccentricity being a vict.
Of this and on the right hand side you can see the differences.
Or this mode of higher eccentricity, and essentially they developed this observable base prior, which is a different approach to priors that aims to mitigate.
This bias, so with that essentially I fit for the orbits of 21 directly imaged companions.
So that's exoplanets and brown dwarfs.
So on the left you can see their individual eccentricities on the Y axis and their mass on the X axis, and then essentially compared what happens to the population level eccentricities of for these objects to see if they had similar or different population level eccentricities, and essentially what.
We found which you can see on the right, is that there is.
A variation actually on whether you have similar.
Different eccentricities for these planets and round dwarfs, depending on where you place the intermediate mass objects which are shown in green.
So essentially what this means is that of course we have uncertainties in the mass. So we don't really know where to place the intermediate mass objects, but also the uncertainties in the eccentricities are so high that we can't quite yet constrain whether planets and brown dwarfs form similar.
Oops, OK.
So essentially what we did after that is simulated how much orbital coverage we actually need with relative astronomy.
To have reliable posteriors for eccentricities, and we found that for both priors actually observable and uniform, you need about 15% of orbital coverage.
However, most objects in our sample had less than 10% in coverage, with the average being about 7.4%.
So again, reiterating that we don't yet have enough data to really constrain these eccentricities, here are some examples of additional data completely changing how eccentricities look for some of these objects. So.
Of course you can have new astronomy, which we got new astronomy from Keck nurture.
You can also add relative radio velocities which essentially give you 3 third dimensional information that you don't really get from a stringetry alone.
So that's getting the radio velocity of the planet itself with high resolution Spectra, and you can see the changes for PZ Tel B with just one newest traometry point HD-1160B with one urv point and then on the right.
This is some work we did with new data from Norton and Kapeck that completely changed the orbit of this object. That has a really long name that I won't quite read.
But it was believed to be highly eccentric and now it is moderately eccentric.
So with that, as I mentioned, we're very interested in getting new data and better data. So that kind of leads to my instrumentation work.
So I was part of the Gemini Planet Imager 2.0 upgrade.
So GPI, the original GPI was in Gemini S the after the several subsystem upgrades, it will be on Gemini North and.
The upgrades will lead to an improvement in contrast by about a factor of 10.
By about a factor of 5, mainly due to the improved stability of the system, which has a lot to do with the upgrade that we had for our wavefront sensor, which went from a Shaq Hartman to the pyramid.
So for that I was in charge of testing the ENCCD camera.
So the detector for for the wavefront sensor and also align its telescope simulator which you can see me working on the bench on the left and on the right is just a simulation of how much.
Gpi two data will help constrain the eccentricity of a planet.
And I just wanted to quickly finish up by saying that there's a lot of exciting things coming up in direct imaging and in finding better orbital or defining the orbital architectures of these systems a lot better.
So Gaia Dr. 4 is coming out next year.
The Roman coronagraph had the world's and I didn't put this on here, but also high spec will be monumental for this. So I just wanted to finish up and thank you and take any questions.
Hey, thank you very much.
I was a bit too loud.
We have a couple of minutes for question.
Any questions to Clarissa?
So what is the principal source of error in determining the eccentricity that prevents you from getting good accuracy with seven on average 7% of the orbit? Yeah, so?
There are several issues happening here. So first of all, I didn't mention this, but there are 6 orbital parameters that we generally fit for and the orbit fits are highly dependent on that, and they're also highly degenerate, so it's really hard to tell exactly, you know what kind.
Of orbital configuration.
You have the reason. Rv's for example are very helpful is they help us constrain.
The plane of the orbit, which already really helps with that.
So I would say it's that plus the fact that we don't have enough data. So once you have enough data, you would be able to do that.
But we just don't.
We're not quite there yet and that's why different kinds of data like R, vs, or even absolute astronomy is very helpful for for determining the orbit because it eliminates these degeneracies.
Any other questions?
Yep, please.
Hi there. My name is Erica Eagan.
I'm from the Johns Hopkins applied physics lab.
I'm curious, in your laboratory work, what are some of the lessons learned that you've had? If you'd like taking hardware from its various steps of development?
Oh yeah, definitely.
So I when I joined the GPI two team, the design had already finished.
So of course there were a lot of delays from a lot of things.
But yeah, one thing that we learned is to really test.
Every single part of our equipment, so of our you know.
Components. The EMCCD camera, for example, came in with the wrong operating mode and we found this out after a while.
Because, you know, we had to do image quality test to find out what was happening.
And so I would say definitely the biggest lesson learned was to really test in everything as as it came.
And yeah, I would say that's that's.
Let's move to the next speaker.
Yeah. OK.