Galen Bergsten
New Stuff in Exoplanet Demographics stellar companions & integrated modeling

The properties of those companions, plus some simulations of the stuff that we weren't sensitive to in order to induce a sort of population level correction effect to re evaluate our occurrence rates.
Here's how that process works.
We started with that original catalog of stars that we had assumed single and said all right, some of these probably have unresolved companions, but we don't know which ones. So we're going to randomly pick them based on observed companion rates, and we don't know what the properties of.
Those companions are, so we're going to randomly draw them based on the stuff that we observed.
And then we're gonna use those properties to rescale our detection efficiency and reevaluate our current rates.
You do that process a couple 100,000 times, and the average answer you end up with is an Eta earth value of about 14%.
Now I don't want you to care about the numbers, but more so the relative increase because that is model agnostic.
So we're suggesting that when you implement this correction for unresolved companions, your Eta Earth could increase by up to a factor of 1.4.
Importantly, that number is actually specific to Earth, because this correction is a function of planet periods and radiation radii, and you can find more about that in the paper that we're hoping to submit sometime in the next couple weeks.
At this point, I'd like to shift gears and introduce another project now in the realm of integrated demographics, which is a fancy made-up word to say that we are doing demographics with data from multiple surveys and multiple detection methods to get better occurrence rates. The idea here.
Is that every detection method has its limits to what it can and can't observe.
One common example is, you know, transits are really only good at studying stuff that's close in direct imaging stuff that's far out.
And then we have techniques like radio velocity and microlens.
Sort of bridge the gap.
So if we ever want to get a holistic and comprehensive understanding of planets over the range of parameter space that they can occupy, well then we have to do it using data from multiple detection methods. So that was our motivation to create our own technique for doing integr.
Demographics and I don't have nearly enough time to explain how it works.
The good news is that if you want to learn some details, you can check out my dissertation talk that's happening on Wednesday morning.
But if you can't make it.
No worries, because it's mostly a bunch of boring old statistical theory that.
In order to enjoy, you'd likely need some combination of delusion, neuro, divergent and a general amount of Stockholm syndrome for the career choices that have led you to this point.
Wait a minute. That's just my CV.
Books aside, I think it's much more useful and impactful to just show you the kind of stuff that we can do with a practical exercise.
So let's look at the distribution of giant planet occurrence rates.
I'll start with some course occurrence estimates from transits and radio velocity, which were the detection methods used by Rachel Hernandez to originally discover the shape of this distribution. And now I'll add in weak data from direct imaging, which we can do now, which pushes this out to about.
100 au.
And so on.
To this plot, I'll add for the first time the kind of model that we can fit to data from all three of these methods.
Simultaneously allowing us to constrain occurrence rates over a much broader range of parameter space and with much greater statistical power than what we can do with one of these surveys on its own. I mostly show this today to say, hey, we got this new technique, let's just do.
Some cool stuff.
There's some exciting work in the pipeline, but also to start garnering hype around the fact that someday, very soon you'll be able to do this too.
We're putting everything we have into a new software package for you to do demographics at home, and there's two big reasons we wanted to do this.
The first is that demographics can support every subfield of exoplanets.
No matter what you do, maybe you're an observer trying to put your objects in context. A theorist comparing your predictions to population trends.
Or you're running yield simulations.
You need a robust input population.
There's always some way the demographics can supplement your work.
And the second reason is that the number of people doing demographics is really small and it doesn't need to be so in an effort to make this field more accessible and support demographics for the range of exoplanet science, we're happy to announce this software package is coming out.
In the next couple months, and if you're someone who might.
Benefit from using a tool like this.
Please hit me up so that we can talk about what you want to do with it and how we can help you get there.
With that, I'll close off with a little summary slide and in the bottom left corner is a box of all the other talks I'm giving this week.
The bottom two of those are on social justice and advocacy talks that I do in the field.
So check out the awesome SAG Splinter session on Monday and the policy work that's happened as a part of that project on Wednesday afternoon.
With that, I'll thank you for your time and I'd be happy to take any questions.
Thank you so much.
We have time for a couple questions.
Jonathan Lane, JPL I don't think statistics are boring.
I don't understand them, but they're very important.
SO 2 quick questions.
One is your first slide went by too quickly and your eta sub Earth, that is, for what population of stars, sunlight or all main sequence stars. This is specifically sun like stars, unlike OK.
Thank you.
And then second question, how will the microlensing survey on Nancy Roman improve these demographics?
Yeah, Roman's gonna be great.
Eventually.
We have to figure out a whole bunch of stuff with microlensing.
There's already people who know how to do microlensing demographics, but Roman's a beast of it's own, and it's really gonna, I think, transform our understanding of planets at intermediate separations, especially hopefully pushing us down to smaller masses, which is currently information that we don't have a lot.
Of.
Hi Daniel.
Yeah. Hello, me, Columbia.
I also don't think stats are boring.
Maybe this is too in depth.
So you could talk after, but when you showed the eta Earth increasing like previously you said that like introducing other stars makes the radius increase.
So I was surprised that you'd get an increase in eta Earth rather than a decrease.
Could you maybe touch on why that is?
Yeah, I'd love to.
So occurrence rates scale inversely with your detection efficiency. The worse you are at finding things, usually the better.
Or the more common those planets actually are 'cause you're not sensitive to them.
So when you drive radii upwards, the detection efficiency contours also get inflated by roughly the same amount, which then drives your currents rates down proportionally.
All right. Let's thank the speaker one more time.