- Hi, everybody. It feels very surreal to be talking to you about SIG2 cause the last time I was talking to you about SIG2 within the whole capital, right, what's happening. So, I'm feeling in a very strange head space, so hello everybody. Okay. So I wanted to provide an update on what SIG2 has been up to in the last six months since then. So again, SIG2 is the longterm science interest group on exoplanet demographics that myself and Michael Meyer are chairing along with our steering committee, Gijs Mulders and David Bennett. So a reminder of our goals, the original goal when we started three years ago now was to extend the quite focused SAG13 work, which was, you know, trying to do a survey of the communities' Kepler occurrence rate results, over a wider parameter space. We brought together experts from all of the different detection techniques to try and see how we could extend these results. So this is an ongoing activity. We have monthly meetings where we got six, the latest occurrence rate results. We get people to present their papers, their surveys, and we try and work out, you know, how we can assimilate these results into the current body of knowledge and in particular, moving onto the second thing. One of the things we're trying to do is work out how these results fit together. Obviously ongoing research in the community right now, but I would say the biggest stumbling block we've had since we formed the SIG several years ago, I think we had this grand idea that we would start synthesizing all of this information and come up with new occurrence rates. But what we found is that surveys simply on average do not publish enough data. They don't publish enough accompanying data about the survey for someone else to then easily incorporate that those data into their results. They're trying to do some meta analysis across multiple techniques. We kept coming up against these wolves. So this is the second point, what is needed to do this? And that's what we spent a lot of time thinking about. And I'm gonna present this first draft of our report that we've put together on what people should publish, how can we make this easier for the community. Moving on from that, if we're telling everybody, "You need to publish all of these results with your data in order for it to be useful for the community." that needs to be some way for them to put them. That's a later activity is analyzing where they should all go. And then finally, one of our upcoming hopeful tasks that we'd like to do is to facilitate a Kepler data challenge and I'll talk about that right at the end. Just to remind you what we have done, we prepared a white paper. Well, the Astro 2020 decadal survey, we hosted a mini-symposium on exoplanet demographics at one of the previous ExoPAGs that feels like approximately a million years ago but within fact, just January last year. And we created demographics gap list that we presented six months ago. And the gap list is kind of what led to the need for this report. There were just too many things missing that we didn't have from the different surveys. Okay. So let's talk more about this report that the SIG is developing. I have an exoplanet survey data set. Now what? How can we help you? So the SIG2 report on the publication of exoplanet survey demographics data, what should you publish? Okay, so far what the report has is for four of the major successful detection techniques, transit, radial velocity, microlensing, and imaging. We've made a list of products that you should make available with your data. If you're publishing a survey and you're publishing occurrence rates results, and you would like anyone else to be able to use those results in a larger analysis, then these are the products that you need to accompany that data you have. We've broken it down into two data tiers. We did this for a specific reason, which is that, not every project has the resources, either time or money or people to make every single product available. So what we've done is broken it into minimum data products and recommended data product. Minimum data products are, if you don't have the time or resources, this is the absolute minimum that someone needs to be able to do something else with your data and they'll probably have to do a bit of elbow work. They're gonna have to work with it again. Recommended is if you want everybody to flock to your data and use it straight away, this is what you would make available. This is can have seamless integration with other people's projects. These are what we would like to see. We further broke the data into three types of data just to keep it all separate and be prepared and make it more clean. The first is Stellar sample properties. So these are what stars did you choose? How many stars did you look at? Why did you choose them? And which ones did you throw away? The second category of data is survey properties. How long did you observed for? What was your signal to noise threshold? What's your detection efficiency? Those kinds of information. And the third one is hooray. You did a survey. You found planet candidates, well done. What are the properties of your catalog? How reliable is it? What's the false alarm rate? What are the parameters of the planets that you found and the space you found them around? So those are like the things you should accompany your planet catalog with so that we can use it for recurrence rates. One of the recommendations that I wanted to pull out of the draft and say, I think I'll say it again, but I'll remind people because one of the difficulties in merging results from all these different techniques together is the fact that everyone is measuring planets in their own native units, whether it be radius or mass or mass ratio or luminosity. Our recommendation is that you keep that you publish things as close to native units as possible. Tell us what your assumptions would that you use to get from those native units to something else, but let other people make different assumptions. Don't just present the results already having been transformed into another space that someone would then have to undo if they wanna use a different like mass radius relationship or something and redo. So publish as close to the native units as possible so that other people can make their own transformation. And just a quick note at the bottom, there is a possibility that... Gaddy recommended that we add astrometry. So we're thinking about that. I'm not sure if we have the depth of expertise on SIG2 for that, but if we can assemble a crack team of astrometry folks who wanna add sections to this report, that'd be great. So if you're one of those tracks people, volunteer. Okay, I'm not gonna spend too much time because it is a draft. But just to say, here's an example. So you can see the sort of thing we're talking about. So for the transit surveys, this is very, very heavily based on Kepler, because that's what we know and what we did well. These are the minimum data products, and these are the recommended data products. And you can see each of them is broken into these multiple tiers. This one, for instance, didn't have any additional product catalog properties. But for an example, Stellar sample the minimum is tell us you looked at a hundred thousand FGK stars and why you charge them? What would be better is if you listed the hundred thousand FGK stars with all of the parameters and way you got the parameters from, and then someone else could come along and for instance, choose a subset of those that matched a different set of criteria. Another one, for instance, we see this a lot. The minimum is that you tell us the overall detection efficiency of your survey. Like you just give us an average, that's the minimum, please, at least give us the average. What would be better is if you give us a per target detection efficiency. The broken down, here we get into the nitty-gritty of transits, but broken down basically the function of signals to noise or in planet property, like what is the per target detection efficiency. And then again, someone can do a lot more with that than they can with this average detection efficiency if they care about a slightly different slice of parameter space, it's hard to work with average detection efficiency. But if you give someone the per target detection efficiency, they can calculate what it is for themselves. Okay. So how did we assemble the report? We broke up into small teams for each detection technique and each team went off and thought on their own, what do we think the products that our detection technique needs to squire. And we came back together and we actually just had a thumb to dark where we like all of the teams looked at each other's lists and like worked out what's missing, "You haven't put this, you haven't put this. This should be recommended, not minimum. This should be minimum, not recommended." And so each lift was kind of interrogated by the other three methods to be like, "Okay, this is how you have sufficiently or insufficiently addressed what I would need." So we had several rounds of feedback, as you can imagine to end up with what is now close to the final list of products. And now we're writing the surroundings texts to make it all coherent and the final draft will be available soon. Likely to be circulated, as Josh said to the fact, when she reports for comments, to at least the wider ExoPAG community before being submitted to exit pack as a final report. The audience for this report is number one is ExoPAG, since we are reporting back to the ExoPAG community about what we have found, likely we will recommend a finding that we would then subsequently vote on, which is that people should follow the recommendations in the report. One of the audience is the exoplanet community. This is gonna be the handy guide that you have to say, "These are the products that I need so that someone could very easily use my data." And finally, one potential audience for the report is journals. This could be a handy guide for reviewers to say, "Okay, someone's submitting this paper to publish the survey data. Here's the list of products that we could recommend that you have so that other people can use the data." All right, I've got one minute. So I'll say when this report is finalized, potentially the next thing we would work on are pending reinvigorating the community for interest in this, which I had done 16 months ago is the Kepler data challenge. So basically this is to get all of the groups in the community to use a single data set that we know the underlying ground truthfully. We simulate a set of planets, and then everybody go and calculate occurrence rates. And the idea of being in this entangled, the current uncertainties, are they just driven by the data or are they driven by the methodology? Are they driven by the assumptions that people make? Because so far everybody's always making slightly different assumptions and using slightly different methods and getting a different answer. Okay. You know, Can we converge? Can we get all of these? Can we get all these assumptions and methodologies on one dataset and see what the answers? So that would be what we could move on to next. Again, pending bandwidth. Okay. So that's it from SIG2, thank you. Let me know if you have any questions.

- Great. Thanks Jessie. And thanks for everybody who's participating in this. I'll just add a sort of double-down comment on what Jesse said. Even quoting your efficiencies. You need to at least someone else would need to know what underlying model you assumed. At least for direct imaging. You have to have that model already built in order to calculate those efficiencies. And oftentimes it's hard to read through the paper and get out what even that model was. Even if you're given the mean detection deficiencies, to try to sort things like that out. So it's wonderful to have this resource for the community. I don't see any questions on the question Q and A, but if anyone wants to shout out from the floor or raise a hand, please feel free. It's just, you said we're all bandwidth limited. So a best effort, right? You're not getting paid the big bucks for this. I know that. At least if you are, you haven't told everybody else.

- Well, I'm glad. I wanna thank the leaders of those four Tiger Teams for the work that they do. Again, it was volunteer but I think it's gonna be really important and really help us move this forward if people just have a set of recommendations, they'd be like, these are the things we need and then we go.

- Terrific. Thank you. So thanks both to Jessie and Josh for those wonderful updates.