- All right, so our next talk is by Jessie Christiansen and we've got the report or a report from SIG#2. So very excited to hear this. Jessie, why don't you take it away?

- Thanks, Josh. Well, first I wanna start by saying something that I said in the chat, but I wanted to say out loud, which is that I'm really excited that this funding is available for seeding some citizen science projects. I remember a few years ago when Exoplanet Explorers was like up and running it off to the races and Paul Hertz was like, "This is amazing, what do we need to do to do more citizen science?" And I was like "Money, that's what we need." So I'm really glad that this is finally happening. But to jump forward, SIG#2. So yes, I'm here today to report on the progress of SIG#2. So as was described this morning, the SIG is the Science Interest Group, which is one of these long-term projects where people in the community come together to think about something and in this case we're thinking about Exoplanet Demographics. So I'm one of the co-chairs, Michael Meyer is my other co-chair and Gijs Mulders and David Bennett form our steering committee. Going forward. So broadly, our original goals were to extend the work of SAG#13 which did this literature meta analysis of ADA earth values. And to broaden that, to both see if we could refine ADA earth and also to broaden the parameter space over which this matter analysis could be done. And the SIG tried to, start attempting this by bringing together people in the community, from different techniques, from different projects in different parts of parameter space, to start to see how we could connect all of these together. We hit some roadblocks which actually turned into the report that I'm gonna tell you about today. In the meantime, we've gotten through some things, so members of the SIG work together on a white paper for the Astro2020 to cradle. We hosted demographics mini-symposium at the January two year W S two years ago, which ultimately became the larger Exoplanet Demographics conference. We created the demographic gap list that we presented at the last January W S. And now at this January, what would have been the W S we're presenting our report that we've put together. This report covers the next two bullets in this list. So one of the roadblocks we needed. We hit was that we realized we didn't have all of the information we needed. All of these different exoplanet surveys had published all of these amazing, incredible data. And a lot of it was unable to be used by other teams to do meta analysis or to do or to redo analyses even. So that was one of the stumbling blocks we hit pretty early. It was very difficult to actually find the data we needed. And that out of that sprang this idea that we should tell people what data we need. Like the demographics community is a, is a subset of the exoplanet community. So there's a lot of exciting detection and characterization and evolution stuff that's happening. And demographics has a different set of needs. So what we did as a group was to come together and write this report about what the demographics community needs you to do in order for us to use your data. And of course you want us to use your data. We want to cite your data, we want to get the best analysis we can. So that's what I'm going to mostly present today is that report. And then I'll talk at the very end about the next task we're thinking of working on. Okay. So I'm happy to announce today that our report is available for community feedback. And the report is entitled, "Enabling Exoplanet Demographics Studies with Standardized Exoplanet Survey Metadata" which was praised as perfect NASA E when I proposed it to try and capture all of the moving bits and pieces. What we find is, is basically what I said before, which is that demographics re-analysis or meta analysis are stymied at the moment by the lack of survey metadata. So this puzzle report presents a list of the data and products we need, that we would love for you to do. That would be very valuable if you included them when you were publishing your surveys. So the report has a few intended audiences. One of them is survey architects, people designing their exoplanet discovery surveys. You might not be thinking in five years time that someone might want to do demographics with your data. But if you're thinking about it in advance, you can set up the products that you need so that you don't have to think of how in five years you've already thought about them. It's aimed at journal offers people who are writing up their survey. We went and used this telescope and found 10 planets. That's really cool, you know, what could you also provide with that so that we could use those data to do demographics. It's also aimed at referees, we're providing a handy-dandy list of things that you could be checking off if someone publishes an exoplanet survey. What could they do to make it useful? Here's the list of things they could do to make it useful, something for you to refer to when you're refereeing papers. And finally, it's aimed at funding agencies. We are asking people to do work, right? We are asking people to make products available, to document them, to get them hosted somewhere and archived. That's not free, that's all effort. And so this is also aimed at funding agencies so that when people put in a proposal and asked for effort to do these things, this is why, this is why we want all of these products to be made available and documented in a certain way. So this is to help justify. You will have to ask for resources because it's gonna take resources. Okay, so to get through into the report, this is how we created the report. So for each of five detection, techniques, transit radio, velocity, micro-lending imaging, and astrometry we had small teams go off and think about what products they would, they would immediately think to make available. So the trans people went off and thought about the transit service. So the five teams came back and they had their preliminary lists of things that they thought would be useful. Then we basically had like a, you know, a thunder dome, all of those lists came together and all of the other techniques were like, ah, you've said this, but actually what we need is this, and what we need is that. So we basically tried to imagine if I had this list of products from the transit technique, could I incorporate it in my rate of velocity analysis? And so we did several rounds of feedback, trying to refine this list. We from those lists, we produced a full report, which was available for open for internal feedback by SIG#2 over the last couple of months. And finally, basically I drew the line in the sand and I was like, we're done, we're making it available. So now the final report is available for community feedback. Okay, how is the report structured? So for each of the five different detection techniques, what basically emerged from the conversations is that there's two tiers of data products. One is the minimally useful stuff that you should make available for us to even attempt a rudimentary incorporation of that in our analysis, like this is the minimum that we could make use of your data. And these are often things that you already have in hand, but you might just not think to like, make that available as part of the publication. And we're saying, please do. A second tier is typically things that involve more effort. Often they're intermediate products, things you've already put effort into making, but think this intermediate product is useless, no one will want it. We want it. So then there's this list of tier two products, which are things that would make it very easy and very valuable to us to incorporate in our demographics analysis studies. So those are the two kind of tiers of data products. Each detection technique that we've broken into three data types. So the Stellar sample description, the survey description and the planet catalog description. And so you'll see this structure kind of throughout for each of the five detection techniques. There's the two tiers. And within the two tiers, there's the three subtypes of data. One finding I particularly wanted to highlight from the report that came out of creating these lists is for each of the different detection techniques, they measure different things, right? So transit measures a planet radius to Stellar radius and radio velocity measures, you know, mass ratio across along the line of sight. So in order to map all of these onto the same parameter space, usually you do some kind of transformation, use a mass radius relation or a luminosity mass relation if you're directly imaging. So you've done something to transform them all into a common set of parameters. Our ask is that you please publish them all in their original native parameters as closely as possible. You can also do the transformation and have that in the paper as well, but it's important to have the native units as close to the observable properties as possible. And then when you transform them, include the information that the assumptions that you've made to transform these observed properties into the derived parameters. That makes it so that someone else can come along, make a different assumption, use a different mass radius relation, use a different evolutionary model of luminosity for direct imaging and, and still use your data. Still still use your survey in their analysis. That's just generally a good idea, but for us, it's very, very useful. Okay, here's an example, just the transit survey. I'm not gonna go through all of this, but what I basically wanted to highlight with some examples of these two tiers. So for instance, in the transit survey data, a tier one product, a basic, we just need this minimum information for the Stellar sample is a description of the Stellar sample, including target selection criteria and number of stars. So if you say we surveyed a hundred thousand stars that fell into these criteria, like this temperature range, this logic et cetera, et cetera. But that's, you need to include that. Ideally what you would include is a full list of the stars, including their fundamental properties. And if you've done any derivations, those derivations and the assumptions you use to make those derivations, right? So that's kind of the difference between tier one and tier two. Another example for the survey useful, minimally useful product that we need to incorporate. It is the detection and vetting efficiency of the survey as a whole. So for instance, for Kepler, the pixel level transit injections that we did, basically, they were very, very detailed and they were like star by star. But then, go over to tier two. What we want is, you know, take one star and inject a hundred thousand things of it. Those were the flux level transit injection from Kepler. So the pixel level transit injections gave you the survey averaged. We put in one thing per star on average, this is the detection efficiency that's your tier one. Your tier two is we took each star and really explored it's detection sensitivity. And so like for instance, for direct imaging, these are the like tongue plots that you see for a detection sensitivity. But that's kind of the difference between the tier one and tier two for detection sensitivity. One is the survey and one is pet target, star by star. Okay, so the PDF has been uploaded to the ExoPAG SIG webpage. So you can either Google ExoPAG SIG, and you'll get there. Or you can go to this website, this presentation is in the Google slot, Google folder. So you can grab the link from there if you want, or just Google ExoPAG SIG and you'll find it. The comment period is, is almost two months. I chose this because it's before our March SIG#2 meeting so that we've got time to get all the comments and start incorporating them. So that's two months for you to look at this. You can send feedback to myself or to the SIG team mailing list. And this document is really for the community, right? Like read this and make sure it does what you need. We probably haven't captured every single person's point of view here. And, you know, we certainly haven't. But the only way we'll do it is if you read it and give us feedback and like read it from the point of view of you're publishing your survey. Have we explained in enough detail what we're talking about? Do you know what we mean when we say per target detection sensitivity? You know, if you have questions or if you want more clarification, or if you think there's something missing, or if you think that this is too much, we had a lot of discussion about at what point we were asking for too much, you know, especially, you know, a grad student working on a project for four years, and then we're dumping like six extra, you know, make these products available. That's kind of where this two tier structure came from trying to, you know, balance our needs versus the needs of people trying to get results out. So please read it and let us know if it captures what you need, or if you have any questions. That'd be really useful. Cause we want this document to be useful. Okay, so that's our report that we've been working on for the last year or so.

- About two minutes left.

- Great, cause this is the last slide. So the next thing that we've been thinking about is a Kepler data challenge. So what this figure here on the left shows is estimates of ADA earth or gamut earth rather sorry, estimates of gamma earth through the years. And you can see that they've spanned several orders of magnitude. They have settled in the last few years, which is good. So over the last two years or so, we've kind of landed between 10 and 50%, but we haven't as yet really done an analysis where everybody's using the same stars, the same planets, the same assumptions about reliability and different methodology. So the idea for the Kepler data challenge is that we would at so a sub-team would basically create a synthetic Kepler sample based on some underlying assumptions about the planet population. Keep all the stars and planets the same to everyone, make all the assumptions the same that we can about false positives and multiplicity and calculate. And then have everybody go off with a synthetic data set and calculate what they think the underlying planet population is. So similarly to the data challenges we've seen before from the direct imaging and a rate of velocity, this would be a capital data challenge. So that's our next task that we'll spool up on this year. So keep your eyes out for that. If you're interested in being part of the data challenge, it'd be really interesting to see, you know, which assumptions in which methodologies make the biggest difference in the answers. All right, thank you everyone. That's the report from SIG#2.

- All right, thank you very much, Jesse. There are still no, there's no questions in the Q and A, so we've got a few minutes, so feel free to type some in. If I might editorialize, I was involved in some of the SIG#2 discussions. I think it's an amazing effort. Jesse has been putting together some really impressive work on doing this. And I think this is going to be an incredibly important document for anyone who's going to be doing any kind of exoplanet discovery going forward. So everyone please read it.

- I see Jake Clark has a question in the chat here as compared to the Q and A tool. Should these best practices also be incorporated into single system plan papers? That's a really good question. So I would say that there are some generic findings that should certainly be included such as the one where, you know, keep the units published the units as close as to the observables as possible, and also include your models to con so that other people can use them. In general, it would be very difficult to incorporate single system discoveries into logic, meta analysis. If it was a single system as part of a larger thing, like there was a new TKS paper today, the test Kepler survey. If there's a survey that's, you know, bringing out single detections, then I would like to see information in each one of those such that, that at the end, you could have a survey paper be like, and here's how you can do this. That would be great. But in absentia of largest survey, a lot of these probably aren't applicable.

- And we have, we also have this question that's now on the shared screen from Jennifer asking about what these data tiers align with the SPD 41 NASA information policy, which is about sort of open and public data.

- I will have to look into that cause I don't know that telephone number off the top of my head. I'd have to go and read it and see if it did, but certainly in this community comment period, these are the sorts of things that would be useful, if there are other external things we should be pointing at and that kind of thing, this is all useful to know. So right now I'm gonna write down SPD 41. And that's the sort of thing I can follow up, thank you.

- Let's see any more questions.

- Well then I'll take a minute to thank everybody who contributed to the report, the tiger teams, you know. Anybody getting anything done in the last two years is wildly impressive to me. So thank you to everybody who wrote text and looked at everybody else's texts and really thought critically about what we needed for this. I appreciate everyone who contributed to the report. And so that's why I really hope that people go off and read it because a lot of people worked on it. And I think it'll be really useful.