Rubin2020_PCW slack archives day3-wed-slot3a-early-science 2020-07-28---2020-08-13

Tue 2020-07-28 06:38PM

@leanne has joined the channel

leanne Tue 2020-07-28 06:38PM

@leanne set the channel purpose: Community Preparations for Early Science

Tue 2020-07-28 06:38PM

@Melissa Graham has joined the channel

Tue 2020-07-28 06:39PM

@Ranpal (she/her/hers) has joined the channel

Tue 2020-07-28 06:39PM

@Emily (she/her) has joined the channel

Tue 2020-07-28 06:39PM

@Robert Blum has joined the channel

Tue 2020-07-28 06:39PM

@Tim Eifler has joined the channel

Tue 2020-07-28 06:39PM

@Pat Burchat has joined the channel

Tue 2020-07-28 06:39PM

@Jeff Carlin has joined the channel

Tue 2020-07-28 06:39PM

@Alex Drlica-Wagner (he/him) has joined the channel

Tue 2020-07-28 06:39PM

@Greg Madejski has joined the channel

Tue 2020-07-28 06:39PM

@Keith Bechtol has joined the channel

Tue 2020-07-28 06:39PM

@Peter Yoachim has joined the channel

Tue 2020-07-28 06:39PM

@Samuel Schmidt has joined the channel

Tue 2020-07-28 06:39PM

@Sugata Kaviraj has joined the channel

Tue 2020-07-28 06:39PM

@Colin Slater has joined the channel

Tue 2020-07-28 06:39PM

@Somayeh Khakpash has joined the channel

Tue 2020-07-28 06:39PM

@Eric C. Bellm has joined the channel

Tue 2020-07-28 06:39PM

@Lynne Jones has joined the channel

Tue 2020-07-28 06:39PM

@Federica Bianco (she/her/hers) has joined the channel

Tue 2020-07-28 06:39PM

@Jim Annis has joined the channel

Tue 2020-07-28 06:39PM

@Jeno Sokoloski has joined the channel

Tue 2020-07-28 06:39PM

@drphilmarshall has joined the channel

Tue 2020-07-28 06:39PM

@Amanda Bauer has joined the channel

Tue 2020-07-28 06:39PM

@Jeff Newman (he/his) has joined the channel

Tue 2020-07-28 06:39PM

@Lee Kelvin has joined the channel

Melissa Graham Tue 2020-07-28 06:40PM

There is already #day3-wed-slot3a-early-science

Melissa Graham Tue 2020-07-28 06:40PM

we should close this new channel

Melissa Graham Tue 2020-07-28 06:42PM

Or perhaps, given that there are 29 people in this channel already and only 5 in the other, rename it and close the original.

Ranpal (she/her/hers) Tue 2020-07-28 07:39PM

Let's keep this one delete the other one, check which one is connected to the session on the web site

Melissa Graham Tue 2020-07-28 07:40PM

If we delete the other one, but then change this one's name to be the other ones name, #day3-wed-slot3a-early-science , will the link survive maybe?

Ranpal (she/her/hers) Tue 2020-07-28 07:41PM

Maybe but if not we can grab the link again and paste it

Melissa Graham Tue 2020-07-28 07:42PM

ok i'll delete the other one and name change this one and report back...

Melissa Graham Tue 2020-07-28 07:42PM

well first i will make sure people who added themselves to the other one get added to this one...

Melissa Graham Tue 2020-07-28 07:44PM

@Melissa Graham has renamed the channel from "day3-wed-slot3-early-science" to "day3-wed-slot3a-early-science"

Melissa Graham Tue 2020-07-28 07:47PM

:thumbsup: all done, website link updated as well, and i tested it

Melissa Graham Fri 2020-08-07 02:36PM

@Melissa Graham set the channel topic: Community Preparations for Early Science with Rubin
<https://project.lsst.org/meetings/rubin2020/agenda/session/community-preparations-early-science-rubin>

leanne Tue 2020-08-11 03:14PM

Zoom Connection Details
Day:_ Wednesday, 12 Aug 2020
Time:_ 09:00 HST - 12:00 PT - 15:00 EDT - 21:00 CEST - 05:00 AET +1
Zoom Meeting ID:_ 957 4572 9442
Password:_ 257500
Zoom Link:_ https://stanford.zoom.us/j/95745729442

Ranpal (she/her/hers) Tue 2020-08-11 06:12PM: or this link has the password embedded https://stanford.zoom.us/j/95745729442?pwd=NnZ2Y2pKZWZMcUZhUmpZVDMyVlRtQT09

leanne Tue 2020-08-11 03:20PM

Reminder that we are still accepting questions in advance of this session. You can either ask them here or use the https://docs.google.com/forms/d/1brGcFxjse4HUX4NvV45RZYYQ2yIuIIj8YEZWvJBsM9Q/edit#responses

Nicholas Walton Tue 2020-08-11 04:53PM

Q: What are the plans for coordinating the publication of the science results from the early comm/ sci verification data? Likely better if all the science from the early phase comes out at once in a special issue of a major journal. All takes a bit of organising of course. Timelines?

Keith Bechtol Wed 2020-08-12 02:59PM

Slides for the session are now posted:
https://project.lsst.org/meetings/rubin2020/agenda/session/community-preparations-early-science-rubin

Robert Nikutta Wed 2020-08-12 02:59PM

It asks for a meeting password...

Federica Bianco (she/her/hers) Wed 2020-08-12 02:59PM

257500

Robert Nikutta Wed 2020-08-12 03:00PM

Ah, thanks. Or use this link https://stanford.zoom.us/j/95745729442?pwd=NnZ2Y2pKZWZMcUZhUmpZVDMyVlRtQT09

Alison Rose Wed 2020-08-12 03:15PM

Please define a "template".

Robert Lupton Wed 2020-08-12 03:16PM

The image that you subtract from your new data to find things that have changed

John Swinbank Wed 2020-08-12 03:16PM

https://www.lsst.org/scientists/glossary-acronyms might be useful for some of the definitions.

Merlin Wed 2020-08-12 03:17PM: Showing 1 to 10 of 1,141 entries
John Swinbank Wed 2020-08-12 03:17PM: ...if only there were a search box... :thinking_face:
Matthew (he/his) Wed 2020-08-12 03:17PM: The bane of searching term and entry...
Merlin Wed 2020-08-12 03:18PM: For those who want to browse or search without the web form:
https://github.com/lsst/lsst-texmf/blob/master/etc/glossarydefs.csv
Alison Rose Wed 2020-08-12 03:18PM: Hi Merlin! Hi John!

Alison Rose Wed 2020-08-12 03:18PM

So templates are your reference images. How often are they refreshed? What's that 'cadence'?

brant Wed 2020-08-12 03:19PM

The data products definition document may also be helpful:
https://docushare.lsstcorp.org/docushare/dsweb/Get/LSE-163/

Michael Strauss (he/his) Wed 2020-08-12 03:21PM

How do you make a template "transient-free"? I can imagine a median of multiple exposures....

Robert Lupton Wed 2020-08-12 03:21PM: It's more subtle than that. Not a well-defined concept
Robert Lupton Wed 2020-08-12 03:22PM: E.g periodic variables.
Michael Strauss (he/his) Wed 2020-08-12 03:22PM: Yup, that's the kind of things I was worrying about...
Eric C. Bellm Wed 2020-08-12 03:53PM: https://dmtn-080.lsst.io/ is the relevant technote describe what is implemented today

Kathy Vivas Wed 2020-08-12 03:22PM

Can you use templates from other telescopes in Year 1? DECam for example...

Robert Lupton Wed 2020-08-12 03:23PM: Tricky. Shallow, different passbands (esp. in the red, and non-existant/v. shallow in u ), probably worse seeing
Alex Drlica-Wagner (he/him) Wed 2020-08-12 03:23PM: Tricky, but better than nothing?
Alex Drlica-Wagner (he/him) Wed 2020-08-12 03:24PM: I know some early GW searches with DECam were using PS1 templates
Robert Lupton Wed 2020-08-12 03:24PM: Not obviously better than single-epoch differencing. Depends on your science
mrawls Wed 2020-08-12 03:26PM: Unfortunately templates from other cameras aren't very feasible because camera/filter/optics details won't match well and you'll get loads of false positives. I don't think DM plans to support this at scale, but nothing to stop you trying it for smaller datasets on your own!
parejkoj Wed 2020-08-12 03:32PM: As @mrawls says, at LSST scale, minimizing false positives is crucial to getting any science out at all. Given the instrument/telescope/survey differences, it would be very hard not to be swamped.

Alison Rose Wed 2020-08-12 03:22PM

Question for clarification: one set of templates per year?

Michael Strauss (he/his) Wed 2020-08-12 03:25PM: I think the templates will get updated in a cumulative way with each yearly data release.
Michael Strauss (he/his) Wed 2020-08-12 03:26PM: Basically what Leanne is talking about now...
Alison Rose Wed 2020-08-12 03:26PM: Hi Michael. Thank you.

mrawls Wed 2020-08-12 03:23PM

Many of the template Qs addressed in the doc Leanne linked to: https://dmtn-107.lsst.io

Knut Olsen Wed 2020-08-12 03:26PM

Will the templates essentially consist of the best seeing photometric images? Could be interesting for static science, especially in crowded fields

John Swinbank Wed 2020-08-12 03:29PM: Best-seeing per-band coadds will be available. I don't think that we've committed that they will be the identical to the templates, though.
Will Clarkson Wed 2020-08-12 03:31PM: Has it been decided yet what "best seeing" means in this context? E.g. top-10% from the sample in that location in that filter? Or top 10% globally?
John Swinbank Wed 2020-08-12 03:33PM: Not been decided yet.

Chris Walter Wed 2020-08-12 03:27PM

Will you choose a target area for this, or just have a template for any location with >= 3 visits?

William O'Mullane Wed 2020-08-12 03:31PM: Operationally we have not said exactly how we will run it but envisage using locations with 3 good visits yes.
Robert Lupton Wed 2020-08-12 03:35PM: The details of all of this are complicated. We'll do our best to satisfy everyone, or at least lots of people
William O'Mullane Wed 2020-08-12 03:35PM: yes I should have said "good" where thats a tricky definition
Chris Walter Wed 2020-08-12 03:37PM: I was thinking about whether having a contagious area (as opposed to whenever the scheduler happened to give you >3 visits) which might be sparse is important or not.
Robert Lupton Wed 2020-08-12 03:41PM: I don't think we've thought enough about the scale on which we'll build coadds-for-templates.

David Buckley Wed 2020-08-12 03:27PM

could other wide-field facilities provide shallower templates for brighter alerts, or is it just too difficult to contemplate?

Federica Bianco (she/her/hers) Wed 2020-08-12 03:27PM: very difficult to implement and I do not think. DM is cosidering it
Federica Bianco (she/her/hers) Wed 2020-08-12 03:28PM: but could be done as User Generated DP
Alex Drlica-Wagner (he/him) Wed 2020-08-12 03:29PM: https://rubin2020pcw.slack.com/archives/C017V2GSZ4M/p1597260153041200

Wesley Fraser Wed 2020-08-12 03:27PM

Will the templates be first generated on single high quality images, or only generated once (and only once) three images of sufficient quality are available?

John Swinbank Wed 2020-08-12 03:30PM: The latter. https://docushare.lsst.org/docushare/dsweb/Get/LSE-61 page 22 has the details, if you're curious.
John Swinbank Wed 2020-08-12 03:31PM: (Oh, what a typo... now fixed)

Francisco Forster Wed 2020-08-12 03:28PM

Question about templates (apologies if it's answered in the document above): will we have access to a record of exactly which images were used to generate each template? Sometimes this information is needed to do a comparison between a difference flux and a model.

mrawls Wed 2020-08-12 03:28PM

Templates will eventually cover the whole sky, best-seeing is one potential option we are testing, but it's a hard problem as all these questions are highlighting :smile: It should be possible to look up constituent images that went into a coadd template, yes, and alert packets will contain info about the template also.

Francisco Forster Wed 2020-08-12 03:30PM: Thanks!
Robert Lupton Wed 2020-08-12 03:30PM: All that book-keeping already done for coadds, although the details may change. It's complicated with discontinuous PSFs, depending on seeing-matching decisions.

mwv Wed 2020-08-12 03:28PM

FWIW, if the formal baseline is that we only use templates from DR1, and that's only the first 6 months, Then that means that there is no DIA in the first 12 months of the surveys. But the current baseline also means there may also be only limited DIA in months 12-18 of the survey because the templates for that part of the sky will not have been in released in DR1 and will have to wait until DR2, which may not be released before the end of Year 2 (from my current reading of the documentation).
https://dmtn-107.lsst.io/DMTN-107.pdf
which says on page 5: "DR2 [is expected to be released] towards the end of LOY2 (LSO- 011)"

I understand/hope that there is not an argument being made that we should actually stick to the baseline, but I wanted to highlight the month 12-18 risk as well, as I haven't heard it previously discussed.

Alison Rose Wed 2020-08-12 03:43PM: Michael, what is FWIW?
John Swinbank Wed 2020-08-12 03:43PM: "for what it's worth"
John Swinbank Wed 2020-08-12 03:43PM: Our glossary doesn't have that one. :grin:
Alison Rose Wed 2020-08-12 03:44PM: Acronyms.

mwv Wed 2020-08-12 03:30PM

(detailed modulo about coverage and season visibility as a function of declination... but the basic point remains)

Wesley Fraser Wed 2020-08-12 03:33PM

When should we expect to submit formal feedback from the science collaborations about early science plans?

Wesley Fraser Wed 2020-08-12 03:41PM: Thanks. And yes you said my name correctly.

Federica Bianco (she/her/hers) Wed 2020-08-12 03:37PM

@Nicholas Walton thank you for your suggestion - noted!

Nicholas Walton Wed 2020-08-12 03:40PM: @Federica Bianco (she/her/hers) depending on rubin/lsst schedule it might be possible to time a big bang science release from early data at a AAS winter or summer meeting - that will help with press/ outreach cut through (or just go with a NSF/DOE sponsored Rubin event).
Federica Bianco (she/her/hers) Wed 2020-08-12 03:41PM: that would be great! it is hard to coordinate because of the lack of funding (and I said enough about that already so i wont hammer it further) but this is a good carrot that should help motivate scientists!
Nicholas Walton Wed 2020-08-12 03:42PM: Rubin is high enough profile that it will be in the AAS interest to help - benefits astronomy/ astrophysics as a whole!

Chris Lintott Wed 2020-08-12 03:37PM

Regarding @Nicholas Walton 's comment, if people are reluctant to take on the overhead of coordinating science collaboration planning, AAS journals have a 'focus issue' feature where papers can be added to a particular issue as and when they're published, which gives you a single place to point to without delaying anyone.

Chris Lintott Wed 2020-08-12 03:37PM

I've also just sent @leanne an email that says we (I'm an ApJ editor) are very keen to work with the collaboration to make as big a bang of your/our science as we can.

Federica Bianco (she/her/hers) Wed 2020-08-12 03:38PM: this is huge @Chris Lintott thank you!
Chris Lintott Wed 2020-08-12 03:38PM: No problem! Let me know when you want to have a proper chat - LSST is a priority for us,.

Federica Bianco (she/her/hers) Wed 2020-08-12 03:44PM

many surveys (including DESC) use far more than 3 images for the templates (is my understanding) - what considerations lead to the number 3 as optimal ?

Federica Bianco (she/her/hers) Wed 2020-08-12 03:45PM

(or did i misunderstand that it is optimal?)

Federica Bianco (she/her/hers) Wed 2020-08-12 03:46PM: nice, thank you!

Alison Rose Wed 2020-08-12 03:46PM

bigger better camera? :slightly_smiling_face:

Mario Juric Wed 2020-08-12 03:46PM

Thanks everyone for the excellent presentations!
Q: The mentioned LCR specifies the DM system will support incremental template building capability. But I'm confused if the decision to do incremental templates in year 1 has been made as well (if not, is there a timeline)?

William O'Mullane Wed 2020-08-12 03:47PM: Thats an operational decision not a construction call. But we intend to yes
Robert Lupton Wed 2020-08-12 03:48PM: As I tried to say, this is tricky. You have to worry about moving faint objects, but you have to put different years on the same photometric system

Tom J Wilson Wed 2020-08-12 03:47PM

Perhaps a simple question that someone could just answer offline: how will differential seeing (both within the template and in the differencing) be handled? I guess the template will use similar and good seeing, and some PSF matching/convolution will be used to make the "best" seeing template match the science image? (Sorry, used to differential images with Hubble where this isn't as much of an issue...)

mrawls Wed 2020-08-12 03:48PM: Difference imaging prefers the situation where template seeing matches or is better than science image seeing. But when it's not, our algorithm deconvolves the template to match.
Robert Lupton Wed 2020-08-12 03:49PM: I don't think that's quite right; the desideratum is actually only that the PSF-convolved image (the likelihood image) not be sharper than the template --- but you have to worry about correlated noise.
mrawls Wed 2020-08-12 03:56PM: OK yes but in practice the code enters "deconvolution mode" if the template PSF is smaller than the science PSF. Indeed correlated noise is an issue.
Robert Lupton Wed 2020-08-12 04:03PM: Hmm, it shouldn't, you can't spatially interpolate deconvolution kernels. How much deconvolution are you allowing?
mrawls Wed 2020-08-12 04:36PM: that is a @Gabor Kovacs question...

Robert Lupton Wed 2020-08-12 03:47PM

Yes, that's standard technology

Chris Walter Wed 2020-08-12 03:48PM

For offline answering: For DIA only, if you used a precursor with worse resolution, could you degrade our images to use them? That would be sort of the opposite of a normal procedure when you use degrade a template co-add right?

Alison Rose Wed 2020-08-12 03:49PM

Sorry: another acronym question. What is DIA?

Chris Walter Wed 2020-08-12 03:49PM

Difference Imaging Analysis.

Federica Bianco (she/her/hers) Wed 2020-08-12 03:51PM

I think the main issue is different filter that captures different SED so the overall light info is different

Federica Bianco (she/her/hers) Wed 2020-08-12 03:51PM

more than the image quality (although warping etc is computationally not trivial too)

mwv Wed 2020-08-12 03:52PM

Yes. Stand by for a philosophical digression:
What we're basically trying to do is to look for things in the Universe that have changed from our static model. This model could be a full catalog with stars + morphology that could forward model any scene.

Robert Lupton Wed 2020-08-12 03:52PM

I answered this in a different thread. Yes, and that's one of the strong points of the ZOGY formulation

Federica Bianco (she/her/hers) Wed 2020-08-12 03:52PM

^^^ another acrynim @Robert Lupton

mwv Wed 2020-08-12 03:52PM

Instead we tend to build pixel-based models of the sky and subtract those because it doesn't require or depend on your forward model being correct or able to model all of the features (galactic cirrus?)

Chris Walter Wed 2020-08-12 03:53PM

ZOGY is one of the subtraction algorithms (as opposed to Lupton + All-someone) I can't rememeber the names it stands for.

mwv Wed 2020-08-12 03:53PM

And, using a template and doing masks with the template allows your easily track completeness in the sky without having to include that alongside your row-based forward-model catalog.

Federica Bianco (she/her/hers) Wed 2020-08-12 03:54PM

(yup... but trying to flag barriers to accessibility )

mwv Wed 2020-08-12 03:54PM

But, you could just compare catalogs from single images to a static catalog for that region of sky.

Robert Lupton Wed 2020-08-12 03:54PM

What Michael (I didn't write MWV) describes is usually called "Scene Modelling". I think it's formally equivalent to a choice about template generation

mwv Wed 2020-08-12 03:54PM

Totally agree.

Chris Walter Wed 2020-08-12 03:55PM

Thanks all.

Robert Lupton Wed 2020-08-12 03:56PM

If you don't know what the subtle tradeoffs are between Zackay et al. (ZOGY) and Alard-Lupton you don't need to know. The point is that the answer comes down to details; I'm happy to discuss them, but all I wanted to say here was, "it depends".

Chris Walter Wed 2020-08-12 03:57PM

Also in case you didn't know what @Federica Bianco (she/her/hers) 's acronym SED means. I took this picture near Super-Kamiokande:

Federica Bianco (she/her/hers) Wed 2020-08-12 03:57PM

hah! i was just thinking that! - apologies Spectral Energy Distribution

Federica Bianco (she/her/hers) Wed 2020-08-12 03:57PM

for all listening

mwv Wed 2020-08-12 03:57PM

Wait, I thought it was "Spectral Energy Density"!?

mwv Wed 2020-08-12 03:58PM

(but I'm not a real astronomer...)

Federica Bianco (she/her/hers) Wed 2020-08-12 03:58PM

possibly.... kinda 6 in on 1/2 dozen i guess

Federica Bianco (she/her/hers) Wed 2020-08-12 03:48PM

TVS agrees with that assesment

Federica Bianco (she/her/hers) Wed 2020-08-12 03:48PM

FWIW

Mario Juric Wed 2020-08-12 03:48PM

I wonder about the timeline for the decision, as not having one makes it difficult to apply for grants for 1st yr of science.

William O'Mullane Wed 2020-08-12 03:50PM: good point
leanne Wed 2020-08-12 03:51PM: @Mario Juric On what timescale would you need a decision? ASAP :slightly_smiling_face: ?
brant Wed 2020-08-12 03:52PM: NSF AAG grants submitted in Nov 2020 would run through ~ June 2024.
Chris Lintott Wed 2020-08-12 03:52PM: For me, it would be very useful to just know when such a decision would be made.
Mario Juric Wed 2020-08-12 03:52PM: November 16th is the AAG grants deadline, and it usually takes ~2-4 weeks to formulate/write proposals, so my guess is ~early October.
William O'Mullane Wed 2020-08-12 03:52PM: ok
Mario Juric Wed 2020-08-12 03:53PM: And thanks for considering this!
William O'Mullane Wed 2020-08-12 03:55PM: I had not appreciated the lead time
leanne Wed 2020-08-12 03:57PM: We will take an action to come up with a timeline for a decision (or a decision)

Reil Wed 2020-08-12 03:49PM

Will we offer the user the ability to generate their own templates and redo the subtraction with their favorite template selection?

John Swinbank Wed 2020-08-12 03:52PM: The "if you get enough cycles" is key -_attempting to do this for the whole survey would likely be compute limited.
John Swinbank Wed 2020-08-12 03:52PM: (But all the data and software to make it possible will be available)
Reil Wed 2020-08-12 03:52PM: I was thinking someone is working on their favorite single (strange) supernova.
jgizis Wed 2020-08-12 03:52PM: But to add to Robert's answer, your answer will be after data release while the official ones will generate alerts very rapidly.

Alison Rose Wed 2020-08-12 03:52PM

Yes Robert.

Peter Yoachim Wed 2020-08-12 03:55PM

How old can images be and still be used in building a template, e.g., can the end of year 4 data release use images from years 2 and 3 to build templates? We don't take that many u-band images per year, and even less if we do a rolling cadence or switch to a 60s u-band visit.

Robert Lupton Wed 2020-08-12 03:58PM: I think that it depends. Does low signal/noise and/or more problems with differential refraction outweigh the changing sky?
Lynne Jones Wed 2020-08-12 03:59PM: "It's complicated" .. yes, probably understanding template generation requirements better would be useful (and their associated impacts on difference imaging).
For example, if we have fewer visits so you drop back to simpler templates (or older templates), do you need us to restrict HA ranges more than otherwise?
Lynne Jones Wed 2020-08-12 03:59PM: (in the survey strategy)
Lynne Jones Wed 2020-08-12 04:00PM: I am hoping the answer is not that template generation requirements and the resulting airmass distribution of new visits ought to have some dependence (in the scheduler).
Robert Lupton Wed 2020-08-12 04:01PM: I can see it going that way. But if the first year seeing is 1.2" median it makes things easier for all of us
parejkoj Wed 2020-08-12 04:01PM: I forget, which science collaboration has control of the weather satellites?
Lynne Jones Wed 2020-08-12 04:02PM: Don't wish for bad seeing :persevere:
Robert Lupton Wed 2020-08-12 04:04PM: I may not need to
Eric C. Bellm Wed 2020-08-12 04:07PM: @Lynne Jones for DCR constraints it would be useful for the scheduler to be able to be able to prioritize observations at "new" airmass + parallactic angles using prior history. Ian sent a metric to that effect to Peter
Lynne Jones Wed 2020-08-12 04:09PM: I guess if it's going to have a big benefit for science output, then it's worth including. But it makes things messier.
Peter Yoachim Wed 2020-08-12 04:12PM: I refined the DCR metric a bit with the AGN folks. It didn't look like intentionally going after DCR helpful observations resulted in a big improvement, but it would be good to have other folks look at it and see if they agree.

Tom J Wilson Wed 2020-08-12 03:57PM

What about any science that might be done early that doesn't involve using templates? :wink:

Robert Lupton Wed 2020-08-12 03:59PM: Sure, if your science is isolated point sources then catalogue differencing is fine.
Robert Lupton Wed 2020-08-12 03:59PM: Or if you don't care about variability, of course. Maybe that was your point, in which case my apologies
Tom J Wilson Wed 2020-08-12 04:00PM: This comment was in jest at how many times the word template has been said this past hour, don't read too much into it!
Robert Lupton Wed 2020-08-12 04:00PM: That was my second theory, which came too late
Eric C. Bellm Wed 2020-08-12 04:02PM: It's actually a good point, though, there is likely to be good static science available in the direct image catalogs prior to DR1. But those catalogs are not made accessible in the current baseline...