Lasair LSST review users meeting - Oxford

2023 September 13 at 14:00 to 17:00
2023 September 14 to 09:30 to 16:00
Denis Sciama Lecture Theatre, Denys Wilkinson Building, Keble Road, Department of Physics, University of Oxford

Accommodation: University rooms website, Booking.com has many options. Area north of Denys Wilkinson toward Summertown is also good area (walk or direct and frequent short bus).

Remote participation :

Join our Cloud HD Video Meeting

ID: 939 2662 0483
Password: 251

There will be a dinner Christ Church on the evening of 13th. As you sign up, please add your dietary requirements. If you signed up AFTER 23rd August you must confirm your place at the dinner by sending an email to cigdem.arnison@physics.ox.ac.uk. Those who signed up before are confirmed.

What is this meeting about ?

We have built and released a new version of Lasair, with a new user interface and documentation that continues to work with the ZTF data stream. The alerts from LSST will be quite different in volume, depth, number of filters and alert packet scientific content (e.g forced flux values). In the second half of 2024, we expect to receive and process the first real data alerts from Rubin.

Now is the time to review what we provide and how we can support the community when LSST data stream starts.

We are running a meeting to engage the astronomy community, to demonstrate the capabilities of Lasair, and to discover what scientists want from the alert stream. The meeting will include:

• Introduction to Lasair and tutorial, including: reducing millions of alerts to the ones you want; intelligent crossmatch with the big catalogues; making a watchlist of your favourite objects; automating response to interesting alerts; sharing filters, watchlists, and classifiers; annotating the Lasair stream.

• Contribute and collaborate by presenting your ideas: share a scientific case study of mining the LSST alert stream and/or run a discussion section on how your science can benefit from the alert stream.

• The LSST:UK project is not only about rapid alerts but all aspects of LSST science. Mining our archive of alerts; how we link to the Science Platform; how can the rapid alerts and data releases work together effectively?

We welcome experienced users and new users at meeting and there will be plenty of opportunities for interaction, small group working and discussion and one-to-one help.

Will the meeting be hybrid ?

Yes, we will provide a remote connection to the meeting - see above

Final programme (all times BST = GMT+1hr)

Wednesday September 13th

Morning to lunchtime : arrival and informal meetings with the Lasair team (Lasair team will be available from morning for any requested discussions).

13:00 - 13:50 Lunch (provided)

15:30 - 16:00 Tea

All 20min talks = 15min talking and 5mins Q&A

Some suggested topics for discussion time : communication between Lasair and TIDES (transient classification),

Evening - dinner in Christ Church college, in the Freind Room.

Thursday September 14th

All 15min talks should aim for 12min + 3min discussion/change over.

11:00 - 11:20 Coffee

12:45 - 13:45 Lunch (provided)

14:45 - 15:15 Tea

16:00 Finish

How will LSST alert data differ from ZTF ?

Alerts will be released 60 seconds after shutter closure

1 yr of history (rather than 30 days) in the first alert packet.

In the 2nd and subsequent alerts, there will be forced photometry data, over 1 yr (SJS has to check if it’s 1 yr or full history). You will get 2 plots - one will be the magnitudes for >=3 sigma detections,

Image stamps will only be 6 x 6 arcsec.

Access to large image stamps of the target data are embargoed for 80 hrs (even to those with data rights). Access to the larger stamps from the reference images is not under the 80hr embargo but they are only available to data rights holders.

Forced photometry will continue for all real objects, after they fall below the 5-sigma limit - we need to check for how long this will continue (Eric Bellm’s document and information)

Some Discussion starting points for the breakout groups

User requirements for Lasair - what’s missing that your science needs

How to alert on interesting sources early - ML for lightcurves and classification

Cross-match catalogues : from watchlists to catalogues

Do you need large cut-outs or is 6 x 6 arcsec sufficient ?

How quickly do you need forced photometry ?

Is documentation sufficient to learn how to use Lasair, and SQL and make queries ?

What are the biggest barriers to writing a filter query to find your type of object ?

What are the biggest barriers to writing a filter or query to find candidate lensed SNe within the Lasair system ?

Questions raised

Is the forced photometry 1yr or full history ?

Saturation limit - what happens with m < 16 sources, are they even flagged ?

Real time x-match against Euclid - possible ?

How will a resolved PSF, such as a lensed SN be treated in the LSST real-bogus classification and will it be rejected with a simple RB cut ?

Can we fill in the non detections without a follow-up real detection ?

Not all users are knowledgable about the Cassandra vs MySQL database difference and that the data are in the database and how to access those values.