Targets, Alerts & Follow-ups

Target and Alert Team

Team Leaders: Ting-Wan (Janet) Chen (MPE ; from 01/04/2019)

Team members - revised list : Kate Maguire (Trinity College Dublin; was TAT leader until 01/04/2019), Andrea Pastorello (INAF-OAPd; was TAT leader until 01/09/2014),  Joe Anderson (ESO Chile), Cristina Barbarino (Stockholm), Maria Teresa Botticella (INAF-Capodimonte), Regis Cartier (CTIO), Annalisa De Cia (ESO Garching), Michel Dennefeld (IAP, Paris), Nancy Elias-Rosa (IEEC, ICE-CSIC/INAF-OAPd), Morgan Fraser (University College Dublin), Lluís Galbany (Pittsburgh), Cosimo Inserra (Cardiff), Giorgos Leloudas (DTU), Joe Lyman (Warwick), Stephen Smartt (QUB), Mark Sullivan (Southampton), Stefan Taubenberger (MPA), Steve Williams (Lancaster)

Classification and Target and Alert Team duties

Target and Alert Team duty overview

The aim of the TAT is to identify suitable targets for spectroscopic classification with ePESSTO+ during an upcoming observing subrun. Discovery information from feeder surveys is ingested by the Marshall and appears firstly in the 'Inbox'. The duty of the TAT is to prioritise the targets appearing in the inbox in the days leading up to an observing subrun. Usually, the TAT is most active in the days preceding observing nights and one to two days before the start of the subrun. Nevertheless, in the ZTF and ATLAS era more discoveries are made and hence the TAT has to handle the 'inbox' targets also between runs and subruns.

Scheduling of targets for classification (adding them to the 'classification targets' queue) is best done no more than one to two days before the start of the observing subrun. This is because otherwise objects can be added for classification for which more data appears that makes them less interesting and then have to be downgraded but can sometimes be missed. Generally, it is fine to 'snooze' objects (see below) well in advance of the subrun because these objects will reappear in the inbox if further information is obtained. TAT tip courtesy of Stefan Taubenberger: "In the inbox, you can sort the objects by various criteria. What I find extremely useful is sorting by last reported non-detection, because this is the best bet to give you young objects. And with ZTF and ATLAS there are actually plenty with non-detections within the last week."

TAT team should coordinate their work through the PESSTO Slack Channel '#TAT'. If you are not a member of this channel, please email the TAT leader (Chen) to get an invitation.

A general (but not-exhaustive) strategy for a TAT member on duty would be to go through the inbox and ask a number of initial questions about each target:

  1. Is the target bright enough for classification at NTT (approx. 19.5-20.0 mag)? Otherwise can be snoozed. Exception for fainter targets can be done if the redshift is well known or object has a reliable non-detection a few days earlier or light curve might suggest that the target will still rise considerably (at least up to 18.5-19.0 mag)
  2. Does the target appear to be real (not an artefact)?
  3. Is it centred on a point source i.e. likely variable star?
  4. Is it a target without a host galaxy that is close to the Galactic plane (see general tips below) i.e. likely variable star?
  5. Is the object currently visible from La Silla? Using e.g. As noted above, please don't archive targets that are >+30 deg if you are also looking at the northern targets.
  6. It can be difficult to observe transients at the centres of galaxies due to the strong background contamination of the host galaxy. This should be taken into account when prioritising candidates for classification.
  7. Will the target be visible from La Silla for enough time to obtain a good follow-up?

If it appears likely to be a true extragalactic transient, then the criteria defined in Section 2 below can be applied to determine the priority.

The snooze button is there to be used for objects where their current light curve evolution is unclear or they are currently too faint for an NTT classification spectrum (>20-20.5 mag). Snoozing an object means that it is removed from the inbox and placed in the 'snoozed' list. When an additional light curve point becomes available that shows the transient to be rising, the object is automatically added back into the inbox for reevaluation by the TAT team on duty.

IMPORTANT - Cleaning of ‘classification targets’ queue: check the targets in the 'classification targets' list before the beginning of a subrun and downgrade them if necessary. It is likely that most of the targets from the previous subrun will be too old to be considered high priority anymore. It’s no problem if they remain there with low priority as backup targets, but old targets should not have high priority unless there is something very special about them.

Priority classification targets

Some science topics will require a "Must classify” targets (i.e. done as soon as possible since science is very timely). Each science PI can send such a request to for approval by the TAT chair or survey PI. The email list comprises the observers, the survey PI and the TAT chair.

Criteria for selecting targets for classification

Light curve and discovery properties for approved ePESSTO+ science projects available here: Object properties

For nearly all transient events that fall into an ePESSTO+ science group, our aim is to observe them as soon as possible after explosion. In most cases (with some notable exceptions), old declining transients are not of interest to us.

Targets that are suitable for classification are added to the 'classification queue' using the top button on the righthand side of the target. A comment should be included describing why it is of interest (or not) e.g 'Offset from z=0.005 galaxy. Deep non-detection three days ago and rising over past two nights. Currently -16 mag. High priority. ' The priority of the target is assigned on the 'classification queue using the 'flame' symbol dropdown menu on the righthand side of the target ('high', 'medium' or 'low'). 'High priority' for classifications should be reserved for the youngest, brightest, faintest, etc. transients to make it easier for the observers to prioritise during the night.

The ATLAS rates project :  we are aiming to classify all, or most, transients for the "ATLAS Volumetric Rate Survey" (AVS) project. This is aiming at all transients within 100 Mpc or z = 0.024 (mu = 35.0 ; all assuming H_o = 73). Transients after peak are still worth classifying, but at medium priority. Guidelines for the TAT and observers for this classification project (note the young transients or otherwise interesting ones should over-ride the AVS priority 

  • If host has a spectroscopic redshift and  z < 0.03 and current mag < 19.0 then set priority as medium
  • If host has a spectroscopic redshift and  z > 0.03 and the object does not fulfil our other criteria (intrinsically bright, faint, fast or unusual) then put it into archive  
  • if host has a photometric redshift an z < 0.06 and current mag < 19.0 then set priority as medium
  • If host has no photometric redshift and current mag < 19.0 then set priority as medium l but leave in queue.  Move it into the archive when mag > 19.0

Old version of selection criteria for targets can be found here: (Criteria for selecting targets for classification up to mag 20.5)

Latest news

2019 June 07
  • Finding different context (host redshift) from marshall and ATLAS page?
  • The ATLAS pages have a rogue spec-z from GLADE. The Marshall associates with a NED galaxy and z - which you can trace and verify by clicking through to NED. We know GLADE has some errors - always trust a NED association above it, and one you can verify by following the links. Until you hear otherwise, trust the PESSTO marshall as the more reliable source (over the ATLAS pages).

2019 May 13
  • Facing problems with marshall?
  • If anyone finds a problem or issue with marshall, don't struggle along, please add bugs to the marshall issues queue: hereand/or email Dave directly.

2019 Apr 29
  • For the "Spectral homogeneity of SNe Ia in the Hubble flow and a subset of SNe Ib/c" project  (PI: Ariel Goobar), the project PI will add targets to classification list with medium priority. With a typical comment: "Target for ZTF Ia program, please classify ASAP" will help the project having a priority among the medium ones. Transients after peak are still worth classifying at medium priority. 

2019 Apr 16
  • If TAT members are in a pinch due to too many targets ingested by ZTF and both the snooze option and the TAT tips reported above are of no help, please use the LASAIR broker 
2019 Feb 08
  • Cleaning of ‘classification targets’ queue: check the targets in the 'classification targets' list before the beginning of a subrun and downgrade them if necessary. It is likely that most of the targets from the previous subrun will be too old to be considered high priority anymore. It’s no problem if they remain there with low priority as backup targets, but old targets should not have high priority unless there is something very special about them.
  • Top TAT tip courtesy of Stefan T.: "In the inbox, you can sort the objects by various criteria. What I find extremely useful is sorting by last reported non-detection, because this is the best bet to give you young objects. And with ZTF and ATLAS there are actually plenty with non-detections within the last week."

2019 Jan 18

  • There are a lot of ZTF objects coming through so two additional cuts were made in November that removes objects with abs(galactic latitude) < 10 degrees and are fainter than 20 mag. There are still a large number of objects but generally priority should be given to objects with official TNS names. Ones with only ZTF names were deemed less likely to real astrophysical transients than those registered on the TNS.

2018 Sept 17

  • Evolving ZTF light curve data can be obtained from LCO via Just enter the object of interest's name under objectId on the lefthand side and the light curve data is returned. Good for seeing if the object is rising or not .

2018 Jan 03

  • Targets marked as 'NT' by the Marshall contextual classifier (visible under the reference image for 'host info' on the Marshall ticket for each object) should NOT be moved to the classification queue during the period Jan - May 2018.
  • The PUNTS team will prioritise 'NT' objects according to their selection criteria and move to the classification queue where appropriate. 

Selection of the TAT for ePESSTO runs

The dates of upcoming observing runs are available here: Observing Runs. A request email for volunteers will be sent by the TAT chair to about one week prior to the start of the next observing run (generally made up of three subruns). Responses received by email within 24 hours will be included for consideration in the TAT schedule. Note: requests for inclusion in the TAT for future observing runs are also possible in advance of this email by emailing the TAT chair, Janet Chen ( with details of your availability. Preference will be given to people who haven't served in the TAT recently (if they so wish).

ePESSTO Marshall

The Marshall is the interface through which targets are compiled from the feeder surveys, scheduled for classification, classifications are added and follow-up scheduling is overseen. Bugs & improvement/development requests can be reported via the 'bug' symbol on the right-hand side of the top menu bar on the Marshall - this takes you to the github issues page

Note on Northern targets in the Marshall:

By default, northern targets are not visible on the Marshall pages. At the top of the page there is a button “NTT” which you can use to turn on >+30deg northern targets. The reason for this is that we have northern targets from ATLAS and PS1 we want to keep track of, and don’t really want to code up another Marshall. Many of you have access to northern facilities so feel free to take spectra of these, but please do put in a comment and/or classify them with the usual Marshall tool.

  • Do NOT archive something just because it is >+30deg. 
  • Feel free to comment and push them into the classification queue. 
  • Other Users can turn them on/off if they want to with the NTT button 

The Pan-STARRS data points you see are generally public (from PSST), so can be used freely. The ATLAS light curve data points you see (for the northern targets) are not public. You can view the PNG plot to guide you if you want to classify things, but ask the ATLAS team at QUB if you want to use data for any papers/analysis.

Tips and tricks (including understanding the feeder survey webpages)

General tips:

  1. Beware of Galactic transients e.g. variable stars that are generally located near the Galactic plane. They usually appear like orphan candidates without a visible host galaxy. The Galactic coordinates are available on the ATLAS and PS webpages for their transients. For a quick check of the sky position, an image like this can be useful (Credit: D. Coe (STScI), D. Schlegel (LBNL), D. P. Finkbeiner (Harvard), M. Davis (Berkeley))

Survey specific tips:
  1. Pan-STARRS (Example:
    1.  Due to the large chips gaps in PS images, non-detections do not mean what you typically might think. They are calculated from the limiting magnitude for each particular skycell for that night. However, they are also displayed for objects that fall on a chip gap or are masked for some reason so a non-detection arrow does not necessarily mean that a true non-detection occurred on that night.
    2.  Galactic coordinates are given on the right-hand side in the top info panel.
    3.  Objects that Ken Smith is confident are real transients and likely SNe are added to the TNS. If a PS1 object has no ‘AT’ designation, treat it with more suspicion.
  2. ATLAS  (Example:
    1. The ATLAS filters are non-standard. Broadly cyan corresponds to ‘g+r’ and orange to ‘r+i’. More details on the ATLAS filters can be found here: ATLAS filters
    2. The second light curve plot on ATLAS pages is from forced photometry so should be more accurate.
    3. Galactic coordinates are given on the right-hand side in the top info panel.
    4. Comments (added mainly by Ken Smith) are available in the top box. Objects that Ken is confident are real transients and likely SNe are added to the TNS. If an ATLAS object has no ‘AT’ designation, treat it with more suspicion.
  3. Gaia (Example:
    1. Underneath the image, there is generally a comment detailing the Gaia scanner’s impression of the target. ‘GS-TEC predicts’ relates to the Gaia spectrophotometry transient events classifier of Blagorodnova et al. (2014, MNRAS, 442, 327). It uses the blue and red grism spectra to estimate the most likely type.
    2. A low-res spectrum is obtained for each epoch. You can scroll through by clicking on the arrows to the right and left of the spectrum.
    3. The spectra are not flux calibrated. For details on interpreting Gaia spectra, see
    4. Gaia is not a SN survey - there will be CVs, YSOs and other transients there too! Checking the galactic latitude of a candidate is probably a good idea!
  4. ZTF (Example:

Information on follow-up target requests

When a classification spectrum identifies an interesting target for follow-up, the science lead of the relevant science group has first refusal on this target. If they are not interested or not able to lead a full follow-up campaign, others may take the lead following an open call within the relevant science group.

Targets can only become follow-up targets with the approval of the Survey P.I., the Science Board chair and the TAT chair. For each science target, a science P.I. is assigned (and their contact details must be listed on the Marshall). Once this is achieved, a summary of the target, Science Group and Science PI must be sent to by the science group lead or his/her delegate.

The science targets can be added to the 'follow-up targets' list on the Marshall. The P.I. is responsible for scheduling follow-up observations. All requests and their details should be entered via the Marshall (not via email!) including a recent estimate of the magnitude of the target. Details on typical exposure times are available at: OBs information  Since SOFI spectroscopy is very time-consuming, requests for follow-up SOFI spectra must be made to the Survey P.I., the Science Board chair and the TAT chair with a justification of the science case.

There are three classes of priority for follow-up: (i) CRITICAL, (ii) IMPORTANT, and (iii) USEFUL. There has been an overuse of 'critical' in the past, this should only be used when the resulting publication is expected to significantly suffer for the lack of these data. A justification on why the target is CRITICAL has to be provided in the marshall. Once the data have been obtained or the time period of the request expires, the priority will be reset to 'NONE'.

Priority Meaning Usage Example
CRITICAL Essential Objects that are disappearing, or phase is critical, or fast rising/rapid declining CRITICAL (3rd December) ; CRITICAL (first sub-run in December)

IMPORTANT Important but not critical to target a phase or expected change IMPORTANT (third sub-run in December) ; IMPORTANT (during December)
USEFUL Data useful addition good coverage already, data would be useful supplement USEFUL (1st subrun in December) ;USEFUL (once in December, date not important)

Optical light curves of follow-up targets

It is ePESSTO+ policy that optical light curves should not be obtained with the NTT unless the target is too faint for smaller facilities (e.g. fainter than >21 mag). It is the responsibility of the Science PI of each follow-up target to organise the light curve coverage, which should be in place (at least in principle) before PESSTO spectroscopic follow-up commences.

Available options for follow-up time include:
  1. LCO (this is the preferred option)
  2. GROND (2.2m, g'r'i'z'JHK simultaneously, unusual SNe, P.I.: Ting-Wan Chen)
  3. Liverpool Telescope (2m telescope, ePESSTO+ related proposal) - please send email to InserraC_at_cardiff_dot_ac_dot_uk.