Welcome! This project recently migrated onto Zooniverse’s new architecture. For details, see here.
FAQ
What is the task I have been asked to do?
You will be looking for optical transients in images taken by GOTO in near real-time. These are relatively short-term astronomical phenomena that change in brightness rapidly, appearing and disappearing again in our images in a limited amount of time. They could be supernovae - the explosive deaths of stars, kilonovae - the electromagnetic counterpart to a gravitational-wave event, or something else entirely!
You will be presented with a set of three images: the "science" image taken recently, the "reference" image taken in the past which shows the background, and the "difference" image which shows what has changed between the two images. By comparing the three images, you will decide if the difference image contains a transient or not. It is essentially a game of "spot the difference"!
How soon after observation is the image available on Zooniverse?
The data provided for this Zooniverse project will be updated roughly every hour (weather permitting). The data from GOTO is timestamped (see the "i" button below the difference imaging), so you can see how recently the image you are looking at was taken!
How do I know if it is a real transient or not?
A real transient will appear in the difference image (on the right) as a bright spot in the middle of the cross-hairs. It will also be present in the science image (on the left), but not the middle reference image.
You can see examples of real and bogus detections in the tutorial and field guide.
What if I get it "wrong"?
These images will be seen by many different people, so don't worry if you think you have made a mistake or aren't completely sure. We take an overall likelihood by combining all the different answers together until we reach a consensus.
Why does this task need citizen scientists?
Currently, we have software that is very good at finding the things it is very sure are real, and that it is very sure are bogus detections. However, we are particularly interested in the edge-cases, i.e., observations where the computer isn't sure what it is looking at. This will help us to improve our algorithms for automatic detection in the future, but also to find the really interesting transients that may not have been picked up!
What happens if I find a transient?
If enough people agree that an observation is a real transient, it will be sent directly to the GOTO team where it will be added to a list for further observational follow-up. We will then analyse these candidates, and share our research and findings with the wider community. Discoveries using citizen science will be credited appropriately.
How will I be credited for my discoveries?
If you are one of the first citizens that discovered a real supernova or kilonova in this project, public credit will be given in our regular newsletters, the project results page, and your name will be included on the Transient Name Server (TNS) announcement. You will also be included in the acknowledgements section of any research paper published by the GOTO collaboration about the object. In order to receive credit, you must be logged into your Zooniverse account when you make the discovery, so that we can identify you.
When you signed up for the Zooniverse platform, you will have had the option to give your real name. If you wish to add or change this, you will need to go to your account settings. If you have chosen to give your real name, this will be used for credits. If not, it will be your username. Please note, we automatically filter out email addresses and profanity from these text strings.
Do I need to comment on the Talk pages for candidates to get credit?
No, you do not need to comment on the Talk pages. Credit is assigned based on the actual classifications (clicking yes/no) rather than on the comments for each subject.
What does the metadata mean?
If you click the 'i' icon in the classification view, you can access some metadata about the candidate you're being asked to classify. A brief description of each field is given in the table below, but for more information you could ask in the Help talk board.
| Name | Type | Description |
|---|---|---|
| ut | integer | Unit telescope that the candidate's discovery image was taken with - goes from 1 to 8 for the 8 tubes on a GOTO mount. |
| s2n | float | Signal to noise (quality) of the candidate detection - higher is better here. |
| tel | integer | GOTO node that the candidate came from: 1 and 2 are from Roque de los Muchachos Observatory, 3 and 4 are from Siding Spring Observatory |
| date_mid | date | Date and time of the image the candidate was discovered in |
| x_position | float | Position in the x-direction of the candidate on the discovery image. Ranges from 0 to 8192 - candidates near the edges have typically worse quality. |
| y_position | float | Position in the y-direction of the candidate on the discovery image. Ranges from 0 to 6148 - candidates near the edges have typically worse quality. |
| recipe_hash | text | Unique identifier for a candidate - internal use. |
| detection_magnitude | float | Magnitude (brightness) of the candidate in the discovery image - lower values are brighter. |
| source_image_identifier | text | Identifier for the discovery image - internal use. |
| image_5sigma_limiting_magnitude | float | Brightness of the faintest sources we can detect in the discover image - lower values are brighter. |
Why are the coordinates of the candidates not given in the metadata?
Whilst we fully understand that some of you would like to do external checks using astronomical catalogs, this is simply not possible as the GOTO data stream is proprietary and we need to make sure the correct processes for discoveries are followed. Giving the full coordinates at this stage potentially allows others to report GOTO candidates without our agreement, prior to when we are ready to announce them to the public. As soon as our real discoveries are made public, we will make this data available on their individual Talk pages and on the results page.
How can we be sure that the science and difference images are the same location?
The set of three images are small cut-outs from the much larger images we actually get from the telescopes. These large images contain hundreds of reference bright stars that we use for localisation and alignment, and our pipeline validates that images are (averaged across the whole image) within alignment to less than a pixel before they arrive in Kilonova Seekers.
Why are there some subjects that are always incomplete?
This is our validation dataset. This comprises a small set of subjects with already known types, and a high retirement limit of 10,000 classifications so they don’t disappear from the project. We are using this labelled dataset to assess the quality of your classifications to build more intelligent ways of identifying the most promising candidates from the project, and improve future iterations of the workflow.
Why do you sometimes upload older data?
Sometimes if the weather is poor, or technical time is needed to update something on the telescopes themselves, we will supplement our daily data with some older data to keep you ticking over. Every day, we only upload a random subset of the data that has been taken the previous night, so this additional back-up data will be a slightly less restrictive set.
Can transients be found in any part of the images?
Our difference imaging works by automatically centring the image cut-outs you see on objects that have changed between the reference and science images. This means that the candidate that has been flagged as a potential transient will always appear in the middle of the crosshairs.