Finished! Looks like this project is out of data at the moment!
This project is no longer active: please visit Space Warps - DES Vision Transformer.
FAQ
Science
Task
Interface
Training
Science
What is a gravitational lens system?
When the light from a distant galaxy gets deflected and focused towards us due to another massive galaxy, or cluster of galaxies, sitting right in the direction of the distant background galaxy, then such a system is referred to as a gravitational lens system.
What does a typical lens system look like?
A typical lens system will show a massive lensing galaxy or a galaxy group surrounded by magnified and distorted arc-like multiple images of a background galaxy. Sometimes the background object is a quasar which looks like a star (i.e. compact circular object). A system with gravitationally lensed quasar will thus show multiple compact objects around the lensing galaxy.
What's the difference between a lens(ing) galaxy and a lensed galaxy?
A lens galaxy or a lensing galaxy is the foreground massive galaxy which warps the space around it and causes light to bend around it. A lensed galaxy is a distant background galaxy from which light rays emerge and get bent or deflected due to the lens galaxy before arriving on earth.
It is a common shorthand to refer to the lensing galaxy as the 'lens' and the lensed galaxy as the 'source'.
Why are most background galaxies blue and most foreground galaxies yellowish red?
The theory of gravitational lensing suggests that there is an optimal distance at which the background source need to be situated for a massive foreground object to produce the gravitational lensing effect.
We find from many imaging surveys that, as we look back in the nearer half of the Universe there are more elliptical galaxies which are not only much more massive than spiral galaxies but are also dominated by older population of stars. These stars emit light closer to wavelengths that appear yellow and hence, the elliptical galaxies appear yellow. Their larger masses also make them more efficient in deflecting light from background galaxies or quasars to produce multiple images.
We also find that there's a rise in the population of spiral (or star-forming) galaxies at a distance when the Universe was one fourth of its current age. Incidentally, this distance happens to be the optimal distance for us to see the strong lensing due to the foreground ellipticals. And, since the spiral galaxies are dominated by young stars which have hot temperatures and emit light in bluer wavelengths, most of the lensed galaxies appear bluish but there can be exceptions.
Why are gravitational lens systems interesting?
Gravitational lenses have numerous applications in astrophysics. Most importantly, they are used for understanding properties of Dark Matter which is very difficult to probe by other methods. Since lens systems also create magnified images of the distant background galaxies, we get a detailed view and hence, better understanding of these distant and faint galaxies which would be impossible otherwise. Other interesting applications include measuring the age of the Universe, finding exoplanets, etc.
Why do you need my help to look for gravitational lenses?
How a lens system will look like in the sky depends on several factors relating to properties of the foreground and the background object. There is enough complexity in these lens systems that the process of detection is extremely challenging to automate completely. One of the major problems is contamination from lens-like systems e.g., spiral galaxies, ring galaxies and other artifacts in the images that mimic arc-like features. Human beings are better at processing and filtering objects that have multiple levels of complexities and objects that are unique or show unexpected properties. Lens systems fall in this category and we need your help both in discovering newer types of lenses and building better algorithms.
How do I know what's a lens and what isn't?
We have put up some examples of lenses, and lens impostors, in the Field Guide. This can be seen as a widget on the right-hand of the classification page. You can click on the images to read the description. You can also learn by discussing with others on TALK.
Task
What is my main task?
The main task for the current project is to inspect images on the classification page and mark any of the gravitationally lensed features you can spot in the image (See the Science section on the definition of a gravitational lens). Lens systems are rare: most of the images won't contain any gravitational lenses, but a few images will have a lens and in extremely rare cases, there might be more than one lens in an image. So, watch out!
Should I try and be inclusive, or discerning?
We don't want to miss the rare gravitational lenses: if in doubt, mark them. The Spotter's Guide can help you in deciding which images are more likely to show gravitationally lensed features.
Why are there three panels for the same objects?
We show three images for each candidate galaxy, with different contrast/brightness settings optimised to help in the visual identification of the faint lensed features.
What do I do if I see more than one lensed image per lens system?
You should place a marker on at least one lensed feature that you see per lens system, for example, arc-like features or compact (circular) objects called quasars. You only need to mark one of the lensed features for your identification to be registered, but you can mark multiple features if you wish.
Will the images have a lens system always at its center?
The images were selected to all contain objects that could be lenses (like massive yellowish elliptical galaxies, or groups of galaxies) - so it's most likely (although not guaranteed) that if a lens is present, it will be in the middle of the field.
Interface
Where should I put the markers?
OR
I marked a simulated lens but was told that I missed it. What's the problem?
We'd like you to mark any lensed features that you see i.e. you can mark one or all of the curved arcs, or multiple compact (circular) quasars. You can choose any one of the four panels for placing the marker and it will be duplicated.
Sometimes the site will tell you that you missed a simulated lens even after marking it. This may be because the marker is not quite on the lensed object, e.g. the marker might have been placed on the lens galaxy at the center instead of the lensed features around. If you are absolutely sure that there is some bug, please report it on TALK along with the image IDs.
Can I delete my markers?
Yes, if you click on one of your markers, a little cross should appear above it. If you click on this cross, the marker is deleted.
When should I use additional tools?
Many lensed images are faint and very close to the bright yellowish lensing galaxies. You can zoom/pan or invert the color scheme to get a detailed and different view of features near the lensing galaxy and be more confident of not missing a faint lens.
I missed a lens! How do I go back?
The last image you saw is in your 'Recents' list, on your Profile. You cannot modify your classification after you have submitted it but you can click on this image to take it to Talk, where you can tag it for other people to see, and discuss it with them.
I found something interesting! What do I do?
The markers are only for lensed features, but you can mark any image as "favorite", whether you think it has a lens in it or not, using the heart-shaped button. This will appear in your favourites list on your profile page. You can then click on that image to take it to Talk, where you can discuss it with other people or add it your own collection.
Sometimes a vertical/horizontal section of the image looks completely dark. Why is that?
Nothing to worry about. These images are at the edges of the survey region where the telescope did not collect data. Since there's no data, these sections look dark.
I can't see some pages on the website. What internet browser should I use?
We offer support for the most recent and second most recent version of all major browsers.
Training images
What are simulated lens systems?
Simulated lens systems are computer generated images of gravitational lens systems using standard models for a foreground galaxy (or galaxy group) and a background galaxy (or a quasar). These simulated lenses are then added to the real images of the sky observed with a telescope.
Why do the training images contain simulated lenses rather than real lenses?
The Field Guide shows some typical examples of real lenses. Since lens systems have varied properties but they are rare, we simulate them in large numbers using computer algorithms. Such a sample can provide further training for the users. We also need a large training sample to understand what kind of lenses are being found or missed by you. This will be used to analyze the real sample of lenses that are discovered by you. Since we don't have a large number of known real lenses, we have to generate them using computer algorithms.
Can I turn off the training images ?
We found that without the training images comprising simulated lenses and lens impostors, we started to make mistakes. We don't want to miss any real lenses, so we thought it best to keep showing the training images, just in case! As you do more classifications we will show you fewer training images. Understanding how you perform on the training images is also important for interpreting the real lenses discovered on Space Warps: we need to know if the sample is biased because certain types of lenses are easier/harder to discover.
We also describe in some detail at this blog post and these papers Marshall et al. and More et al.
Can I turn off the feedback messages?
As gravitational lenses are rare and you have to inspect very many images, we found that without the feedback messages, we tend to lose focus and are likely to make mistakes. We don't want to miss the precious lenses, so we thought it best to keep the messages turned on!