This blog post marks the 31st anniversary of the launch of the Hubble Space Telescope. For more information and resources regarding Hubble anniversary, please visit hubblesite.
Introduction
by Joseph DePasquale
Today, we celebrate 31 years since the launch of the Hubble Space Telescope by unveiling a brand new image of one of the Milky Way’s most luminous stars known as AG Carinae, or “AG Car” for short. The naming of this star comes from the fact that it is a variable star (actually a special class of variables known as a luminous blue variable) in the constellation of Carina. The “AG” prefix is a naming convention of variable stars, which are stars that undergo periodic changes in brightness over time. There are about 20,000 light-years between us and this star as well as a large amount of galactic dust, all of which combines to make the star appear much dimmer than it actually is. To put that distance in perspective, if you were to board a commercial jet headed for AG Car, it would take longer than the current age of the universe to get there—about 15 billion years!
Hubble has captured AG Car and its nebula in exquisite detail. The surrounding nebula is the result of an ancient stellar outburst around 10,000 years ago in which the star ejected an estimated 15 times the mass of our sun in stellar material. Don’t we all need to blow off a little steam now and then? That gas and dust now glows as it is irradiated by the intense light of the star itself. Hubble observed AG Car in several different wavelengths from ultraviolet to optical.
The contrast of blue and red in the image is the result of combining those different wavelengths into one image where the blue light shows us the dust in the nebula reflecting starlight, while the red light actually comes from emission from hydrogen gas. This is actually separate and distinct from the blue light in that the hydrogen molecules are excited by light from the star and are emitting their own red light while the blue is simply a reflection of starlight.
We chose to observe the star all the way down to ultraviolet wavelengths in an attempt to investigate the possibility that the nebula may show a different morphology (shape and appearance) at those wavelengths. A similar study was done on the nearby star and nebula system Eta Carinae which showed interesting structures in UV that provided astronomers with a better understanding of the nebula’s overall appearance. Ultimately, the UV data did not show us anything new, but that does not mean the time was wasted. In scientific pursuits, a null result is still a result! Read on to learn more about how this exciting new image was assembled from the Hubble data.
The Goldilocks of Exposures
by Alyssa Pagan
Much of what makes AG Car an attractive target also makes it a challenging one. Capturing a field containing a bright star adjacent to a faint nebula requires finesse. To do so, Hubble needs to collect just enough light to capture the faint ejecta without completely over-saturating the star and the area surrounding it. With this in mind, the outreach-imaging team scheduled two sets of observations at different dates, giving us and Hubble the opportunity to pivot (literally!) if necessary. What resulted were two quality datasets that could be processed separately and later combined to maximize the field of view and mitigate artifacts by leveraging the redundancy in the observations.
Stretching: Translating what Hubble Sees
With all the beautiful and intricate Hubble images floating around, it may be a surprise to discover that Hubble data starts off black and white. Much of the intricate detail we are used to seeing in a Hubble image is tucked away in the vast dynamic range of light values Hubble can detect. To reveal these details, the brightness values have to be scaled to a range our screens can display. This process is known as stretching and for an image like AG Car, where the range between the diffuse light of the nebula and bright light of the star is particularly great, this step is even more critical. Given the object’s description, luminous blue variable, a bit of saturation from the star is to be expected; the goal instead is to minimize this as much as possible. By manipulating the mid-tones of the image, we can preserve the bright data, but bring out the diffuse structure of the surrounding ejecta.
Assigning Color
Once the appropriate stretch is applied to all the filters, the files are imported into a photo-editing program where we can begin assembling an RGB image for each data set. Colors are assigned based on their chromatic order or ascending wavelength starting from the blue end of the spectrum and ending at the red. The filters used to observe AG Car include F275W, F547M, F657N, F845M, and are prescribed a blue, green, orange, and red hue, respectively. The colorized filters are then combined additively to form a full-color image. At this point, it is important to align the separate visits and select one data set that will be altered to match the other in color and tonality. Once this is complete, an initial color balance using curves and levels adjustments can be applied globally.
Making Artifacts Disappear
After we have matched and aligned our two separate images the magic truly begins! We can finally remove the glaring ghost, the bright donut eclipsing AG Car’s right side, from the first data set and the charge bleeds. This is done by masking out the artifacts in one image revealing the clean data of the other. For the remaining artifacts which do not overlap with clean data, such as the cosmic rays and less prominent ghosts and charge bleeds, we apply the same techniques covered in this blog post—namely, replacing artifacts with the average background level.
Diffraction Spikes: Four is a Crowd and Six is a Party
Even though diffraction spikes are also artifacts, we often leave them in the image almost as a Hubble signature and a cue to the viewer that they are looking at a bright object in space. They would also be a pain to remove from every star in a field, but that is another story. In this case, however, where the star of the show is, in fact, a bright star, there was much deliberation on whether they should stay or go. Complicating matters more, due to the rotation between two visits, artifacts ended up overlapping and we incurred a second pair of spikes in the clean-up process! To remove the spikes, good pixels near the affected region must be carefully sampled and used to replace the offending data.
In the end, since AG Car was such a unique object, the team settled on the image that was most foreign to us and yet the closest to reality—the version without the spikes. Removing the artifacts allowed the more subtle details to shine forward, and the viewers to get a closer look of how this object would actually appear in space.
Orientation: Is There an “Up” in Space?
Up until this point, the images you have seen in this blog follow the astronomical convention of North-up and East-left. If I were to flip the image now, to your eye, it may seem wrong or upside down. In reality, however, notions of up and down in space are merely relative. In fact, even our view of the sky and perception of ‘up’ varies according to our latitude.
While North-up convention is certainly helpful for analyzing and comparing data from the same object, sometimes this default orientation can detract from an image. In cases like these, aesthetics trump convention and an orientation is selected that enhances the object and the overall composition. Some of AG Car’s distinguishing characteristics include its cone-shaped region of gas and dust as well the unique color variance and structure in the nebula. In the original orientation of the image, the cone stretches upward, making the composition feel awkward and unbalanced. Instead, if we rotate AG Car about 130 degrees counterclockwise, we then end up with the cone on the left side of the image which balances nicely with the visual weight of the bright reflection nebula. The weight distribution also promotes a sense of movement that mimics are eye’s natural tendency to travel left to right.
Finalizing the Image
One of biggest challenges for an image processor can be deciding when to stop. What makes an image complete? Often times the science and story behind the object can inform the image processor of where and how far to take the image. In the case of AG Car, the distinctive difference in the color and morphologies of the nebula guided the remainder of the processing. The final adjustments focused on emphasizing the color separation between the two types of nebula and the faint structures.
When processing an image, one can get lost in the data and lose sight of how remarkable it is that Hubble is still producing such quality data, and that despite all the odds, AG Car was observed in such fine, exquisite detail. As it has done for so many years, and hopefully years to come, Hubble continues to inspire, broaden our understanding of universe and facilitate groundbreaking science.
Great article!