Removing the twinkle of stars

This month on Conversations in Science, I spoke about something a little more closer to home. I had fun going back to my PhD roots and spoke about what I had spent nearly eight years of my life researching: how one can remove the twinkle from stars.

Twinkle Twinkle Little Star... Not!
(First Aired on KLRNRadio, Monday, November 7, 2016)

When I was studying for my PhD, it was a little inside joke between my supervisors and myself. Whenever anyone asked what it was we did for a living, we would answer, "We remove the twinkle from stars." The reactions were hilarious. Most people would see it for the joke it was, but there would always be the one person who would get incredibly offended.

"But I like the twinkle in the stars. You can't remove the twinkle. That's just rude and I'll create a petition to have your research shut down."

Okay, that particular reaction never actually happened, but there were some people that took us seriously. Like we could really remove the twinkle from the stars... (Shall I snort now or later?)

(To make matters even funnier, my main supervisor would often introduce himself as an astrophysicist to those he didn't want to talk to, but as an astronomer to those that he did. It's amazing how the one change in word shifts the dynamics of a conversation.)

My field of research was in adaptive optics. We were looking at ways to stabilize the light entering into a telescope to improve resolution of images and the data collected. In reality, that twinkle of stars that people love so much is actually detrimental to astronomical observations.

In 2007, I was at the Mount John University Observatory collecting data while Comet McNaught was in the night sky. This was just one of the many shots that I got while walking around the mountain. The dome in seen is for the 1-m McLellan Telescope. (Photo credit: Judy L Mohr)

In 2007, I was at the Mount John University Observatory collecting data while Comet McNaught was in the night sky. This was just one of the many shots that I got while walking around the mountain. The dome seen is for the 1-m McLellan Telescope. (Photo credit: Judy L Mohr, 2007)

Imagine trying to take a picture of something that is constantly moving, but moving in an erratic way. All you get is a blurry mess. It might turn out to be quite an artistic image, but for someone actually trying to take pictures of the stars... Yeah, not a good thing.

So, I spent nearly eight years of my life looking at how one can compensate for the twinkling of the stars in real-time. It's not an easy task and the mathematics did my head in. However, the concepts behind adaptive optics are actually quite easy to understand.

The atmosphere can be thought of as glass, but not just any glass. I'm talking about that rippled glass that is commonly used in bathrooms and toilets. If you were to shine a light through that glass, you'll give a rippled pattern where parts are brighter than others.

Now imagine that the rippled pattern in the glass was constantly moving. The bright sections of light would be flickering — they would be twinkling.

The effect of atmospheric turbulence. (a) The diffraction-limited image of a star where no distortions are induced by the atmosphere. (b) The real situation where a speckle pattern is seen in the short exposure image due to the aberrations induced by the atmosphere. (Source [1])

The effect of atmospheric turbulence. (a) The diffraction-limited image of a star where no distortions are induced by the atmosphere. (b) The real situation where a
speckle pattern is seen in the short exposure image due to the aberrations induced by the atmosphere. (Source: Reference [1])

Simulated twinkling star. The dancing image would make it difficult to image properly. (Credit: Judy L Mohr, 2009)

Simulated twinkling star. The dancing image would make it difficult to image properly. (Credit: Judy L Mohr, 2009)

In reality, the atmosphere is not a single sheet of glass with a moving rippled pattern. It's actually multiple sheets of glass, each with a different characteristics. In some cases, that layered sheet of glass is incredibly think. However, if you were able to determine exactly how the layered sheets in the atmosphere were rippled and how they were changing with time, then you could compensate for the resulting twinkling effects on stellar light.

Before you can dream of compensating for the twinkling effect, you have to model it. So that's what I was doing. For eight years, I was attempting to model the effects that the atmosphere has on stellar light above Mount John University Observatory in Tekapo, New Zealand. I collected four years worth of data, looking for seasonal trends. After an additional three years of data analysis, here is what I found.

Five years of design and this is what I came up with in the end. It may not look like much, but that was the point. One of my research goals was to prove that you could still use one of the recognised methods for measuring optical atmospheric turbulence but using a system made from off-the-shelf components and will very little budget. That rig cost $2000 USD to build. A far cry from the $100K USD that some systems cost. Left: System as mounted on the back of the 1-m McLellan Telescope. Right: The SCIDAR system and its designer (i.e. me). (Photo credit: Judy L Mohr)

Five years of design and this is what I came up with in the end. It may not look like much, but that was the point. One of my research goals was to prove that you could still use one of the recognized methods for measuring optical atmospheric turbulence using a system made from off-the-shelf components and with very little budget. That rig cost $2000 USD to build. A far cry from the $100K USD that some systems cost. Left: System as mounted on the back of the 1-m McLellan Telescope. Right: The SCIDAR system and its designer (i.e. me). (Photo credit: Judy L Mohr, 2007)

Because of the way the atmosphere moves above Mount John, the 1-m telescope had the exact same resolving power as a 5-cm telescope. Basically, if the 5-cm telescope was able to collect the same amount of light as the 1-m, then they would have the same scientific capability. Here we have the only scientific observatory in the country, one that accommodates the largest telescope in the country (1.8-m MOATel), yet, the site was no better than your run-of-the-mill children's binoculars.

Here's the thing: even the telescopes on Mauna Kea, located in Hawaii, one of the most respected observatories in the world, are plagued by the twinkling of the stars, resulting in uncorrected resolutions no better than say a 10-cm telescope.

This is where adaptive optics comes in. Using mirrors and lenses that deform on the fly based on predetermined correction factors, the stellar light can be stabilized to dramatically improve the resolution of the telescope. In some cases, corrections have been reported to be in the order of 40 times improvement.

But adaptive optics systems are incredibly complex, containing multiple mirrors and lenses, along with lasers and high-powered computers. And they are expensive, with the adaptive optics system often costing more than the telescope that it's installed on. With the cost and the complexity, many ask why do we bother. Well, the only other solution is to use telescopes that are outside the atmosphere, such as Hubble.

I saw all different kinds of weather at Mount John University Observatory, and many with fond memories. It never ceases to amaze me how wondrous the world is. Top left: The dome for the 1.8-m MOATel (the largest telescope in New Zealand) and 60-cm Boller & Chivens Telescope. Top left: Cloud often hung low over Lake Tekapo and the township below. Good thing that Mount John had an elevation of 1029 m (3376 ft) above sea-level, a whole 329 m above the township below. Bottom left: Snow on the mountain frequently created a haze near ground, but crystal-clear skies above. Bottom right: The dome for the 1-m McLellan Telescope and the living quarters for observers. (Photo credit: Judy L Mohr)

I saw many different kinds of weather at Mount John University Observatory. It never ceases to amaze me how wondrous the world is. Top left: The domes for the 1.8-m MOATel (the largest telescope in New Zealand) and 60-cm Boller & Chivens Telescope. Top left: Cloud often hung low over Lake Tekapo and the township below. Good thing that Mount John had an elevation of 1029 m (3376 ft) above sea-level, a whole 329 m above the township. Bottom left: Snow on the mountain frequently created a haze near ground, but crystal-clear skies above. Bottom right: The dome for the 1-m McLellan Telescope and the living quarters for observers. (Photo credit: Judy L Mohr, 2005 - 2007)

However, space telescopes have their own problems. Let's ignore the cost that it takes to launch a satellite into space, and let's totally forget about all the engine designs needed to keep something like that in orbit, but move it around so it can look at different regions of space. The size of any space telescope is 100% dependent on the restrictive size of the payload that we can launch. Hubble's primary mirror is 2.4 meters (94.5 inches) in diameter. But even Hubble won't have the resolving power of a 2.4-m telescope. There will be flaws in its optics which restrict its resolution. But the main restriction on Hubble is its light collecting power; it will only ever be able to collect the light of a 2.4-m telescope.

Why is this important? The more light you can collect, the fainter the star you can see. The fainter the star, the further away you can see. Basically, Hubble will never be able to see as far into space that SALT can see. SALT (Southern African Large Telescope in South Africa) has a diameter of 10 meters.

Yes, SALT employs an adaptive optics system. All telescopes above a certain size will employ adaptive optics. It would be a waste of a perfectly good telescope if they didn't. (Remember Mauna Kea has an uncorrected resolution of 10 cm.)

As any astronomer knows, clouds are the bane of our existence, however, sometimes you can get some gorgeous sunsets. (Photo credit: Judy L Mohr, 2006)

But the twinkling of stars is not the only way the atmosphere plays with light.

Think about driving down a long straight road on a sunny day. In the distance, you see rippling above the road. The further into the distance you look, the greater this mirage effect is. This is caused by the heat rising off the ground. However, the closer you get to the origins of the mirage, the effects become diminished and practically disappear. This is because the effects that air movement has on light need distance to propagate before we can detect it. It's the twinkling of the stars all over again, but along a horizontal path instead of vertical.

For those trying to take pictures over long horizontal distances, and I'm talking miles, the mirage effect is a serious problem. However, like in the stellar situation, if you could model the effects that the atmosphere has on the light, then you can compensate for it. The equations for the horizontal problem are significantly more complicated, but it is doable.

In fact, in 2002, researchers from ADFA (Australian Defence Force Academy) released corrected images of a house that was 10 km away from their imaging system (over 6 miles).

Single frame from a telescopic sequence obtained over a horizontal distance of 10 km (over 6 miles). (Source: Reference [1].)

Single frame from a telescopic sequence obtained over a horizontal distance of 10 km (over 6 miles). (Source: Reference [2].)

Imaging over long distances has been an area of research for many long years. Granted large portions of the research have been initiated by military and spy-related activities, but other sectors also benefit from the technology: astronomy (which is how I came to be in this field myself), telecommunications, civil aviation, and users of Google Earth just to name a few. You read that correctly: Google Earth has taken advantage of satellite imagery for years.

I recently wrote another post on this topic for Dan Koboldt's Science in Science Fiction series.

With the advances in imaging technologies and increased understanding of the effects that the atmosphere has on light, I can't wait to see what sort of images are collected in the future.

Feature Image:
This pair of images of the galactic center, the rotational center of the Milky Way galaxy, shows how adaptive optics technology can sharpen a telescope's view. The image on the right shows the level of clarity achieved with Keck Obsevatory's current Adaptive Optics systems compared to the image without AO on the left. The position of the super-massive black hole at the very center of the galaxy is marked in the AO image. (Credit: UCLA Galactic Center Group/WMKO)

References:

[1] Mohr (2009) "Atmospheric Turbulence Characterisations Using Scintillation Detection and Ranging." PhD Thesis, University of Canterbury.

[2] Jahromi, et al. (2002) “Image Restoration of Images Obtained by Near-Horizontal Imaging through the Atmosphere.” DICTA2002: Digital Image Computing Techniques and Applications, 21–22 January 2002, Melbourne, Australia.

Posted in Astronomy, ConvoScience Podcasts (Archive), Science and tagged , , , , .

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.