Blog

For all of the non-astronomers out there (hi dad!) who have been patiently wading through our blog because they love us, here is a very brief crash course on AO. We thought that this might help you to understand why what we’re doing here at MagAO is so exciting.

The first important concept is that, in an ideal world, the resolution of telescope images is directly proportional to the wavelength of light you’re looking at and inversely proportional to the diameter of your telescope. In other words, the formula for resolution is:

resolution = wavelength / telescope diameter

Small resolutions are the best (we generally call this “high resolution” when we’re talking about cameras or TVs), so you can improve by either (a) going to a bigger telescope (increase diameter = divide by a larger number = smaller resolution) or (b) going to shorter wavelength light.

The first and even second and third generation AO systems all operate on infrared light, which has relatively long wavelengths.  We are the first 8m-class telescope system to push to shorter wavelength visible light (hence the name of our blog, VIS-AO). If there is a resolution advantage to shorter wavelength light  (visible over infrared), why did/does anyone bother with infrared AO?

The reasoning comes down to another formula, this time the formula for the atmospheric coherence length (also known as the Fried parameter). The official definition of this quantity is “the area over which an rms wavefront aberration is less than 1 radian”, which is a little opaque so let’s break it down. RMS is “root-mean-square” and is essentially an average. “Wavefront aberration” is the amount that a wave’s position deviates from perfection (perfectly flat, which is why we’re always talking about “flat wavefronts”). The atmosphere is the culprit here, bending our wavefronts from a perfectly flat shape.

One way to think about the Fried parameter is as the size over which you can expect the atmosphere to “behave” the same. The Fried parameter is proportional to wavelength to the 6/5 power. This means that at infrared (long) wavelengths, the Earth’s atmosphere is coherent over bigger patches. So if you want to correct at visible (short) wavelengths, you have to correct on smaller spatial scales. Adaptive secondary mirrors (ASMs), which are physically large compared to the tertiary mirrors used in other AO systems, can fit enough actuators on the back to correct at the necessary spatial scales for VisAO. In short, VisAO is hard, and we are heroic individuals!

The great irony, no matter what wavelengths you’re using for your AO correction, is that light from astronomical objects spends the vast majority of it’s journey to Earth with perfectly flat wavefronts. It’s only in the last 300 miles, the part where it’s traveling through the Earth’s atmosphere, that it becomes distorted. 300 miles seems like quite a distance on the surface (hah!), but it’s actually only a teeny tiny fraction of the entire journey made by light from an astronomical object.

In fact, let’s quantify that. If we’re looking at the very nearest star, Alpha Centauri, its light will have taken about 4.5 years to reach us. At a speed of 186,000mi/sec, that’s:

4.5yr x 186,000mi/sec x 3600sec/hr x 24hr/day x 365days/yr = 26 trillion (26,000,000,000,000) miles

to get to us. But the Earth’s atmosphere, which does all of the wavefront bending, is only 300 miles thick, so the portion of the light’s journey that is spent in the atmosphere is only 300/26 trillion, or 0.00000000001 (=0.000000001%).

That means that 99.999999999% of the journey was complete before the light found itself in need of our AO services, and that was for the very nearest star. If we’re looking at more distant objects, that fraction only increases!

This, of course, is why people put telescopes in space. If you don’t bother with those last 300 miles, then you don’t have to correct your wavefronts at all.

So why do we put telescopes on the ground? Well, there are lots of reasons. My two favorites are (1) you get a lot more bang for your buck on the ground and (2) you can upgrade the technology on a ground-based telescope whenever you want, so it will never become obsolete.

Hope that helps clarify what we’re up to a bit. Thanks for sticking with us!

Today we were down 50% of our tenured professors and 50% of our Italians.  Phil left yesterday and Simone and Enrico left this morning.  Therefore, I was a little worried because their presence is definitely missed.  Luckily, it was a good night anyway!

We weren’t socked in with high cirrus like we were yesterday, so we were able to keep the loop closed much of the night.  We focused through several of the filters in Clio (Clio has 10 color filters, 2 neutral density filters, and 2 cameras — so it’s quite a lot of modes to commission!  Not to mention the coronagraphs, prism, and non-redundant masks!).

We also got nodding fully functional on Clio.  Nodding is when you move the star to another part of the detector chip.  This is important in the thermal infrared where you have to subtract adjacently-nodded images to subtract the warm sky background.  We nod Clio by nudging the telescope pointing a bit.  This causes the star to also move on the AO wavefront sensor and on the other science camera VisAO.

The AO system can handle the nods, and gets the star back on-axis for VisAO, but it takes a second or two.  During that time, Jared has to deal with bursts of poor image quality on VisAO.  So nodding is working, but now we are working on traffic control.  We are working on getting a smooth operation between the two science cameras that allows Clio to nod without affecting VisAO’s image quality too much.  We are working on taking darks with VisAO’s shutter closed during the bursts of poor image quality.  Watch this video to see how we are dealing with the problem:

 

Here we’re enjoying some good closed loop images:

 

“You can look on either camera for good images!!!” – T.J.

“We’re nodding.  Deal with it.” — Katie

“The loop is back.  Deal with it.” — Alfio

“It’s ballpark super-well focused.” — T.J.

“Your advisor is telling you to take a break???” — T.J. to Jared Re: Laird

“Eat your vegetables at dinner.  They’re good for you.” — Phil (T.J.’s advisor)

“Those are pretty high gains… it’s like we’re doing real AO!” — Laird (1.4 tt and 0.4 ho)

“Look at what Clio is doing to VisAO!  It’s criminal!  See, this is why you never use two science cameras simultaneously!” — Laird

“It’s a frickin’ miracle!” — Laird, commenting on having two science cameras operating simultaneously behind an AO system.

The unofficial battle cry of the MagAO project over the last few days has been “More Cowbell!” as we try to get a high-order basis set that keeps our actuators happy.

To get the truly amazing image quality that our system is capable of, we need to find a set of shapes for our mirror that is stable on sky. We aren’t there yet, but we also need to test what we have on real stars to make sure we are on the right track. So we went on sky again tonight, but first we had to pull the CRO.

Kate found us a nice quadruple system to check our closed loop image quality:

Simone and Enrico are leaving this morning. Right before he left, Simone showed me this. Maybe it contains the answer to the riddle: “how do you simultaneously minimize force and maximize rejection?”

“I got a fever, and the only prescription is more cowbell.” – Laird Close

(If you don’t get the cowbell jokes, click this link.)

“I put my pants on, just like the rest of you — one leg at a time.
Except once my pants are on, I make extreme AO systems.” – Simone Esposito

 

“All that brain power concentrating on the same thing – how can it go wrong?” Marcia Males (Jared’s Mom)

I’ve already looked up “Strehl’s,” and I’m on my way to “modes.” – Morzinski family friend, and fan of the blog

Today we worked more on interaction matrices and calibrations during the day.

In the evening, Phil and T.J. spent some time optimizing and testing Clio.  Clio got to move to the big computer in the Clay control room for the first time (instead of running it off Phil’s laptop).  Here are some pictures of the big event.

 

Jared  is taking the slopes from the wavefront sensor and multiplying them by the reconstructor to determine the phase of the wavefront (including adding back in the high order terms that are above our cutoff frequency).  He writes the residual wavefront error to the VisAO FITS headers.  He also records the gimbal position so we know the position of the star in the image with respect to the center line set by the AO Pyramid.  Jared has calibrated the gimbal position and the wavefront error using the focal plane.  Soon, we will only need the VisAO FITS headers, not the images themselves!

 

Sunrise from this morning:

Jared: I record the wavefront error and position of the star in the FITS headers.
Katie: So what do we need the VisAO CCD for anyway?
Kate: To calibrate the FITS headers!

And it won’t stop us tonight either! Rest assured that our PI had the foresight to keep the CRO on today, so MagAO commissioning continues unabated.

Look how short that post was!