Goldstone: Desert outpost performs radio imaging of close-passing asteroid 2005 YU55

Anticipating the close flyby of asteroid 2005 YU55 yesterday, the Jet Propulsion Laboratory invited media to tour Goldstone, one of three facilities that make up NASA's Deep Space Network. I've always wanted to see these massive radio dishes up close, so I jumped at the chance! The tour was on Monday, the day before the asteroid encounter. I queued up the audiobook version of Dava Sobel's new book and drove the 150 miles to Fort Irwin, beyond Barstow, California.

I suppose it's not a very insightful comment, but I'll make it anyway: the 70-meter dish at Goldstone is really, really big. From where we parked to walk around it, it was impossible for me to get the whole thing in one camera frame. So I took several mosaics and used Photoshop's Photomerge function to stitch them together. (All my photos are posted on Flickr.) Here's a side view. If you look about two thirds of the way up the pedestal, you can see several red doors -- those are standard-sized doors, for people.

Here's a front view:

I was taken, first, to the station's operations center. As JPL likes to do, they made one of its four walls out of glass, permitting tours to see what's inside without disturbing the operators. Of course, it's pretty obvious when a group of people is standing outside your window gawking at you, so lead space operations controller Ross Hayes -- sitting at the console dedicated to DSS-14, the big 70-meter dish -- turned and waved at us. There were five such stations in the room, and operations manager Kim Massey said that in fact most of the operators monitor several antennas at once, supporting 35 spacecraft and 130-140 contacts per week over 6 antennas. He also said that they get 2,000 kids per year touring the facility, and that he hopes that some of them are inspired by what they see to pursue careers in science and technology.

Then we were given hardhats and invited to climb a few stories up a set of fire-escape-style stairs on the outside of the pedestal. Our guide for this part of the tour was Robert Harrelsson, whose title I didn't catch, but who was an incredibly enthusiastic interpreter of the advanced technology it takes to accomplish feats like talking to spacecraft more than five billion kilometers away, or imaging hundred-meter-wide asteroids. Here, he's got his hand on the steel ring that's lubricated by a quarter of a millimeter of oil, which suppports the entire rotating, swiveling assembly of the 70-meter dish, weighing in at 4 million kilograms. That assembly was recently replaced, which was a major operation.

He took us to an interior room where they can monitor the constantly changing position of the antenna and the operational state of its various "cones." The display he's pointing to here shows one LED lit, informing us that cone P1 was radiating, sending a 400,000-Watt signal toward a 40-meter asteroid.

Speaking of those cones, here's a crossed-eye stereo view of them. There are three; two are for S- and X-band, and one for X-band and K-band. (More on that here.) The telescope uses only one cone at a time, rotating it into position. The one in use is the one closest to the camera. For a sense of scale, you can see a ladder attached to its right side; each cone is about three stories tall, and the whole rotating assembly six or seven stories. Radar astronomer Lance Benner came outside to chat with us and told me that ordinarily, when they do monostatic observations of asteroids (broadcasting and receiving from the same antenna), the gold-toned metal plates on the tip of that cone go clunk, clunk, flicking back and forth every six seconds as the antenna switches from transmitting to receiving modes. But at the moment they were doing bistatic observations, broadcasting from DSS-14 and receiving at the 34-meter DSS-13 at the far end of the Goldstone installation.

We walked down the stairs and then into the base of the telescope's pedestal, where I was rather surprised to find actual radar astronomers actually looking at data in nearly real-time -- you can see on the screen at left the picture of asteroid 2005 YU55 that they released that day. I've always seen the DSN as part of a data pipeline, one that receives and then forwards data to other locations where scientists examine it; but here, it's like a telescope with operators and astronomers working together in a control room. Even optical astronomers are often not physically located in telescope control rooms any more! But here they were.

After the tour was over, we collected outside to wait for interview opportunities with the radar astronomers. Although they weren't quite finished with their observing run, Lance took pity on us standing in the chilly wind outside and came out to share his excitement about the data they'd amassed so far.

Here's my notes about what he said: The radar observations of 2005 YU55 taken in the last few days from Arecibo and Goldstone have allowed them to significantly improve the precision with which they can predict the near-Earth asteroid's future path. Before November 4, we only knew its orbit from 1827 to 2011; now, we can extend our predictions by 64 years. Those predictions have revealed that in 2029, YU55 will pass even closer to Venus than it passed by Earth yesterday. That close pass increases its positional uncertainty after that by a large amount. But the uncertainty is still small enough for them to predict another close encounter with Earth, in 2075, in which it might (but might not) pass closer to Earth than it did yesterday. There is no risk of it impacting Earth at that time. Beyond that encounter, it's impossible to reliably compute its future path. Thus, Lance said, this is an object that we (meaning humanity) will have to observe indefinitely; some day in the future, it's very likely that it will, finally, hit Earth (though of course it could also be taken out of commission by Venus, since it approaches that planet closely too).

So much for the orbit; now, onto the radar imaging. I explained how delay-Doppler imaging works in a previous post so I won't belabor that here. Here's the photo that Lance had in his hands on Monday:

He said that so far, they had confirmed the diameter derived from Arecibo observations (400 meters), and that its rotation period is about 18 hours. He was delighted with the images' quality, a result of a new system recently installed on DSS-14 that improved its imaging resolution by a factor of five. This is an astonishingly detailed radar image; most radar images of asteroids I've seen before are much noisier (more speckled) and have far fewer pixels. The pixels are 3.75 meters wide. Lance said that there are radar-bright spots on the asteroids that are most likely reflections off of large boulders like those seen on the surface of asteroid Itokawa by Hayabusa. He said these radar-bright spots are only visible when the echoes are very strong (when asteroids are very close). I asked him what other asteroids they'd been seen on, and he named 1998 CS1, 1992 UY4, 2006 VV2, and 1999 RQ36.

This wasn't even the closest image that they got. Any closer ones will have the same resolution, 3.75 meters, but will have even better signal and less speckle effect from receiver noise. That's because, unlike with optical imaging, there isn't a direct relationship between distance to the target and image resolution with radar imaging. From November 7 to 10, they will be acquiring 3.75-meter-resolution images, the highest resolution of which the upgraded Goldstone system is capable.

To try to explain why resolution doesn't change over that period: the measurable delay and Doppler shift don't become smaller with distance -- an asteroid rotating at a particular rate will return a signal with the same range of Doppler shifts of the signal's frequency regardless of how far it is from the telescope. So if the signal is strong enough, the sizes of pixels stay constant and are limited only by the capability of the system; Goldstone's can achieve 3.75-meter resolution. In practice, however, the increasing distance does have a strong effect on the strength of the signal. The signal strength goes down as the fourth power of the distance. So when things get far away, they have to bin their data in order to boost the effective signal over the background of noise in the receiver. Lance told me that "We used 18.75 meter resolution on Nov. 4, 7.5 m on Nov. 6, 3.75 m on Nov. 7, 9, and 10, and we'll probably go back to 7.5 m on Nov. 11."

Here's what an asteroid located much farther away, imaged with the same system, looks like. 2010 JL33 is about five times larger than 2005 YU55 but was located 8.5 million kilometers away, roughly 30 times the distance of YU55. At that great distance, the signal was much fainter, so they produced images with a resolution of 75 meters, 20 times coarser than the YU55 image above. The speckly appearance of the background is noise from within the antenna system.

Shortly after that show-and-tell, the 70-meter dish finished the day's observing run. In between runs, the dish returns to a neutral position, pointing straight up; it's in this position that maintenance can be performed on it. We were told ahead of time to watch for the dish to move upright.

I hiked off into the deserty area to the west until I got to a point where I could fit the entire dish into my camera's field of view, and sat down to wait. It was chilly and windy, so I huddled on the ground to try to conserve body heat and prevent myself from shivering while I recorded video (without success, unfortunately). The dish stayed pointed as it was for several minutes, then the machine hum that had been droning away since I arrived ebbed, and suddenly it was much quieter. That was the clue that the dish had begun to move.

It was astonishingly quiet as it steadily lifted and rotated. From where I sat, the gibbous Moon came out from hiding behind the dish as it rose. I kept on losing the sense of scale -- kept having to reorient myself to visual cues to tell me how large this thing really was. Here's my video. I did record it in stereo, and am trying to figure out how to share that effectively.

After it stopped moving, the machine hum began again. I shot one more mosaic of the thing pointing straight up, with the Moon near the horizon in the distance.

When we departed, we paused to get a distant view of the antenna and its neighbor, the 34-meter DSS-15.

Driving out, other farms of 34-meter dishes kept popping into view from their separated hollows.

And that was it. I'm so glad I finally got a chance to see one of the DSN installations. This network is so important -- without these great antennas to receive their data, none of our deep-space missions is of any value. Some work is being done to keep them operational, like the replacement of that bearing assembly, but not enough. I worry that, like freeway bridges, these dishes will suffer from insufficient maintenance funding until some calamity makes it impossible for us to receive all of the data being produced by our deep-space missions. I'm not sure how best to advocate for it, but we need to support the DSN!