SPT Blog

McGill's South Pole Telescope Deployment Blog

It is the South Pole Summer 2006/7 and a cutting edge cosmology instrument, the South Pole Telescope, is being deployed and commissioned at the bottom of the planet. This blog is one researcher's journal of the expedition. As with most blogs, the first entry is at the bottom ... so if you're visiting for the first time, start there. Contact Matt.Dobbsatmcgill.ca for more information.


This project is funded in the U.S.A. primarily by the National Science Foundation (NSF), with additional support from the Kavli Foundation of Oxnard, California, and the Gordon and Betty Moore Foundation of San Francisco. Canadian collaborators receive funding from the Le Fonds québécois de la recherche sur la nature et les technologies (FQRNT) and the Natural Sciences and Engineering Research Council of Canada (NSERC). The McGill cosmology instrumentation lab receives funding from the Canadian Foundation for Innovation and the Quebec government. It wishes to acknowledge in-kind support from Xilinx Canada University programs.


Feb 18, 2007

Well, that didn't last long. As the thermometer dipped downwards, the US air guard decided to evacuate us on Feb 17. Fortunately the weather was too nasty for them to take off from McMurdo that day, so our departure was delayed until today - Feb 18. The last three flights of the season came in today, and we boarded the 3rd one for Mac-town (McMurdo), where our C130 propeller plane was met by a C17 jumbo cargo plane for a connection back to New Zealand. The entire station came out to the runway to see us off. The winter-overs smiled and everyone exchanged big bear-hugs through our thick extreme weather clothing. It was a desolate sight to see the 60 winter occupants of the station waving good-bye to the last plane they'd see in eight months. As our plane lifted off and circled back over the station, the pilot banked quickly to the left and back again so that the wing would dip down and give us our last view of the 10m South Pole Telescope - commissioned and ready to bring back a year's worth of exciting new science.

Feb 16, 2007

Today is closing day for the South Pole Station - all the summer inhabitants are being evacuated and the station goes into winter mode. All the summer inhabitants, that is, except us. We've been granted a "soft close". This means that we can stay an extra week, pushing the season and the temperatures. The agreement is that we place all of our baggage in cargo today and it gets strapped to a pallet, ready to be loaded on the military cargo plane that will take us back north. We spend the rest of the week living out of a carry-on bag, with the promise that we will be ready to fly out at any time with three hours notice. The air guard that operates the US military planes will watch the temperature. As soon as it starts to dive towards -50 C, they will take off from McMurdo station to fly here and evacuate us - the last to leave. This extra week is crucial for us for our final commissioning exercises with the telescope before we leave it to operate for the winter.

Feb 15, 2007 First Astronomical Light for the South Pole Telescope

By evening we were back in business after the parking garage incident last night. After tuning up the system, we went after a more exciting target, Jupiter. We aren't so interested in making measurements of planets, but instead we use them to calibrate our system. Jupiter is very bright, so by scanning across it, we can easily measure the time-response of our detectors and map out their beam shapes. The exercise went very well, and we were able to point the massive telescope up at jupiter, hitting it immediately, and following it across the sky lighting detector pixels. We created maps of the planet at 2 and 3mm, achieving first light for the 10m South Pole Telescope.

2007-02-14_SPT-Deployment_069small.jpg 
John Carlstrom looking at timestream traces from a scan across planet jupiter

The fun isn't over - in fact it's just beginning. There's a lot of work ahead of us before we start to record the data that will help us understand the growth history of the universe. We must first accurately measure pointing errors of the telescope, define the focus of the telescope more precisely, measure beams for all of the detectors, measure time constants of the detectors, verify the noise properties, and finally start to go after some galaxy clusters. It's going to be an exciting year!

NSF Press release: http://www.nsf.gov/news/news_summ.jsp?cntn_id=108413&org=olpa&from=news

Feb 14, 2007

A wild day. Late last night we let the SPT camera loose on the sky for the first time and began tuning up the SQUID amplifiers and bolometers. We first experimented with the effects of the telescope's motors and motion on the SQUIDs (remember - the SQUIDs are special superconducting amplifiers that use quantum interference to convert really small electrical currents into voltages that are big enough that standard room-temperature electronics can measure them), and found the SQUID readout to be stable.

Today, around 3pm we were ready to look for an astronomical signal - naturally we chose the easiest thing out there, the moon. At mm-wavelengths the moon is big and bright - easy to see. Scanning our detectors across the moon our real-time display shows the amplitude of the signal coming from each detector as a function of time. When a detector crosses the moon, it's signal shoots up, stays there for a second or so, then comes back down, producing a top-hat shape. After all of our trials getting the telescope up and running, it was a magnificent top-hat indeed! Around 4pm we did our first scan across the moon. We spent most of the evening using our scans of the moon to adjust the focus of the telescope. Though this is an observation of an astronomical object, we choose not to call it first light yet - we'll reserve that title for something a little more meaningful like a planet.

But the excitement for today was not near to being over. We adjust the focus in a similar manner as you would with a regular camera - by adjusting the distance between the primary mirror and the camera. This is accomplished with a massive optical bench that is driven around inside the cabin by a set of 5 actuators (motors). The camera takes up almost the entire cabin, such that in many places there are only a couple inches of clearance. Thus, with the actuators, we are effectively driving the camera inside what amounts to being a very small garage. The actuators are controlled remotely by computer. One of our actuators wasn't working properly, so late this night we replaced the encoder that measures the location of the actuator. When we powered the actuator up again a perverse combination of software/hardware malfunction (which we prefer to refer to as a simple act of god) caused this actuator to drive itself fully in the wrong direction. This caused the optical bench to twist, such that it crashed into a protrusion in the ceiling of the cabin, with one of the high pressure helium lines hitting first (yes, you could say we had a parking garage accident while driving the camera inside the telescope cabin). The helium line broke, and high pressure helium began spraying out of our refrigerator system into the cabin causing our camera to warm up in a hurry. Disaster!- remember it took 7 days to cool this thing down, and we're scheduled to leave within a week. People moved quickly - pausing for brief football style huddles to discuss the best course of action. The damaged helium line had to be removed, and a fitting on the cold head in the camera replaced - no small task considering the camera is 2 stories above the floor and can only be accessed with ladders and a man-lift. Once disassembled like this, the refrigerator system is contaminated, and needs to be purged with ultra-pure helium gas. Mid-way through this process, we ran out of helium gas and had to scramble around the South Pole in the middle of the night scrounging for helium. Once the system was refilled with helium gas, we had to fix the malfunctioning actuator as well, since it was blocking access to attaching the replacement for the damaged hose. About 3 hours later the system was repaired, with the cold plate having risen from a temperature of 4K to 50K. There's never a lack of adventure in experimental cosmology! Fortunately we acted quick enough to save the system from warming up too much, and we'll be back down to our operating temperature in less than a day.

Feb 12, 2007

The weather forecast for today, posted by our ever-good humoured weather team at the south pole, doesn't look too promising.
SP_Forecast2007-02-11.PNG

Feb 7, 2007

2007.01.13.008MattIcedUpSmall.JPG
Photo:  Matt- a little iced up after some time outside. (Photo by Adrian Lee)

The temperature is continuing to drop, and those who spend a lot of time outdoors are getting little bits of frost-nip on their noses where goggles meet balaclavas. There's a lot of action at the station as the summer employees start to clear out and the winter-overs (those polies who will spend the 8 winter months isolated here, mostly in darkness) begin to arrive. Tonight they transmitted a recording of the Superbowl in the galley. People were surprised that I was totally uninterested and I teased them by asking, "Superbowl, is that baseball? Did any American teams make it this year?" A bunch of the station-women found outfits in the costume shack and dressed up to cheer-lead the game.

The last few days were hectic working on the SPT camera. When we cooled the camera for what was meant to be the last time this year, we found that the coldest sub-Kelvin stage wasn't able to get to its usual 250mK temperature and was instead sitting a little hot, about 400mK. Trying to diagnose a problem like this is a little like trying to diagnose a heart condition without x-ray or ultrasound (not that I would know anything about that). We can measure at of the electronics outputs, send electrical signals around, and measure heat gradients--but we can't do what we'd really like to do--just look at the guts of the receiver with our eyes. After a day of testing we hypothesized a thermal link between our coldest stage and another warmer heat exchanger stage (we call this a "touch" in the jargon--this just means, not surprisingly, that one part of the guts of the cryostat is touching another part. If the two parts are supposed to be at very different temperatures this is a bad thing because the warmer part will heat up the colder part.) We warmed the camera and brought it up to atmospheric pressure--finally allowing us to open it and see inside. Indeed we found a nasty touch and we were able to repair it. Today the camera is being cooled again so that we can continue the final tasks in our commissioning. This problem cost us a precious week- but one thing you can count on with new technology is unplanned problems... this is part of the fun and adventure of doing cutting edge science under extreme conditions.

Here's a picture of the telescope, with folks standing on top of the receiver cabin for scale. The inner ground shields are in place (the white wings jutting up from the cabin). This is the only season that the telescope will be visible in this way. Next summer a huge outer ground shield the size of a tennis court will be installed around the telescope to keep reflected light from being seen by the telescope when the light bounces off the ice.

2007-02_SPT-AndPeople_small.jpg(photo credit unknown, taken from the station's common drive)

Feb 4, 2007

Wanna learn more? The Sunday February 4 interview on CBC radio's "All in a Weekend" program is archived athttp://www.cbc.ca/allinaweekend/listenagain.html (scroll down to Feb 4). The real audio link ishttp://www.cbc.ca/montreal/media/audio/allinaweekend/20070206MATT_DOB.ram .

Feb 1, 2007 (Thursday)

Things are about to switch back into high gear down here - the South Pole Telescope's camera will be cold at about midnight tonight. It is mated to the secondary cryostat, and will see the outside world for the first time tomorrow. Currently the pair are sitting in the SPT control room, immediately below the telescope cabin. We'll cross-check the filter band-passes, look for excess optical loading, and roughly characterize the optical efficiency of this subsystem before hauling it up (while keeping it cold) into the telescope cabin - probably on Monday.

The telescope cabin is coming together nicely.Though it hasn't been officially handed over from the antenna contractor (Vertex/RSI), we've been in it quite a bit- setting up the electronics racks, working on the window mechanism, and other general prep work. The Vertex guys (Tim, Peter, and Kevin) have been great, and our team has integrated really well with them.

So what are we going to do once we get this beast all tuned up? The primary purpose for the next 2 or so years of SPT observation time is to perform a blind survey of Galaxy Clusters using something called the Sunyaev Zeldovich effect (SZe). Here's a rough sketch of how it works. Our telescope detects mm-wavelength light. Light left-over from the big bang, called the Cosmic Microwave Background (CMB) radiation, sits at this wavelength and back-lights the entire universe. Studying this light for its own sake is a worthy and important cosmology pursuit - and one of the things we'll be doing. More exciting for us is the possibility of using it to find galaxy clusters which sit between us and the time at when this CMB light was released. When our telescope looks out through a galaxy cluster, some of this mm-wavelength light will have been absorbed and re-emitted at a higher energy. This means that in our wavelength band some of the light will be missing at the locations on the sky where galaxy clusters sit. The clusters effectively act like shadow puppets, obscuring the CMB light which is arriving from behind. By surveying large regions of the night sky and looking for these signatures, we can identify distant galaxy clusters. One important feature of this signature is that is depends only very weakly on how far away the cluster is -- instead the amplitude of the signature is proportional to the mass (or total cross section for inverse Compton scattering) of the cluster. Thus, unlike standard telescopes that look at light emitted from an object, our observations will be able to see far away clusters as easily as those that are close by. So who cares?- well, lots of people. Cosmologists care a lot, because, after following up the observations with other techniques to determine the cluster's distance, we can map out the abundance of clusters as a function of time. This will place new constraints on the expansion history of the universe - and hopefully help us understand the relatively new findings that indicate the universe seems to be accelerating its expansion (this phenomenon, not really understood, is dubbed "dark energy"). If theseaccelerating universe results continue to hold up (and they're holding up very, very well so far), we've got a whole lot more to learn about our universe!-

Jan 29, 2007

It's starting to get colder down here. From about christmas onwards, the temperature just drops and drops. Up until now it hasn't been much colder than a cold Montreal day - about -28 C. Now it's heading downwards and we're starting to wear our balaclavas and ski goggles for the walk to the dark sector lab.

A few photos to give you a feel for the place:

2007-01-17_SPTdeploy_002small.jpg
The scenery at the pole is a big blue sky meeting the horizon an eternity away. There are no land features visible.

2007-01-17_SPTdeploy_010small.jpg
The new USAP South Pole Research Station, with the ceremonial geographic south pole in front.

groupPhoto2007.01.14.003small.JPG
Members of the SPT collaboration in front of the telescope: 
Back row: Ken Aird, Zak Staniszewski, John Carlstrom, Clem Pryke, Joaquin Vieira, Martin Lueker, Brad Benson, Erik Leitch, Bill Holzapfel, Jeff McMahon, Adrian Lee. 
Front row: Tom Crawford, Kathryn Miknaitis, Tom Plagge, Matt Dobbs, Ryan Salem, Stephen Padin (very front). 

Jan 25, 2007

After two around the clock days opening the receiver cryostat and adding a small change to the cold detector-squid circuits, we have closed up again and mated the receiver to the optics cryostat. The optics cryostat is a monster. It contains the telescope's secondary mirror, just over 1m in diameter, and cools it to about 4K (-269 C). We do this so the mirror doesn't radiate extra heat (loading) that would reduce the sensitivity of our sky images.

This was the first time the receiver/optics pair had ever been mated, so we were a little nervous.

2007-01-25_SPTdeploy_070small.jpgThe SPT receiver (red) being mated to the optics cryostat (white).

After mating the cryostats, we did a test lift, hoisting them up from the DSL lab into the receiver cabin where they will operate. The mating and lifting come off without a hitch, thanks mostly to the careful uber-engineering of Steve Padin, the SPT project scientist.

2007-01-25_SPTdeploy_081small.jpg
The receiver/optics cryostat being hoisted up into the SPT 10m cabin.

With all this excitement, it was time to unwind a little with a trip to the station sauna (yes, all the luxuries of home at the USAP south pole station!) followed by a quick venture outside to cool off. Polies call this the 140 club due to the +120 F to -20 F temperature differential.

Jan 21, 2007

Today is day 5 of our receiver commissioning in the lab. Things are coming together -- and it's a good thing -- our team's been going 24 hours a day, with a few of us stepping out at a time for some quick sleep. Our buddies on the other side of the lab declared the optics cryostat all ready to go -- so we've got a day to finish our work with the receiver and start the process of mating the two cryostats.

2007-01-25_SPTdeploy_018small.jpg
The SPT receiver, with the 4K cavity visible, but focal plane removed.

2007-01-25_SPTdeploy_026small.jpg
Brad Benson with the focal plane that he designed.

2007-01-25_SPTdeploy_030small.jpg
One wedge (160) of SPT bolometric detectors are shown.

2007-01-25_SPTdeploy_052small.jpg
Bill holzapfel, framed by the SPT receiver aperture.

The station is getting ready for winter- stocking up on supplies and fuel for the 8 months during which no flights are possible. There's a certain lingo that goes along with living at the station. Polies (citizens of the South Pole) gossip incessantly about the arrival of freshies (fresh fruit & veggies flown in from New Zealand) and going far north (New Zealand). This being a research station, there are plenty of beakers (scientists) around. The South Pole Telescope is referred to as "the ten meter".

Jan 19, 2007

Inside the receiver are 960 bolometric detectors. Each bolometer consists of an absorber and a thermistor. The absorber is a small piece of metal in a spiderweb pattern that absorbs millimeter light and converts it to heat, raising the temperature of the piece of metal. The thermistor is a small piece of superconductor that is locked in a state midway between superconducting and normal. It is very sensitive to temperature changes. Since we're expecting very small temperature changes, the detectors need to be kept very cold - about 250 mK (-273 C). The bolometers will look out at the night sky through a series of filters that define the bandpass of the instrument (the wavelengths of light that we're looking at - in this case millimeters). The light is focused through a lens, then reflected off the aluminum mirrors, of which the big 10m primary mirror is the first to collect the light from the sky.

I'll explain more about the SPT receiver in a future post.

2007-01-17_SPTdeploy_028small.jpgTom Plagge, Adrian Lee, and Matt Dobbs next to the receiver (capped off so no light can enter) and the rack of readout electronics.

Foraging for food can be a pass-time out here, since we've often got our heads in the receiver out at the dark sector lab. Fortunately, I understand humans can survive for months on goldfish alone: 2007-01-17_SPTdeploy_037small.jpg

A few overlanders - skiers who had made the journey from the Antarctic coast under their own power - glided into the station today. At lunch one of the icecube scientists commented, "wow, what a tough journey that must of been - really the hard way to get here compared to our easy plane ride" ... to which one of the Berkeley graduate students commented "are you kidding, not compared to my journey. I've suffered through 5 years of graduate school to make it here!". We're never to busy to joke around a little - being sleep-deprived helps with that.

It's late into the night (still fully bright outside of course) after another long day with the receiver. It's easy to stay awake here since it's always light out and the sun energizes you. I'm trying to decide whether to set my alarm for 5am, when the iron-workers (welders working on the telescope structure) get off their last shift of the week (marking the completion of the iron structure). They're planning a big party in the summer-camp lounge to start right after their shift. Since the sun turns the days upside down, a 5am party is not as strange as it sounds. Judging from the game of high speed full costume dodge ball that was going on this morning at 7am in the station gym, it's bound to be a memorable party and serve as interesting breakfast entertainment. Did I mention that the station has a full room (not heated) of dress up costumes to make festive occasions a little more interesting? - great vintage 70s dresses and wigs to match the most outlandish personality.

Jan 16, 2007

The USAP station sits just meters (err, I guess it would be feet to them) from the geographic South Pole. Since the 3km thick glacier is constantly moving, the physical marker for the South Pole needs to be displaced a little every year. It is denoted with a ceremonial globe.2007-01-17_SPTdeploy_013small.jpg

But enough fun and games - we've been working hard to get the SPT receiver up and running, and tomorrow will be it's first operational day at the pole.

I'm part of the team that built the receiver and readout electronics for the South Pole Telescope. The receiver is a special "camera" that is able to detect millimeter wave light. The receiver first needs to be pumped down to a pretty good vacuum - about 10^-7 Torr, then cooled with a series of cryo-coolers to 250 mK. That process takes about 5 days - and so tomorrow is the first day that we'll start to test it. We start with a dark test (meaning the receiver is not open to light) in the laboratory. The next step is to mate it to a second cryostat that houses a cold mirror, then finally we can look out at the night sky and see the beauty of the universe for the first time with this telescope.

Once "we're cold" the fun begins. The six of us receiver folks will break into two shifts, so that we can test and debug the system 24 hours a day. I'm on day shift with Adrian and Tom. We typically like to overlap with the other shift by several hours to exchange info and ideas, meaning this will be all we do with ourselves for about a week until we're ready to warm up and begin the process of mating the receiver to the optics cryostat.

Jan 13, 2007

We boarded a military C130 propeller cargo plane for a loud three hour flight to the South Pole. Three quarters of the way there the plane boomeranged (turned around) because of high winds and low visibility reported at the landing strip. We breathed a breath of relief when it turned around a second time after ten minutes, resuming the journey to the pole when the ground crew reported that the weather was clearing. The C130 looks like it is about 50 years old. The cockpit is full of old button and analog meters. It's louder than an Metallica concert. Despite all this, it is the only big cargo plane that the US has that is outfitted to land on skis.

2007-01-13_SPTdeploy_003small.jpg
Military C130 propeller plane.

Walking down the ice run way it was magnificent to look up and see the impressive structure of our 10m South Pole Telescope rising up out of its foundation on the ice. After more than three years of work on the project, it was a great feeling to see it in form. The crew did a great job assembly the antenna over the last 3 months.

2007-01-13_SPTdeploy_021small.jpg
Matt in front of the South Pole Telescope and the Dark Sector Lab (blue).

I met up with the rest of the South Pole Telescope team in the galley of the main station where the day crew was eating breakfast and the night crew eating dinner.

The dark sector lab, where the SPT is located, sits about a kilometer away from the main station. The walk way is marked with flags - except where it crosses several hundred meters of runway. The wind was up and the visibility was low by mid afternoon. It can be disorienting walking across that hundred meters of white, with nothing to look at.

2007-01-25_SPTdeploy_012small.jpg The South Pole!

I was assigned a sleeping berth in an old Jamesway hut, about a half kilometer from the main station. The Jamesways are big half-cylinder canvas structures about 50 feet deep and 20 feet wide, housing 10 or so rooms separated by canvas walls. Heat is provided by an oil burning furnace. My room was all tricked out with a window cut in the wall and carpet strung up on the walls for soundproofing. Yellow snow isn't allowed at the pole, so going to the bathroom at night is a big ordeal, dressing up in layers of extreme weather gear for the short walk.

2007-01-13_SPTdeploy_013small.jpg
A Jameway hut - sleeping quarters out at the South Pole Station "summer camp".

Today being saturday, it is a big night at pole-station, since all the construction crews have sundays off (not so for the scientists) - so the galley was full of cheer, beer, and music. I sat in for a few numbers playing bass with the station's resident blue grass band before retiring to the Jamesway for sleep.

Jan 12, 2007

Our project is part of the United States Antarctic Program (USAP), as it sits at the American operated Amundsen-Scott Research base, -90 degrees latitude.

The USAP issued us our "extreme cold weather" gear- a two hour process of fitting a mountain of clothing. Most of it was pretty well chosen, so I wasn't disappointed that I hadn't brought much of my own cold weather gear.

This morning we showed up at the airport's antarctic terminal at o-five hundred hours, as instructed by our US Navy chaperons. We got dressed in our -50C gear (mandatory for the flight), and sat around sweating for 4 hours in the +15C weather until we could board the huge C17 military cargo jet bound for McMurdo station on the antarctic coast.

McMurdo is the base for US antarctic programs, and sits just in from of Mt. Erebus, Antarctica's active volcano. It is an ugly mini-town, full of shipping containers and oil storage facilities -- but the surrounding landscape is breathtaking. Mountains, glaciers, and sea ice as far as the eye can see.

2007-01-12_SPTdeploy_032small.jpg 
!McMurdo Station, on the Antarctic coast, as seen from Scott's hut.

It was also the launching point for Scott's expeditions and many of the historical structures are still standing.

January 11, 2007

After a 5 hour flight to LA, waiting 8 hours for a 12 hour flight to Auckland, New Zealand, then a short flight south to Christchurch, the "gateway to the antarctic", I was glad to have a day layover in New Zealand - but also anxious to get to work down South on our project.

Christchurch is small and sleepy - but not fully asleep. I found three pubs with live music, so that meant with 2 nights in town and a little walking I was able to see six bands. Five fell under the unfortunate category of outdated rock cover bands or singer-songwriters who hadn't written any of their own songs... but the sixth one came through - a funk-grungy local band called OrKid that put on a tight show with a full sound (especially for a trio).

January 3, 2007

Several members of the SPT collaboration have been down at pole since mid-november assembling the telescope and main reflector. They have been doing a great job and here's some evidence.

2007-01-03_SPT-DishInstalled_small.jpg 
The SPT primary reflector being mounted on the telescope.

SptWithShield.png A computer model showing what the telescope will look like when it is fully assembled.


Comments