Observing from a large telescope is a rather different experience to what you might imagine. It's not a case of being outside with the equipment, your eye to the eyepiece, capturing your data. We're sat in a control room several floors below the VLT platform, with UT3 doing its work remotely on the mountain top. In fact, all the telescopes are controlled from here - there are horseshoe-shaped desks for each telescope (the UTs), each one with a telescope operator, the support astronomers and, in our case, space for the visiting astronomers. There's also space for the ELT (which will be controlled from here, even though its platform, Cerro Armazones, is kilometres away) and the interferometry. It’s all very high-tech, but essentially a large open plan office where the astromers sit for the full night.
We have dinner in the Residencia at around 5pm, then drive the 10 minutes up the mountain in ESO's fleet of old, slow Fiat Puntos. Sunset happens pretty fast at this latitude of 24S, so we head up to the VLT platform to watch the sun vanish over the Pacific ocean, shrouded in an array of brightly coloured clouds. In fact, the clouds always obscure the ocean itself, hiding it beneath an unending mat of white. To the east, deep purples and pinks show the shadow of the mountain, with Jupiter rising (we're really close to opposition right now). Venus has been glowing bright as the Sun disappears, but there's little time to waste as we head back down to the control room.
VISIR Observation Strategy
Padraig and I have had a plan each night for the sequence of VISIR 5-20 µm observations, with the ultimate aim of using Jupiter's ten-hour rotation to map out as many longitudes as possible, generating full maps of the planet in lots of different spectral settings. The benefit of mid-infrared observing is that we don’t need to wait until it gets dark to start taking data - the sky is always bright in the mid-IR. We can start during twilight, which ends around an hour after sunset. The telescope operator, Ben, would find a nearby bright star to guide on and to perform the "active correction" - deforming the main 8.2-m mirror with a series of actuators to correct for distortions caused by flexure and temperature changes. Then our support astronomer, Florian, would load up the first of our observation blocks for execution. These start with an acquisition template, a first glimpse of Jupiter so that we can move the telescope around a little, centring Jupiter, before starting the main observations.
Then it's a case of clicking go, and watching along as VISIR changes its filters and moves through the spectrum. We have to get the timing right, as Jupiter's 10-hour rotation is surprisingly fast when you're running against the clock. Sometimes we’ll break and do 20 minutes of "lucky imaging" (taking a video of the scene at 5 µm so I can then stack together only the sharpest frames, freezing the seeing), or 20 minutes of spectroscopy using VISIR's low-resolution mode. These tend to need the most user intervention, but for the rest of the time it's a matter of making sure that nothing is going wrong as the images are acquired.
Staying Awake
There's been some moments of excitement. On our first night, we were taking data at exactly the same time as Juno was executing its close flyby. That meant that we caught the flyby longitude (29W) perfectly, about 45 minutes before Juno flew over it on the terminator. We spotted that Juno flew over one of the 5-micron hotspots on the North Equatorial Belt, the first time that's been accomplished during the mission. We also had several opportunities to view the Great Red Spot in all its glory - the GRS is really unusual at the moment, after the passage of the South Tropical Disturbance. It looks like a spoon holding an egg. And the southern aurora was really vivid in all of our images (not so the northern aurora). These all brought gasps from the other astronomers in the room - I don't think any of them are used to being able to see such incredible data in the raw images.
Glenn and colleagues were also observing on Subaru for two of the nights, so at one point we had a live Skype conversation going as VISIR finished its Jupiter observations and COMICS picked them up from Hawaii - two of the world's best observatories working in tandem. We also had the Oxford service mode run on MUSE operating one night, so both UT3 and UT4 were looking in the same direction.
But there were some bad moments too. To get decent images of Jupiter we really want to be chopping the telescope by more than 45 arcseconds, but we've always been limited to 25 arcsec in the past. We tried to be cheeky and push this to 30 for some observations, which worked on the first night, but on subsequent nights it caused severe upsets. The M2 mirror lost all power and had to be rebooted - literally turning it off and on again to get it working, which meant we lost 30-60 minutes of time on the second night. Furthermore, the spectroscopy and burst mode imaging with VISIR proved to be exceedingly difficult. I spent most of my daytime hours trying to figure out the complexities of their reduction, coming back to the mountain the next night with suggestions for improvements. On the plus side, we'd be experts in VISIR data acquisition by the end of the run….
The first couple of nights spoiled us with excellent conditions - low wind, perfect seeing, incredibly low water vapour. Heading up on the mountain on the third night, we could see light cirrus overhead and the temperature had dropped. We were going all-guns-blazing until around 21:30, when the wind went above 12 m/s, meaning that we had a 'pointing restriction' and had to turn the telescope away from Jupiter. Above 18 m/s and they have to close, so we remained open, but executing service mode observations pointing away from the wind direction out of the north. I can’t tell you how frustrating it is to sit watching the minutes (and Jupiter longitudes) flick by while we can’t observe… but on the plus side, the support astronomers were generous. When the wind dropped at about 1am, they gave us the extra few hours back, and we continued with Jupiter until around 4am in the morning.
Night's End
At the end of each half-night we'd hand back over to the interferometry mode, which uses all four Uts. We'd collect our things and head outside to the Puntos, but it was astonishingly dark outside when the moon had set, providing us with a stunning view of the night sky, the Milky Way, and the Magellanic Clouds. Almost enough to want to stay up longer, until sleep deprivation took its hold…
