Last weekend, I had the pleasure to be shift leader for ATLAS. It was a real pleasure for many reasons: being right in the middle of the action, surrounded by an international team of enthusiastic and dedicated people, and taking part in great teamwork. The shift crew (about ten people plus dozens of experts on call) must keep the detector running smoothly, tackling every problem, big or small, as fast as possible.
Data was coming in at a high rate and all sub-detectors were humming nicely. Not a glitch in hours so we were getting slightly sleepy nearing the end of the shift around 22:00. So when a colleague from the trigger system (the system that decides which events are worth keeping) called to inquire about recurrent splashes of data, I was rather puzzled.
I quickly went around, asking a few shifters to check their system. Nobody had a clue. Then I took a closer look at this plot that I had not scrutinized before since everything was so seamless.
The two lower curves in beige and green show the instantaneous luminosity measured by the two largest detectors operating on the Large Hadron Collider (LHC), CMS and ATLAS. This is a measure of how many collisions are happening per second in each experiment from the two beams of protons circulating in opposite direction in the LHC tunnel. If you look closely at these curves, they both have small dips at regular intervals. Since both ATLAS and CMS were registering these dips, it had to be coming from a common source, the LHC.
So I called the LHC control room to find out what was happening. “Oh, those dips?”, casually answered the operator on shift. “That’s because the moon is nearly full and I periodically have to adjust the proton beam orbits.”
This effect has been known since the LEP days, the Large Electron Positron collider, the LHC predecessor. The LHC reuses the same circular tunnel as LEP. Twenty some years ago, it then came as a surprise that, given the 27 km circumference of the accelerator, the gravitational force exerted by the moon on one side is not the same as the one felt at the opposite side, creating a small distortion of the tunnel. Since the moon’s effect is very small, only large bodies like oceans feel its effect in the form of tides. But the LHC is such a sensitive apparatus, it can detect the minute deformations created by the small differences in the gravitational force across its diameter. The effect is of course largest when the moon is full or during the new moon, when the sun and the moon combine their tidal forces for being all aligned with the earth. But the same happens twice a day like the tides and the operators must correct for it.
What came as a surprise to me was to witness the dynamic aspect of it. As the moon was rising in the sky, the force it exerted changed ever so slightly, but even these infinitesimal changes were big enough to require a periodic correction of the orbit of the proton beams in the accelerator to adapt to a deformed tunnel. And each time the operator corrected the orbit, he also had to reoptimize the beams position to maintain the best collision rate by doing a small scan of the two beams to realign them, which caused the dips in the luminosity plot.
Other surprising disturbances were also observed in the LEP days like one that appeared every day at fixed times. It took months and a train company strike to figure it out. These perturbations were created by the passage of the fast train linking Geneva to Paris, the TGV, since it releases a lot of electrical energy into the ground. The LHC is also sensitive to the hydrostatic pressure created by the water level in nearby Lake Geneva that also deforms the tunnel shape.
Life is full of surprises when dealing with such a sensitive piece of equipment.
To be alerted of new postings, follow me on Twitter: @GagnonPauline or sign-up on this mailing list to receive and e-mail notification.
Addition to the initial post
Since there have been so many questions on this, let me try to be clearer.
Why is this effect bigger during the full moon?
As Paolo explained in his comment, it’s because when the moon is full (just like when the moon is new as many pointed out) the tidal effect due to the sun reinforces that due to the moon (they are all aligned). So it is not because the moon is closer to the earth, this distance being rather fixed. It is just that the sun and moon combine forces on full and new moon.
Sorry for not mentioning the new moon in the initial blog, but it works just like with the tides: full moons and new moons bring the biggest tides, this is what my comment meant. But I agree: it was misleading.
Here is the graph used by the LHC operators to compensate the accelerator displacement. Each up and down represents a day, with a high and a low tides. The external modulation comes from adding in the position of the moon with respect to the earth and sun during the month. Since the moon takes 28 days to go around the earth, twice a month it is aligned with the sun. This occurs at full moon and at new moon. This is also when the tides are the strongest.
Is this effect there only on full moon?
No. The LHC operators have to correct for this effect every day, following the tides. Just as Christopher Grams, Phil Plait and others suggested. But on full moon, the effect is bigger. When I observed it myself during my shift on Saturday June 2nd, it was made even more visible because the operator took the opportunity to optimize the beam position every time he applied a correction to the orbit since it was a big correction.
What is moving?
The accelerator is moving, not the proton beams. The accelerator is moving with the earth crust. We all know the moon creates tides. This happens because the moon pulls on the ocean as it circles around the earth. The earth crust feels the same pull but since water is much easier to move than the earth crust, almost nobody ever notices the small earth deformations. But the LHC operators do because the accelerator is both very large and very precise. They see the protons moving outside of the accelerator orbit, as the accelerator shifts.
To recap: The LHC accelerator was pulled away from the proton beams that circulate inside it by the tidal forces. The operator corrected the orbit to bring back the protons in the center of the beam pipe. Each time he did that, since it was a big correction, he also swept the beams with respect to each other to find the place of maximum collisions. This sweeping of the beams caused small dips in the luminosity plot (the measure of the collisions rate), making it very visible.I hope this helps.
Thanks to all of you who offered clarifications or asked the right questions!