An analysis done in 3 weeks on 7 TeV data?

12 May 2010

Results on charge multiplicity obtained on 900 GeV and first 7 TeV data


When the ATLAS management kind of sprung this idea on the Minimum Bias group, the reactions were very varied: From ‘let’s not even talk about it’ and ‘this is not possible’ to ‘yes, let’s do it’ . But as biased as some pessimistic opinions were at this first moment, as fast did the scepticism change to shear enthusiasm - in some cases over night. “All physicists like challenges and it was interesting to see the evolution of people‘s thoughts on if we would make it or not. Most people changed their mind at least twice, some more” said Alison Lister.

The first measurements from 7 TeV proton-proton collisions recorded with the ATLAS detector were collected on March 30th using a minimum-bias trigger. So why not go from the 900 GeV Minimum Bias paper straight to the final analysis? Why this intermediate step repeating the previous analysis on 7 TeV data?

“There are many reasons for this, one of them involving us showing off to ourselves and the other experiments that we can do good physics in three weeks.” she explained with a grin. The goal was to be able to show some 7 TeV physics results already at the first conference after the data were taken and that was end of April at the DIS 2010, where Heather Gray presented these results.

“This time-scale was too tight for us to properly understand the tracking down to the transverse momentum we want to ultimately go to, so management suggested a ‘re-run’ of the 900 GeV analysis at higher energy. We thought about this for a while and decided that was not what we wanted to do as we didn’t want to go ‘backwards’ and remove all the changes that were made, but instead more ‘forwards’ and concentrate on doing work that would anyway have to be done for the final analysis.”

On the other hand, the decision to do an intermediate analysis redirected resources away from the final paper. But it paid off: “Of course it was very hard work and thus draining people, but the motivation is still high to get the next paper out which is keeping us going.” Alison said. The analysis followed very closely the strategy used in the 900 GeV data publication - that was presented in a previous ATLAS e-News article.

The final 7 TeV results are compared to these published values as shown on the above figure. Several improvements were made to both the track and the vertex reconstruction algorithms - the new tracking setup did what it was expected to, a bit better than the old one but essentially without changes in the region used by the analysis. The Inner Detector looks as good now as it did in December 2009.

Did any unforeseen problems get in the way?
“Luckily not! The only problem not encountered before was the issue of pile-up, which we worked on very closely with the vertexing experts. But that was expected, just not studied before as there wasn‘t any pile-up in the 900 GeV data.” says Alison. In fact everyone in the group knew that if there was one major problem found on the way, they would not be able to make the dead-line... “But it‘s not only luck, it‘s the hard work of many, many years of preparation that allowed this to happen”, adds Alison.

“From the results, we learnt that the ATLAS Monte Carlo tunes (MC09 and MC09c) do a much better job than expected at 7 TeV. It really is quite a jump in energy and to get some of the distributions agreeing to 10-15% is very good at this stage. Of course more work needs to be done to fine-tune the models to agree for all distributions but if there is anything to take away from this analysis I think it‘s how well the energy extrapolation actually worked!”, continues Alison.

Talking about learning, Alison said she learnt a few valuable lessons. “The most important one is that if one is organizing an analysis team that‘s on high alert, at the same time as attacking part of the analysis that was not required for the 900 GeV analysis... well it‘s too much for one person to do.”

The team worked until late every day (or early depending on your definition of day), getting very little sleep during those three weeks. A 5 minute-break counted as lunch-break and a Sunday off felt like holidays.

But no time for rest yet! The next publication is due very soon and aims to extend the measurement region. As one PhD-student put it: “Lucky that there is bad weather on the weekend, so we can work...”

 

Birgit Ewert

ATLAS e-News