Searching for the unexpected
This plot shows the largest discrepancy between data and theoretical predictions. The inset plot is a close up of the larger one. This particular plot adds the transverse momentum of an electron and muon with opposite electrical charge and "missing transverse momentum" that usually indicates a neutrino. The colored areas are the predictions, while the points are the measurements. The discrepancy is most likely due to an inadequate modeling of how muons are measured and is not a discovery.
Sherlock Holmes solved crimes by looking for the thing out of place: a smudge of dust on a finely pressed suit; cuts on the side of a leather loafer. A case could change, depending on what he saw that was unexpected.
In the very finest Sherlockian tradition, DZero scientists have been carefully studying data, looking for something unexpected. At first, you might wonder why this is newsworthy. After all, that's what scientists do, right? However, most searches for new physical phenomena start out with scientists looking to see if a particular theoretical prediction is true. To do this, particle physicists carefully study the prediction and decide how to best record data in which that process occurred. We then discard the data that were caused by other physical processes. This strategy allows us to precisely study a particular theoretical prediction.
A problem can arise if there is some new physical process present in the data but the theory explaining it hasn't been invented yet. What then? In order to make sure that we do not miss something that is there, we can employ a different technique called a model independent search, or MIS. In this case, we simply make many data plots and compare them with Standard Model predictions. Plots in which the data and Standard Model predictions disagree could be the sign of something unexpected.
Using two programs, called Vista (pioneered by CDF) and Sleuth (pioneered by DZero), we made plots using very violent collisions, the collisions where new phenomena are most likely to be found. We tested this approach by comparing our data to a theory from which top quarks were removed, essentially rediscovering the top quark. We were happy to see that this technique did indeed find top quarks.
We repeated this process for nearly 10,000 data plots, comparing each of them to the full Standard Model. Almost all plots agreed with theory. The few that disagreed were more closely inspected. It became clear that the discrepancy stemmed from having an imperfect simulation of the DZero detector.
(Scientists knew about this issue.)
So, no hint of new physics was found. But this study makes us much more confident that it's not just because we're looking in the wrong place.
This team was responsible for this interesting analysis.
||The computing support team is crucial for smooth data analysis. Adam Lyon and Robert Illingworth handle day-to-day operations. Alan Jonckheere and Vladimir Sirotenko compile the reconstruction and simulation programs. Jim Linnemann is responsible for the experiment data bases.