Tomorrow I am traveling to CERN, where I have been invited to give a seminar at a meeting of the LHCb experiment. My talk will discuss the issue of the energy calibration of b-quark jets, a topic to which I have devoted a good part of my research time for the last thirteen years. The talk will of course be centred on the explanation of the analysis Julien Donini and I, together with a few colleagues, performed in CDF a few years ago, the search for Z boson decays to b-quark jet pairs.

I have in the pipeline a simpler article than the present one, where I explain why b-quark jets are special, and what makes them so important at hadron colliders (it is all about the Higgs, the top quark, and new physics searches, as you will learn in the next piece). Here, however, I wish to get away with just a slide show: I am offering to you a preliminary version of the slides of my seminar, with minimal commentary. Of course, many of you will find the slides unintelligible, and I cannot do much to prevent that. This is thus an "experts only" post, but if you hang around anyway you might find something interesting and understandable in it nonetheless... So here we go.

The Talk Slides



Slide three (well, yeah. Slides one and two are just a header and a summary...): the Tevatron collider. Here I want to give some information on the running performance of this incredible machine, and point out that if it continues to perform as it is presently doing, a first evidence of the Higgs boson might be at reach of CDF and DZERO.


Slide four: Just a descriptive summary of the CDF detector and its history.


Slide five: some detail of the important parts of CDF. From top to bottom the pictures show the silicon inner detector, a drawing of CDF with the inner region exposed (the tracking system is inside the red layer, which is the electromagnetic calorimeter), and the central drift tracker.


Slide six: the trigger system of CDF. Here the three-level architecture of the system is discussed. The trigger, in case you are wondering, is the system which decided online which collisions to filter and save to mass storage.


Slide seven: SVT is the heart of the Run II CDF trigger system. With it, b-quarks are collected with high efficiency. On the left a cartoon shows the function of the SVT associative memory bank, which matches patterns to the reconstructed track segments. The center plot shows the large sample of hadronic B meson decays, which among other things granted CDF the first measurement of B_s oscillations; on the right, the inventor of the device, Luciano Ristori. Incidentally Luciano has recently been awarded the Panofsky prize for the SVT, together with Aldo Menzione.


Slide eight: just an introduction on hadronic jets and their reconstruction. The cartoon on the right shows what happens when protons and antiprotons collide: a parton jet arises on one side of the event, and particles are created, which then interact and are destroyed in the calorimeter, where we measure the energy of the jet, and try to figure out what the energy of the originating quark was.


Slide nine: here I make a couple of points about the goals of jet clustering.


Slide ten: this is a roll-out display of the CDF calorimeter. In the grid -so-called eta-phi plane, basically a two-dimensional plane constructed on two angles describing the outgoing direction of the particles- the towers show the energy deposits from an energetic event. As you see there are clusters of energy scattered around, and different clustering algorithms "see" -interpret- these clusters as jets in different ways.


Slide eleven: some details of the algorithm of choice by CDF in Run II.


Slide twelve: an explanation of the algorithm that corrects the measured jet energy to make it match as closely as possible the energy possessed by the parent quark which originated it.


Slide thirteen: a few plots that describe how the response of the CDF calorimeter to jet energy is equalized for disuniformities as a function of angle of incidence (bottom left), and how the study of the calorimeter response to single particles (top right, single particles in data; and center right, pions from test-beam runs) allows to tune the simulation, which is then used to get an absolute correction function (bottom right).


Slide fourteen: this graph shows the energy behaviour of the biggest sources of systematic uncertainty on jet energy measurement. As you can see, at low energy the out-of-cone uncertainty -the unknown fraction of energy that flows out of a cone of fixed R=0.7 radius- dominates all others.


Slide fifteen: this introduces the issue of b-jets, and why they are different. The graph demonstrates that the fraction of energy of a b-jet which is measured in the calorimeter (in blue or red) is smaller than for generic jets (the black curve).


Slide sixteen: some more detail on the various sources of uncertainty from the b-jet energy scale, and the impact on the uncertainty on the measured top quark mass (which uses the b-jet energy measurement, of course).


Slide seventeen: still some detail on generator-level issues -herwig and pythia, the two simulation programs, possess some tunable parameters which have an impact on the measured b-jet energy.


Slide eighteen: the way the b-quark originates a stream of hadrons as it "fragments" is well-known from studies performed at LEP; however, some uncertainties from the modeling of fragmentations still exist.


Slide nineteen: the pattern of possible decays of B hadrons has a clear effect on the fraction of energy that we can measure from a b-quark jet: that is because we do not measure neutrinos -which escape unseen- and also muons provide a small calorimeter response; electrons, on the other hand, over-compensate.


Slide twenty: the Z decay to b-quark pairs was found already in Run I at CDF -I should know it, I did it by myself for my PhD thesis! This is just an introductory slide of the problem. On the top right you can see a figure of historical value: it is the first publication which measured the top quark mass, by CDF in 1994 -when the top quark had not been discovered officially yet. On the picture below the plot, you can see the detector which made it possible, the SVX -a four-layer barrel of silicon microstrip sensors.


Slide twentyone: this shows the signal I found in 1998 from the 100 inverse picobarns of Run I data. The plots on the right show two different means of extracting the signal: a counting experiment (top), and a unbinned likelihood fit (bottom).

The other twentyone slides will be posted tomorrow...