1 of 64

Experimental physics at the LHC:

from collisions to results

Kārlis Dreimanis

Baltic School of High-Energy Physics�and Accelerator Technologies

​

Palanga, Lithuania

August 10th, 2023

Supported by, Latvian Council of Science, State Research Programme project: VPP-IZM-CERN-2022/1-0001

​

2 of 64

What is experimental HEP?

2

  • Broadly speaking, experimental HEP has two tasks:

1) To measure precisely the various parameters and predictions of the Standard Model (SM);

2) To search for signals of yet-undiscovered New Physics, lurking Beyond the SM.

​

  • But why?
    • Particle physics seeks to answer the most fundamental questions about the Nature�of our Universe, by studying its elementary building blocks - particles (and their fields).

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

3 of 64

What is experimental HEP?

3

  • Broadly speaking, experimental HEP has two tasks:

1) To measure precisely the various parameters and predictions of the Standard Model (SM);

2) To search for signals of yet-undiscovered New Physics, lurking Beyond the SM.

​

  • But why?
    • Particle physics seeks to answer the most fundamental questions about the Nature�of our Universe, by studying its elementary building blocks - particles (and their fields).

​

​

“Why climb Mount Everest?”� “Because it’s there!”� (George Mallory)

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

4 of 64

What is experimental HEP?

4

  • Broadly speaking, experimental HEP has two tasks:

1) To measure precisely the various parameters and predictions of the Standard Model (SM);

2) To search for signals of yet-undiscovered New Physics, lurking Beyond the SM.

​

  • But why?
    • Particle physics seeks to answer the most fundamental questions about the Nature�of our Universe, by studying its elementary building blocks - particles (and their fields).

​

​

“Experiment without theory is blind, but theory�without experiment is a mere intellectual play”� (Immanuel Kant)

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

5 of 64

The Standard Model

5

  • The Standard Model consists of 27 unique elementary particles [experimentally speaking…]:

​

    • 6(+6) (anti-)quarks;
    • 3(+3) charged (anti-)leptons;
    • 3 neutral leptons;
    • 6 force carriers;

​

  • Since the discovery of the Higgs,�its proposed particle content is complete!

​

  • However, as a most fundamental�theory of Nature, the SM is still lacking!

​

  • To study the SM experimentally, we need to study all of its constituent particles!�To do that we must first create them in particle colliders!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

6 of 64

When beams in a collider cross …

6

  • In an e+e- collider (such as LEP) life is “simple”:

​

    • The entire particle participates in the collision;
    • The full energy is converted into pure energy;
    • Collisions can be tuned for a specific resonances:
      • bb{bar} (SuperKEKB*);
      • Z, WW, ZZ (LEP);
      • ZH, tt{bar} (FCC-ee**);
    • Clean signatures!

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

* using asymmetric {energy} beams.

** hopefully, all of you will contribute to the creation of this!

Image source

Image source

7 of 64

When beams in a collider cross …

7

  • In an e+e- collider (such as LEP) life is “simple”:

​

    • The entire particle participates in the collision;
    • The full energy is converted into pure energy;
    • Collisions can be tuned for a specific resonances:
      • bb{bar} (SuperKEKB*);
      • Z, WW, ZZ (LEP);
      • ZH, tt{bar} (FCC-ee**);
    • Clean signatures!

​

  • But e+e- colliders have a drawbacks:
    • Greater synchrotron radiation losses → energy limit!
    • Cannot be solved simply by stronger bending magnets → need larger radii!�(but even then, P goes as 1/r2 and still as γ4)

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

* using asymmetric {energy} beams.

** hopefully, all of you will contribute to the creation of this!

Energy loss per revolution:

8 of 64

When beams in a collider cross …

8

  • In an e+e- collider (such as LEP) life is “simple”:

​

    • The entire particle participates in the collision;
    • The full energy is converted into pure energy;
    • Collisions can be tuned for a specific resonances:
      • bb{bar} (SuperKEKB*);
      • Z, WW, ZZ (LEP);
      • ZH, tt{bar} (FCC-ee**);
    • Clean signatures!

​

  • But e+e- colliders have a drawbacks:
    • Greater synchrotron radiation losses → energy limit!
    • Cannot be solved simply by stronger bending magnets → need larger radii!�(but even then, P goes as 1/r2 and still as γ4)

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

* using asymmetric {energy} beams.

** hopefully, all of you will contribute to the creation of this!

Energy loss per revolution:

9 of 64

When beams in a collider cross …

9

  • Collisions in a hadron collider are much more complicated:
    • Only a fraction of the particle participates in the collision;
    • Extremely messy environment;
    • Impossible to tune the collision energy for certain resonances.

​

  • But hadron colliders have their up-sides:
    • Low synchrotron radiation losses:
      • Higher achievable centre-of-mass energies;
      • Increased luminosity → higher statistics;
      • Further energy gains available with further magnet development!

​

  • Hadron collision process can be broken down to its main parts:
      • Main hard scattering
      • Underlying event
      • Collision product decay
      • Hadronization and hadron decays

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source

10 of 64

Interlude - PDFs

10

  • Protons are constituent particles;

​

  • We tend to think them as consisting of 3 quarks (uud),�but in truth these 3 quarks are just the valence quarks,�held together by gluons, whilst swimming in a quark-gluon sea!

​

  • The greater the energy with which one probes the proton (ie. the greater the Q2),�the greater the chance of encountering a non-valence constituent of the proton!

​

  • These constituents are called partons and their distributions�(likelihood of being encountered with a given Bjorken-x at a given Q2)�are governed by the parton distribution functions (PDFs).�

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

11 of 64

Interlude - PDFs

11

​

  • PDFs can be measured best at ep experiments,�where e deeply probes the proton;

​

  • PDFs measured at given Q2 can then be evolved to other�Q2 values using the DGLAP equations;� (and I’ll leave it at that here!)

​

  • But, importantly, this leads to LHC being,�essentially, a gluon-gluon collider!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: PDG 2016

12 of 64

When beams in a collider cross …

12

  • Main (hard) scattering event:
    • Large momentum transfer (Q2);
    • Perturbatively calculable!

​

  • Secondary (underlying) event/-s:
    • Small Q2;
    • Impossible to calculate using perturbative methods!
    • Must be provided by the experiment!

​

  • Our detectors can see only the green and yellow!
    • Hadron decays;
    • Photon emission;

[and prompt and non-prompt leptons (not shown)].

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source

13 of 64

“Detecting particles” at LHC

13

  • The aim of the LHC experiments�is to study exotic particles.

​

  • In HEP, exotic often means more massive.

​

  • Such particles are extremely fickle - they�decay into more mundane particles very quickly:
    • Z,W±: τ~10-25s;
    • t: τ~10-25s;
    • H: τ~10-22s.

​

  • These particles travel infinitesimally small distances�from the primary vertex (PV) before decaying.

​

  • Example: the LHC beam-spot at ATLAS

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

14 of 64

“Detecting particles” at LHC

14

  • The aim of the LHC experiments�is to study exotic particles.

​

  • In HEP, exotic often means more massive.

​

  • Such particles are extremely fickle - they�decay into more mundane particles very quickly:
    • Z,W±: τ~10-25s;
    • t: τ~10-25s;
    • H: τ~10-22s.

​

  • These particles travel infinitesimally small distances�from the primary vertex (PV) before decaying.

​

  • The beam-spot contains all the pp interactions within it�(both transversely & longitudinally);

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

~10um

~10um

LHC beam-spot (head-on/transverse view)

15 of 64

“Detecting particles” at LHC

15

  • The aim of the LHC experiments�is to study exotic particles.

​

  • In HEP, exotic often means more massive.

​

  • Such particles are extremely fickle - they�decay into more mundane particles very quickly:
    • Z,W±: τ~10-25s;
    • t: τ~10-25s;
    • H: τ~10-22s.

​

  • These particles travel infinitesimally small distances�from the primary vertex (PV) before decaying.

​

  • The velocity at which any this H would have to travel�to even escape the beam-spot is 0.9999999999999999c;

(protons accelerated by LHC to 0.999999991c are ~100x less massive!)

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

~10um

~10um

LHC beam-spot (head-on/transverse view)

Example�Higgs event!

Image source

16 of 64

“Detecting particles” at LHC

16

  • In fact, it is even worse that!

​

  • LHC beams are not perfectly uniform nor is�the beam position always the same;

​

  • The detectors must be placed a safe�distance from the beam to avoid costly damage;

​

  • At the LHC, this safe distance is ~3 cm;

​

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

17 of 64

“Detecting particles” at LHC

17

  • In fact, it is even worse that!

​

  • LHC beams are not perfectly uniform nor is�the beam position always the same;

​

  • The detectors must be placed a safe�distance from the beam to avoid costly damage;

​

  • At the LHC, this safe distance is ~3 cm;

​

  • Thus, the average distance a particle must travel to be directly detected is not this …

​

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

~10um

~10um

18 of 64

“Detecting particles” at LHC

18

  • In fact, it is even worse that!

​

  • LHC beams are not perfectly uniform nor is�the beam position always the same;

​

  • The detectors must be placed a safe�distance from the beam to avoid costly damage;

​

  • At the LHC, this safe distance is ~3 cm;

​

  • Thus, the average distance a particle must travel to be directly detected is not this …

​

… but rather something like THIS!

​

Which is not really viable!

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

~10um

~10um

19 of 64

“Detecting particles” at LHC

19

  • We never actually detect these heavy particles!

​

  • We only detect their decay products!�(and decay products of decay products of …)

​

  • Here, ttH production via a gluon-gluon collision process at a hadron collider (like LHC) is shown;

​

  • t quarks and H boson are produced;

​

  • These decay in W bosons, 𝝉 leptons and b quarks;

​

  • These then further decay into various leptons and hadrons.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source

20 of 64

“Detecting particles” at LHC

20

  • Thankfully, some exotic particles are more compliant:
    • 𝜏 leptons: τ~10-13s;
    • D mesons: τ~10-13s;
    • B mesons: τ~10-12s.

​

  • At the LHC, B mesons can travel as far as ~1 cm!

​

  • One detector at the LHC can, in theory,�“see” these particles directly;

​

  • LHCb’s VELO detector can be moved-in towards�the collision point to a distance of 5 mm;

​

  • VELO stands for VErtex LOcator, as the detector specialises in precise reconstruction of secondary vertices.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: LHCb collaboration

21 of 64

Vertexing at LHCb

21

  • In truth, the tracks are still reconstructed from the decay product interaction with the detector;

​

​

​

​

​

​

​

​

​

​

​

​

​

​

  • But the closer to can start the decay product detection, the better your vertex resolution!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: LHCb collaboration

22 of 64

Vertexing at the GPDs

22

  • For ATLAS and CMS, the first detecting layer is 3.3 and 2.9 cm from the beam-spot, respectively.

​

  • The challenge of precise secondary vertex�reconstruction is even higher!

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source

23 of 64

Vertexing at the GPDs

23

  • For ATLAS and CMS, the first detecting layer is 3.3 and 2.9 cm from the beam-spot, respectively.

​

  • The challenge of precise secondary vertex�reconstruction is even higher!

​

  • Especially, given the�pile-up at these detectors!

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: CMS collaboration

24 of 64

Vertexing at the GPDs

24

  • For ATLAS and CMS, the first detecting layer is 3.3 and 2.9 cm from the beam-spot, respectively.

​

  • The challenge of precise secondary vertex�reconstruction is even higher!

​

  • Especially, given the�pile-up at these detectors!

​

  • Here, the nPV is only ~100;

​

  • At HL-LHC it will�be up to 200!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: CMS collaboration

25 of 64

Track reconstruction

25

  • Precise track reconstruction is vital for good quality physics results.

​

  • Trackers are usually* the nearest-to-beam instrumentation at the LHC;

​

  • Trackers are immersed in a B-field,�which bends the particle paths:

​

  • The bending radius is used to measure charged particle momentum.

​

  • An ideal tracker is infinitesimally thin to avoid ‘corrupting’�the measured particle path through material interactions;

​

  • Ideally, the tracker would also be infinitely granular, but in reality, understanding of your the B-field can be the driver of the�momentum uncertainty with a modern pixel tracker.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

* muon stations are, essentially, also trackers, but are the farthest.

26 of 64

Calorimetry

26

​

  • As the name suggests, calorimeters aim to�measure the total calories (the total energy)�of the incident particles;

​

​

  • Thus, completely opposite to trackers, calorimeters�want to maximally ‘corrupt’, or stop,� the detected particles;

​

​

  • Usually split into ECAL and HCAL to allow�for a separation between e/𝛾�and the hadronic particles.�

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

27 of 64

Particle Identification

27

​

  • Particle Identification (PID) is usually a combined effort of�multiple detector layers and technologies.

​

  • As before, ECAL/HCAL split allows to separate e/𝛾�from the hadrons.

​

  • Adding the tracking information allow to split electrons from 𝛾�and charged hadrons from neutral hadrons.

​

  • Finally, placing muon stations as the outermost layer,�allows to identify muons (and their tracks in the tracker).

​

  • But this approach struggles to separate, various hadronic particles�from each other → particularly relevant for flavour physics at LHCb!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: https://doi.org/10.1016/j.nima.2011.03.009

28 of 64

Particle Identification

28

  • Use ingenuity - Cherenkov light!

​

  • Cherenkov light is produced when a charged�particle moves in a medium faster than the�speed of light in that medium;

​

  • It creates a light-cone with an angle�related to the particle’s velocity:

​

​

​

​

  • Combining this information with the momentum information�from the tracking system, one can extract the particle’s mass!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: LHCb collaboration

29 of 64

Correcting the measurements

29

  • All particles are detected with a given efficiency, 𝜀det;

​

  • Various parameters need to be corrected for,�like reconstruction efficiency, ID efficiency, etc. …

​

  • This is usually done by binning your distributions�against some relevant variable:

​

    • Total momentum, p;
    • Transverse momentum, pT;
    • Pseudorapidity, η;

… etc.

… and finding the 𝜀det in each bin.

​

  • Another tricky correction is accounting for bin migrations.

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: CMS collaboration

30 of 64

Monte-Carlo simulations

30

  • A vital part of particle physics experiment is the Monte-Carlo (MC) simulation;�MC, broadly speaking, is a theory prediction of particle behaviour.

​

  • MC generator precision is limited by the order to which�the particle interactions are simulated:
    • Leading order (LO), a.k.a. tree-level;
    • Next-to-leading order (NLO);
    • Next-to-next-to-leading order (NNLO);

…

  • The current state-of-the-art is N3LO;

​

  • There are many MC generators in use:
    • PYTHIA;
    • Powheg;
    • Sherpa;

… among others.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

31 of 64

Monte-Carlo simulations

31

  • MC simulation on its own gives us the physical�processes at the truth or generator level.

​

​

  • A particle collision as it truly happened,�from the hard scattering itself, to the�potentially observable physics�objects.

​

​

  • Notice, it also includes non-perturbative�soft components, such as the MPIs.

​

  • Generator level MC is very powerful, but not�immediately useful for the experiment on its own.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: https://arxiv.org/pdf/2203.11601.pdf

32 of 64

Detector simulation

32

​

​

  • Detecting == interfering!

​

​

  • We must have a complete (!) understanding�of what the generated particles encounter;

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

33 of 64

Detector simulation

33

​

​

  • Detecting == interfering!

​

​

  • We must have a complete (!) understanding�of what the generated particles encounter;

​

​

  • We use a tool* called Geant4, to fully simulate�our detectors, including both active and passive�materials within them.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: Geant4 collaboration

* other tools, such as FLUKA are also used for more specialised needs.

34 of 64

Combination: full simulation!

34

  • To get to how nature shows-up in our detector,�one must combine the physics generator�with the detector simulation …

… and add digitisation and other steps …

​

  • All HEP experiments have their own huge�software packages, painstakingly built,�continuously updates and improved:

​

    • ATLAS Athena;
    • CMS CMSSW;
    • LHCb Gauss.

​

​

  • Finally, at the end of all this, we arrive�at our reconstruction or detector level�MC simulation.

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: CMS collaboration

35 of 64

MC simulation versus collision data

35

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

What information is readily accessible?

36 of 64

MC simulation versus collision data

36

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

✔

What information is readily accessible?

37 of 64

MC simulation versus collision data

37

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

✔

✔

What information is readily accessible?

38 of 64

MC simulation versus collision data

38

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

✔

✔

✔

What information is readily accessible?

39 of 64

MC simulation versus collision data

39

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

✔

✔

✔

Generator / truth level�information

╳

What information is readily accessible?

40 of 64

MC simulation versus collision data

40

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

✔

✔

✔

But the truth level information in real collision�data is exactly what we seek!

�Catastrophe!!!

Generator / truth level�information

╳

What information is readily accessible?

41 of 64

MC simulation versus collision data

41

  • To avoid the catastrophe, we must correct the measured distributions for detector effects;

​

  • We must also understand our backgrounds very well;

​

  • In fact, the vast majority of time it takes to complete a physics analysis at LHC goes to understanding backgrounds, correcting your distributions, etc. …

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

✔

✔

✔

Generator / truth level�information

╳

What information is readily accessible?

42 of 64

MC simulation versus collision data

42

Generator / truth level�information

  • To avoid the catastrophe, we must correct the measured distributions for detector effects;

​

  • We must also understand our backgrounds very well;

​

  • In fact, the vast majority of time it takes to complete a physics analysis at LHC goes to understanding backgrounds, correcting your distributions, etc. …�� … and convincing your collaborators that you have done everything correctly!

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

✔

✔

✔

╳

What information is readily accessible?

43 of 64

Unfolding

43

  • The knowledge of MC reco, MC truth and data reco distributions can be used to recover data truth,�via a process called detector unfolding.

​

  • Consider a reconstructed distribution, Mr(MC), and a true distribution, Mt(MC);�these can be related to each other through some response matrix R via Mr(MC)=RMt(MC).

​

  • Hence, by definition, the relation Mt(MC)=R-1Mr(MC) also holds true; R-1 is what one calls the unfolding matrix U, which can be obtained directly from MC and applied to Mr(data) to get find Mt(data)=UMr(data)

​

  • For MC, the link between truth and reco particles can be retained, ie. one can identify:

​

    • Correctly reconstructed particles (exist at both the truth and reco levels);
    • Missed particles (exist only at the truth level);
    • Ghost particles (exist only at the reco level).

​

  • Using this information one can easily construct both R and U.

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

44 of 64

Unfolding

44

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

Build R or U

Validate reco.�level dist.

Apply R-1 or U

45 of 64

Unfolding

45

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Monte-Carlo simulation

Collision data

Detector / reconstruction level information

Generator / truth level�information

Detector / reconstruction level information

Generator / truth level�information

Build R or U

Validate reco.�level dist.

Apply R-1 or U

46 of 64

Unfolding example

46

  • Multiplicity distributions in (e⨯η) [LHCb unpublished]:

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: https://cds.cern.ch/record/2304736/

MC rec.

Data rec.

MC truth

Data unf.

47 of 64

Unfolding example

47

  • Multiplicity distributions in (e⨯η) [LHCb unpublished]:

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: https://cds.cern.ch/record/2304736/

MC rec.

Data rec.

MC truth

Data unf.

48 of 64

The goal - physics

48

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

Image source: CMS collaboration

49 of 64

The goal - physics

49

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: ATLAS collaboration

50 of 64

The goal - physics

50

  • There are a vast array of various�physics measurements to be done�to further validate (or finally discredit!)�the SM!

​

    • Particle production cross-sections;
    • Particle decay channels and widths;
    • Particle masses;
    • Coupling constants;
    • Angular distributions;

… etc. …

​

  • But we also perform spectroscopy�(bump-hunting) and many (many!)�NP searches !

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Image source: CMS collaboration

51 of 64

The goal - physics

51

  • Spectroscopy can be exciting, but has very little discovery potential without a considerable jump in collision energies (or theory suggestion/model for some odd particle-combinations!).

​

  • Alternatively, we need MUCH MORE data!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

Source: LHCb collaboration

52 of 64

Example

52

An event recorded by the CMS experiment

​

  • Two reconstructed high-pT photons:
    • Pointing towards a single vertex;
    • Combined mass ~124.7 GeV/c2.

​

  • What was, more likely than not, produced in this collision event?

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

53 of 64

Example

53

An event recorded by the CMS experiment

​

  • Two reconstructed high-pT photons:
    • Pointing towards a single vertex;
    • Combined mass ~124.7 GeV/c2.

​

  • What was, more likely than not,�produced in this collision event?

​

Let’s see!

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

https://doi.org/10.1140/epjc/s10052-014-3076-z

54 of 64

Example

54

An event recorded by the CMS experiment

​

  • Two reconstructed high-pT photons:
    • Pointing towards a single vertex;
    • Combined mass ~124.7 GeV/c2.

​

  • What was, more likely than not,�produced in this collision event?

​

Let’s see!: at 124.7 GeV/c2 …

​

… the # of SM background events ~1.75x103;

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

https://doi.org/10.1140/epjc/s10052-014-3076-z

55 of 64

Example

55

An event recorded by the CMS experiment

​

  • Two reconstructed high-pT photons:
    • Pointing towards a single vertex;
    • Combined mass ~124.7 GeV/c2.

​

  • What was, more likely than not,�produced in this collision event?

​

Let’s see!: at 124.7 GeV/c2 …

​

… the # of SM background events ~1.75x103;

… the # of Higgs decay events ~0.2x103;

​

​

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

https://doi.org/10.1140/epjc/s10052-014-3076-z

56 of 64

Example

56

An event recorded by the CMS experiment

​

  • Two reconstructed high-pT photons:
    • Pointing towards a single vertex;
    • Combined mass ~124.7 GeV/c2.

​

  • What was, more likely than not,�produced in this collision event?

​

Let’s see!: at 124.7 GeV/c2 …

​

… the # of SM background events ~1.75x103;

… the # of Higgs decay events ~0.2x103;

​

  • This single event is ~9x more likely to be a random SM background event!

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

https://doi.org/10.1140/epjc/s10052-014-3076-z

57 of 64

The goal - finding the answers!

57

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

  • The SM may by self-consistent, but it still has internal unanswered questions;
    • What is the origin of the specific masses of the fermions?
    • Why are said masses so different between generations?
    • Why are there (and, indeed, are there?) only three generations of fermions?
    • Are the neutrinos Majorana or Dirac; is their mass-hierarchy normal or inverted?*

​

  • In total the SM has 19 (26*) free parameters:
    • Irreducible sets of 7 free parameters from the electroweak sector;
    • 6 quark masses and the 3 angles and 1 complex phase of the CKM matrix;
    • QCD renormalization scale and the Θ-parameter, arising from the strong CP problem;
    • 3 masses and 4 parameters from the PMNS matrix from the neutrino sector*;
  • These must be experimentally determined and input into the SM!

* - technically the neutrino sector is already an extension to the SM!

58 of 64

LHC Run 3 - a great time to join hep-ex

58

karlis.dreimanis@rtu.lv Baltic School of High-Energy Physics and Accelerator Technologies, Palanga, Lithuania, August 10th, 2023

You are here

LHC: Runs 1 to 3 will total ~300 fb-1

HL-LHC: Runs 4 to 5 will total ~3000 fb-1 !!!

59 of 64

F particle

59

karlis.dreimanis@cern.ch

Source: ATLAS collaboration

  • In 2015, ATLAS reported a 3.6σ excess at 750 GeV in the di-photon spectrum!
  • Could this have been a heavy Higgs?�Could this be proof of Supersymmetry?
  • CMS data reported an excess at 2.6σ!
  • Surely this is a major discovery!

​

60 of 64

F particle

60

karlis.dreimanis@cern.ch

  • In 2015, ATLAS reported a 3.6σ excess at 750 GeV in the di-photon spectrum!
  • Could this have been a heavy Higgs?�Could this be proof of Supersymmetry?
  • CMS data reported an excess at 2.6σ!
  • Surely this is a major discovery!
  • Prompted 500 theory papers on arXiV in the span of ~2 weeks! Massive hype in media!

​

61 of 64

F particle

61

karlis.dreimanis@cern.ch

  • In 2015, ATLAS reported a 3.6σ excess at 750 GeV in the di-photon spectrum!
  • Could this have been a heavy Higgs?�Could this be proof of Supersymmetry?
  • CMS data reported an excess at 2.6σ!
  • Surely this is a major discovery!
  • Prompted 500 theory papers on arXiV in the span of ~2 weeks! Massive hype in media!

​

  • Alas, with more data taken, the bump disappeared entirely!
  • We must be careful with our announcements!

​

62 of 64

A true discrepancy?

62

  • Measurement of the ratio of B-meson decays into a kaon+muons and a kaon+electrons;
  • Corrected for mass should be == 1;
  • Multiple measurements pointing in the same direction; >3σ significance;

​

  • This is a tentative evidence of BSM physics!�Lepton non-universality;

​

  • Must be cautious as this could also�go away!

karlis.dreimanis@cern.ch

Source: CERN courier

  • Personal opinion - tantalising! A genuine chance of New Physics!�(but rumours are swirling that this is less prominent in higher q2 regions)

63 of 64

A true discrepancy?

63

  • Measurement of the ratio of B-meson decays into a kaon+muons and a kaon+electrons;
  • Corrected for mass should be == 1;
  • Multiple measurements pointing in the same direction; >3σ significance;

​

  • This is a tentative evidence of BSM physics!�Lepton non-universality;

​

  • Must be cautious as this could also�go away!

karlis.dreimanis@cern.ch

Source: CERN courier

  • Personal opinion - tantalising! A genuine chance of New Physics!�(but rumours are swirling that this is less prominent in higher q2 regions)

'E's not pinin'! 'E's passed on! This *ANOMALY* is no more! He has ceased to be! 'E's expired and gone to meet 'is maker! 'E's a stiff! Bereft of life, 'e rests in peace! If you hadn't nailed 'im to the *ARXIV* 'e'd be pushing up the daisies! 'Is metabolic processes are now 'istory! 'E's off the twig! 'E's kicked the bucket, 'e's shuffled off 'is mortal coil, run down the curtain and joined the bleedin' choir invisible!! THIS IS AN EX-*ANOMALY*!

64 of 64

Thank you