Spring 2021 Semester

1/28/2021 - Reformatting Logbook and Project Choice Research

• • Reformatting Logbook

  • Changing page name

    • • Click to edit page (top right)

• • Creating new page

    • • Click to create a new page (top right)

• • Moving page

    • • Click More Options 

• • Project Choice Research

  • Research project choices and write a summary paragraph for each

    • Preferably ASAP

    • Definitely before next week

1/29/2021 - Project Choices Research 

2/4/2021 - Group Assignment

• • Assigned: Learn how to build and manage one of the computing clusters of the Worldwide LHC Computing Grid (WLCG)  based at UMD

  • Prepare presentation; research terms: LHC Research Grid, UMD tier 3 research grid

  • Discussion of free will

    • Casimir effect

    • https://en.wikipedia.org/wiki/Casimir_effect

    • Wolfram physics project

    • https://writings.stephenwolfram.com/2020/04/finally-we-may-have-a-path-to-the-fundamental-theory-of-physics-and-its-beautiful/

    • Everything may be determined from the beginning, but the result is not something any of us may know. 

    • So, I choose to believe in this illusion called free will. Even if the result is set in stone, I can only continue to do what I do to see it through.

  • Research group discussion:

    • Useful links:

      • UMD Cluster Homepage:

      • https://sites.google.com/a/physics.umd.edu/umdt3/home

      • UMD Cluster Monitoring page:

      • http://hepcms-hn.umd.edu/~jabeen/Monitoring

      • CERN LHC Computing Grid

      • https://home.cern/science/computing

      • CERN Computing CLuster Tier Listing

      • https://cds.cern.ch/record/1997396?ln=en

    • Topics to research:

      • CentOS

      • Apache Hadoop

      • Big Data

      • HTCondor

      • Foreman and Puppet

      • NIS Protocol

• Prepare a short presentation on the assigned research topic by next week

2/11/2021 - Group Topic Introductions

• • Group Presentations

    • LZ Experiment

      • Dark Matter Summary

        • Extra mass that can't be seen

        • Galaxies exhibit extra mass at the edges of the galaxy (usually)

      • Candidates

        • Doesn't interact electromagnetically

          • Probably not baryonic

        • Has a lot of mass

        • Primordial black holes?

        • Some new particle?

      • Background

        • LZ Experiment

          • Far underground, surrounded by water

            • to shield from unwanted particles

          • Xe core which absorbs energy from particle collisions

            • Scintillation

              • light released can be used to identify particles

            • WIMPs

              • Weakly Interacting Massive Particles

      • What Students Did Last year

        • Analyze simulated data

        • Setting up Linux environment

        • Coding

      • Some Tools

        • Python, C++, and ROOT

    • IceCube

      • What is IceCube

        • Observatory in the south pole

        • Detects neutrinos (high energy)

          • through reactions with water and ice

        • Multinational collaboration (like LHC)

      • IceCube @ UMD

        • Sponsored by Dr. Erik Blaufuss

          • UMD Research Scientist for over 20 years

      • IceCube Detector

        • observes neutrinos from violent astrophysical sources

        • Detector made up of 5160 digital optical modules

        • When neutrinos pass through the ice, electrically charged particles produced

        • Particles emit Cherenkox light

        • Expected detection rate?

          • ???

          • can detect low energy down to 10 TeV

    • Computing Group (My group)

    • Lookup

      • oVirt

      • Open Science Grid

    • Top Quark at LHC

      • Background Information

        • t --> W-b

      • Tools

        • MadGraph5

          • Monte Carlo Simulation

      • The Top Quark

        • Heaviest standard model particle

      • Week One

        • invariance of two-body decay kinematics

        • https://arxiv.org/abs/1209.0772

2/25/2021 - Group Presentations and New Tasks

• • Group presentations

    • IceCube Group

      • How many neutrinos are passing through a given volume of space at a given time? How does that compare to the amount of neutrino detected at IceCube

        • Solar neutrino flux: ~3e12 per m^2 per sec

        • background neutrinos: ~100,000 per year

        • Signal Neutrinos: ~100 per year

      • Does the direction of a neutrino's origin or source affect the results detected at IceCube

        • direction matters for sifting out noise from the signal

        • Leptons emit Cherenkov radiation in ice

          • cosmic-ray sources: downward direction

          • Neutrino sources: any direction

          • Energy helps paint a better picture

        • If the path starts within the array, then it's probably a neutrino

          • if it comes from above, then probably not

          • if it comes from below, then probably something interesting

        • Want to detect neutrinos from celestial events

          • High energy is more likely

          • From below is more likely

            • from above has a bigger chance of being a neutrino from a cosmic ray decay (not what we want)

      • What is Glaciology? Why is it important?

        • Old ice, not like normal ice

        • The study of the optics of the ice is important in lessening the error of the detector

        • Not a major component of systematic errors due to the neutrinos having such high energy

      • Cherenkov Radiation

        • the movement of the particle through the medium is faster than light in the medium

        • kinda like a sonic boom

      • Project updates

        • mainly in python

          • NumPy

          • Matplotlib

        • Access to complete public data release from IceCube

        • Machine learning

        • a lot of Background reading

    • LZ Dark Matter Experiment

      • The theory behind Dark Matter

        • doesn't interact or emit light

        • ~27% of the universe

        • Based on the theory of general relativity

      • A brief history of the universe

        • Temperature goes down as time elapses

        • Cosmic microwave background radiation

          • black body radiation

            • hot things emit light

          • association between temperature and time used to estimate particle composition of the universe

            • so ~27% of the universe

      • Dark matter evidence (rotational curves)

        • Looking at the galaxy, orbital speed is flat because mass is distributed evenly

        • Galaxy clusters, flat rotation curves as well

      • Research Goals

        • LZ experiment extends previous work by LUX and others

        • Want to reproduce results on sensitivity with the LZ detector

          • with simulated data

      • Simulating Data

        • including background noise

          • half the sensitivity calculation is estimating background noise

          • the other half is simulating what data looks like

      • Neutrinos discovery limit

        • Neutrinos are very similar to WIMPs

      • The shape of the sensitivity graph

        • detection is limited at small masses because it's hard to detect things at small energies

        • Why does it get harder as the mass gets larger? (to be addressed)

    • Computing Group Presentation (My Group)

      • .privnet

    • resolve.con file

      • umd.edu vs. privnet?

      • presentation of node 

    • if kicked out of VM

      • how would you get back into it?

    • Top Quark at LHC Group

      • Research goal

        • reproduce paper

      • Some introductions

        • top quark decay

          • extremely shortlived

        • use conservation laws to...?

      • Special relativity

        • Lorentz transformation

          • used to analyze particles in different frames

      • Two Body Decay: An Ideal Case

        • B decays into two particles

          • one of which is massless

        • Process of analysis

          • convert lab frame to rest frame

          • use natural units

          • calculated energy-momentum in the rest frame of B

        • Energy momentum of the lab as related to energy-momentum at rest

          • the Lorentz transformation and angle of decay is different for each experiment

        • Assuming B is unpolarized, cos(\theta) distribution is constant

          • simplifies the energy-momentum conversion to depend on just one variable

      • What if both particles have mass?

        • introduce another variable to account for it

3/4/2021 - Week 3 Group Presentations

• • Top Quark at LHC Group

    • honestly, idk...

    • Want to find upper and lower bounds for the theta based rectangle

      • how does this rectangle relate to the top quark?

      • probability distribution piece of rest energy

    • Rectangles symmetrical under log(x)

    • will be using Madgraph soon

    • Getting a distribution function for the rectangles

      • get the extremum (max) of the distribution function 

    • what if there's more than one extremum?

      • proof not yet complete

  • Computing Group

  • Task: experiment with sites migration

  • LZ Dark Matter Group

    • Detecting particle interactions

      • Goes through a xenon gas chamber

      • Creates two flashes of light per particle interaction

        • S1 and S2

    • Generated Data

      • S1 signal is short and weak compared to the S2 signal that's longer and larger

      • Amplification

        • photomultiplier tube

    • Interaction Types

      • Single Scatter

      • Multiple Scatter

      • Kr85m

      • S2 Only/ Pileup

    • A glimpse into the Code

      • create an enum for interaction types

      • filter out interactions that happen outside the xenon chamber

      • Only consider single scatter events

        • Only interacts once

          • because multiple interactions are highly improbable for dark matter

    • Moving forward

      • getting familiar with the code

      • recreate plots from simulated data

  • IceCube Group

    • Updates

      • Weakly interacting, have mass, but not WIMPs

      • What affects them?

        • Gravity, General relativity, Weak Force

      • What doesn't affect them?

        • EM, strong force

      • Cross-section

        • higher energy, higher the cross-section (likelihood to react)

      • Neutrino Flavor (Type)

        • Electron, Muon, Tau

        • Neutrino decay into its corresponding lepton...?

    • Neutrino Oscillation (so cool...)

      • • Neutrinos don't keep their original flavor

        • Probabilities change as the neutrino travels...?

        • So, must have mass

    • Muon

      • Straight track

      • Hard to tell their energy

      • Less than 1 deg error uncertainty

      • don't love interacting with matter

      • Travel about 1 km before losing all energy

    • Electron

      • Cascade

        • lots of interactions

      • Easier to tell the energy

      • 10 deg of angular uncertainty

    • Tau

      • Double bang

      • Not yet observed

        • not created in cosmic ray interactions

        • Neutrino oscillation after pion decay 1:2:0 (el, mu, tau)

      • Travel about 100 m before decaying

    • Point Source Analysis

      • When Cherenkov light is detected in the tank, position and energy are recorded

        • used to determine point source of particle

      • Binned Methods

        • segment sky into buckets

        • Compare the number of events to the expected number of events from background scrambling

        • If one bucket has a lot of events, telescopes should be focused on that sky area more

      • Tools Available

        • Python

        • Docker

3/11/2021 - Week 4 Group Presentations

• • Housekeeping

  • The goal of the project assignment due midnight today

  • For our computing group:

    • look into past logbooks for details

    • Using GPUs for Machine learning

  • Assignment: Book report for Discovery of the God Particle

  • Curiosity Journal linked in the syllabus: contains past papers for reference

  • IceCube Group

    • Point Source Search

      • Hard to find because the energy level tends to be similar to the background

      • Assume background neutrinos as constant

    • Spatial distributions

      • Point source forms a single point where neutrinos are consistently found to be from

      • The background is mostly constant

    • Probability

      • consistent with background + signal vs background only

    • Software, Installation, and Run

      • Docker

        • containerization Software

      • JupyterLab

        • better than Google Colab...? I think not.

      • Using Docker engine

        • publish an image on a server

        • ssh to it

      • Astronomical Terms

        • MJD

          • modified Julian Date

          • Days from midnight on Nov 17, 1858

        • RA

          • celestial longitude

        • Dec

          • celestial latitude

        • Azimuth and Zenith

          • RA and Dec of the Sun

  • LZ Dark Matter Group

    • Eliminating background signals

      • cosmic ray reduction at 1 mile below the surface

      • passive shielding from the water tank

      • veto system

        • eliminates everything other than a single scatter

      • Other methods

        • clean surface

        • pure xenon

    • Outside sources that cannot be eliminated experimentally

      • neutrino scattering

      • Xenon decay

    • Getting rid of backgrounds

      • remove radioactive contaminants (as much as possible)

      • Underground

      • Multiple filter layers

        • water - muons interact and decay into neutrons

        • neutron interacting liquid

      • Radioactive decay with short lifespans can be eliminated

    • How much background should be expected

      • electron/nuclear events

      • 1131/1.03 total events from normal sources over a 1000-day run

        • after analysis cuts: 5.66/0.52 events

      • if the reality is significantly different from the expected background, then we may have found dark matter

    • Future directions

      • code review done

      • begin remaking signal plots

  • Computing Group

  • Embedding markup language?

    • Use an HTML Box

      • • Insert > HTML Box

        • Include <script src="//cdn.jsdelivr.net/gh/tutts/google-sites-markdown/index.js"></script>

          • then you can just write markdown between the script tags which will be parsed as markdown

  • Top Quark at LHC Group

    • my bad... was figuring out embedded markup stuff...

    • The goal is to recreate the mass histogram of top quark decay

3/25/2021 - Weekly Presentations

• • Computing group presentation

  • FINISH NODE SET UP BY NEXT WEEK!

  • LZ Dark Matter Experiment

    • Nuclear decay chains

      • radon could be long-lived or short-lived, which makes identifying the interactions more difficult

    • The Data

      • most interaction occurs near the edge of the detector before the particles reach the center

      • a weird tail that doesn't match the paper graph

    • Future directions

      • more simulations and plot generations

  • Ice Cube

  • Where are events coming from?

    • Southern Sky, no earth to filter so all taken with a grain of salt

    • Northern Sky, earth in the way, assumed to be neutrino events

    • Horizon, a little bit of earth as a filter, better than the southern sky in terms of filtered events, and better than the northern sky in terms of capturing higher energy neutrinos

    • data artifacts

      • any error in perception or representation of any information, introduced by involved equipment or technique

      • regular distribution of optic detectors results in periodic peaks and troughs of detection despite even distribution of events

        • a more curved sunflower-like distribution should even out this effect

  • Top Quark

    • Attempting event simulation with MadGraph and Matematica

      • currently working on the code

      • file transfer from room to Matematica? 

4/8/2021 - Weekly Presentations

• • Ice Cube

    • Bins and points

      • points don't disappear, but smaller bins mean that the error of any single point is not properly represented

      • too many bins => divide by zero errors

    • Optimizing the number of declination bins

      • tried all values for declination bins in a range of values

      • tried finding the natural number of declination bins in the sky and using that number

    • Unbinned Analysis

      • The likelihood approach

        • assumes probability distribution of data is the most probable distribution

      • k-means clustering

  • Dark Matter

    • Event counts

      • most events don't interact (not detected)

        • especially in the background; possible filtering?

    • The background is weird...

      • events mostly clustered in the 0 bin? not a normal distribution

        • The other category is very large in the background dataset

          • reserved for events that can't be classified into other categories

  • Computing Group

    • Build a NN from scratch

  • Top Quark

    • Varying collider energy

      • peak stays put, but shape changes

    • not enough time... to be finished in the next class

4/15/2021 - Weekly Presentations and Quantum Weirdness

• • Weekly Presentations

    • Top Quark

      • at varied collider energies, peaks should be at the same place

        • but the simulated data doesn't seem to show this

      • debugged log scale graphing

      • The next step is to generate graphs for initial state radiation, in which a gluon is radiated at the beginning of the process

    • Computing group

    • Build NN Demo by next week

    • Dark Matter

      • large bin at zero explained

        • radon decay has zero parent energy since the radon is not moving at relativistic speeds, and radon makes up a significant portion of the background signals

      • WIMP and H3

        • almost no H3 background at 18 GeV, whereas WIMP signal generally increases as GeV increases, so it's a good place to look for WIMPs

      • Next steps

        • reproducing the s1s2 graph from the paper

    • IceCube

      • tested analysis on TXS 0506+056, which is a known high neutrino source

        • however, the point source didn't manage to stick out from the background

        • so, oof

      • Making the binned method better?

        • possible, but not very much, and kinda arbitrary or computationally difficult

      • Maybe just go with the unbinned method

  • Quantum weirdness

    • Bell's experiments

      • only rules out local hidden variables

      • Pilot Wave Theory? (Maybe something to look into)

    • Entanglement experiments

      • observing lights through gravitational lensing?

      • create the entangled particle yourself

      • Measurements questions:

        • timing?

EOF Stub