UC Berkeley Previously Published Works

Deep Generative Models for Fast Photon Shower Simulation in ATLAS

(2024)

Abstract: The need for large-scale production of highly accurate simulated event samples for the extensive physics programme of the ATLAS experiment at the Large Hadron Collider motivates the development of new simulation techniques. Building on the recent success of deep learning algorithms, variational autoencoders and generative adversarial networks are investigated for modelling the response of the central region of the ATLAS electromagnetic calorimeter to photons of various energies. The properties of synthesised showers are compared with showers from a full detector simulation using geant4. Both variational autoencoders and generative adversarial networks are capable of quickly simulating electromagnetic showers with correct total energies and stochasticity, though the modelling of some shower shape distributions requires more refinement. This feasibility study demonstrates the potential of using such algorithms for ATLAS fast calorimeter simulation in the future and shows a possible way to complement current simulation techniques.

Software Performance of the ATLAS Track Reconstruction for LHC Run 3

(2024)

Charged particle reconstruction in the presence of many simultaneous proton–proton (pp) collisions in the LHC is a challenging task for the ATLAS experiment’s reconstruction software due to the combinatorial complexity. This paper describes the major changes made to adapt the software to reconstruct high-activity collisions with an average of 50 or more simultaneous pp interactions per bunch crossing (pile-up) promptly using the available computing resources. The performance of the key components of the track reconstruction chain and its dependence on pile-up are evaluated, and the improvement achieved compared to the previous software version is quantified. For events with an average of 60pp collisions per bunch crossing, the updated track reconstruction is twice as fast as the previous version, without significant reduction in reconstruction efficiency and while reducing the rate of combinatorial fake tracks by more than a factor two.

The present and future of QCD

(2024)

This White Paper presents an overview of the current status and future perspective of QCD research, based on the community inputs and scientific conclusions from the 2022 Hot and Cold QCD Town Meeting. We present the progress made in the last decade toward a deep understanding of both the fundamental structure of the sub-atomic matter of nucleon and nucleus in cold QCD, and the hot QCD matter in heavy ion collisions. We identify key questions of QCD research and plausible paths to obtaining answers to those questions in the near future, hence defining priorities of our research over the coming decades.

A RAB7A phosphoswitch coordinates Rubicon Homology protein regulation of Parkin-dependent mitophagy.

(2024)

Activation of PINK1 and Parkin in response to mitochondrial damage initiates a response that includes phosphorylation of RAB7A at Ser72. Rubicon is a RAB7A binding negative regulator of autophagy. The structure of the Rubicon:RAB7A complex suggests that phosphorylation of RAB7A at Ser72 would block Rubicon binding. Indeed, in vitro phosphorylation of RAB7A by TBK1 abrogates Rubicon:RAB7A binding. Pacer, a positive regulator of autophagy, has an RH domain with a basic triad predicted to bind an introduced phosphate. Consistent with this, Pacer-RH binds to phosho-RAB7A but not to unphosphorylated RAB7A. In cells, mitochondrial depolarization reduces Rubicon:RAB7A colocalization whilst recruiting Pacer to phospho-RAB7A-positive puncta. Pacer knockout reduces Parkin mitophagy with little effect on bulk autophagy or Parkin-independent mitophagy. Rescue of Parkin-dependent mitophagy requires the intact pRAB7A phosphate-binding basic triad of Pacer. Together these structural and functional data support a model in which the TBK1-dependent phosphorylation of RAB7A serves as a switch, promoting mitophagy by relieving Rubicon inhibition and favoring Pacer activation.

Anthropogenic heat from buildings in Los Angeles County: A simulation framework and assessment

(2024)

Anthropogenic heat (AH), i.e., waste heat from buildings to the ambient environment, increases urban air temperature and contributes to the urban heat island effect, which leads to more air-conditioning energy use and higher associated waste heat during summer, forming a positive feedback loop. This study used a bottom-up simulation approach to develop a dataset of the annual hourly AH profiles for 1.7 million buildings in Los Angeles (LA) County for the year 2018 aggregated at three spatial resolutions: 450 m, 12 km, and the census tract. Building AH exhibits strong seasonal and diurnal patterns, as well as large spatial variations across the urban areas. Building AH peaks in May and reaches a maximum of 878 W/m2 within one of several AH hotspots in the region. Among the three major AH components (surface convection, heat rejection from HVAC systems, and zonal air exchange), the surface convection component is the largest, accounting for 78% of the total building AH across LA County. Higher AH is attributed to large building density, a high percentage of industrial buildings, and older building stock. While AH peaks during the day, the resulting ambient temperature increases are much larger during the night. During the July 2018 heatwave in LA County, building AH (excluding the surface component) leads to a daily maximum ambient temperature increase of up to 0.6 °C and a daily minimum ambient temperature increase of up to 2.9 °C. It is recommended that reducing summer building AH should be considered by policy makers in developing mitigation measures for cities to transition to clean energy while improving heat resilience.

Measurement of the Z boson invisible width at s = 13 TeV with the ATLAS detector

(2024)

A measurement of the invisible width of the Z boson using events with jets and missing transverse momentum is presented using 37 fb−1 of 13 TeV proton–proton data collected by the ATLAS detector in 2015 and 2016. The ratio of Z→inv to Z→ℓℓ events, where inv refers to non-detected particles and ℓ is either an electron or a muon, is measured and corrected for detector effects. Events with at least one energetic central jet with pT≥110 GeV are selected for both the Z→inv and Z→ℓℓ final states to obtain a similar phase space in the ratio. The invisible width is measured to be 506±2(stat.)±12(syst.) MeV and is the single most precise recoil-based measurement. The result is in agreement with the most precise determination from LEP and the Standard Model prediction based on three neutrino generations.

Efficiency of ETA Prediction

(2024)

Computational-Rock Mechanics in Pedagogy and Practice

(2024)

Point cloud modeling of rock slopes using LIDAR and Structure-from-Motion digital stereophotogrammetry provides,

at a minimum, thousands of facets and facet normals that can be used to identify the densities of orientations of rock mass

discontinuities, the geometries of potentially removable blocks, and the character of the excavation face. As part of the Engineering

Geology graduate curriculum at the Civil and Environmental Engineering program at the University of California, Berkeley we teach

graduate students an integrated methodology for [a] gathering point cloud information be laser or camera; [b] computing facets and

facet normals form point clouds for stereonet presentation and geometric analysis of block dimension; [c] extract rock mass

discontinuities from stereonet data to analyze key blocks, assess discontinuous deformation analysis (DDA) behavior, and model

rock slope stability. These new methods require a suite of different software tools discussed in the paper to move through the workflow

process. Computational rock mechanics provides data sets that are orders of magnitude richer in detail and result in better

understanding of rock slope and tunnel key block behavior. Full application of computational rock mechanics methods should reduce

the cost of bolting by identifying critical support orientations and design loads.

Estimating the effect of realistic improvements of metformin adherence on COVID-19 mortality using targeted machine learning.

(2024)

BACKGROUND: Type 2 diabetes elevates the risk of severe outcomes in COVID-19 patients, with multiple studies reporting higher case fatality rates. Metformin is a widely used medication for glycemic management. We hypothesize that improved adherence to metformin may lower COVID-19 post-infection mortality risk in this group. Utilizing data from the Mexican Social Security Institute (IMSS), we investigate the relationship between metformin adherence and mortality following COVID-19 infection in patients with chronic metformin prescriptions. METHODS: This is a retrospective cohort study consisting of 61,180 IMSS beneficiaries who received a positive polymerase chain reaction (PCR) or rapid test for SARS-CoV-2 and had at least two consecutive months of metformin prescriptions prior to the positive test. The hypothetical intervention is improved adherence to metformin, measured by proportion of days covered (PDC), with the comparison being the observed metformin adherence values. The primary outcome is all-cause mortality following COVID-19 infection. We defined the causal parameter using shift intervention, an example of modified treatment policies. We used the targeted learning framework for estimation of the target estimand. FINDINGS: Among COVID-19 positive patients with chronic metformin prescriptions, we found that a 5% and 10% absolute increase in metformin adherence is associated with a respective 0.26% (95% CI: -0.28%, 0.79%) and 1.26% (95% CI: 0.72%, 1.80%) absolute decrease in mortality risk. INTERPRETATION: Subject to the limitations of a real-world data study, our results indicate a causal association between improved metformin adherence and reduced COVID-19 post-infection mortality risk.