Periodic Reporting for period 4 - PrecisionNuclei (Strong interactions for precision nuclear physics)
Reporting period: 2022-08-01 to 2023-01-31
Nuclear physics is a cornerstone in our scientific endeavour to understand the universe. Indeed, atomic nuclei bring us closer to study both the stellar explosions in the macrocosmos, where the elements are formed, and the fundamental symmetries of the microcosmos. Having access to a a precise description of the interactions between protons and neutrons would provide a key to new knowledge across 20 orders of magnitude; from neutrinos to neutron stars. Despite a century of the finest efforts, a systematic description of strongly interacting matter at low energies is still lacking. Successful theoretical approaches, such as mean-field and shell models, typically rely on uncontrolled approximations that severely limit their predictive power in regions where the model has not been adjusted.
In this project I will develop a novel methodology to use experimental information from heavy atomic nuclei in the construction of nuclear interactions from chiral effective field theory. I expect this approach to enable me and my team to make precise ab initio predictions of various nuclear observables in a wide mass-range from hydrogen to lead as well as infinite nuclear matter. I will apply Bayesian regression and methods from machine learning to quantify the statistical and systematic uncertainties of the theoretical predictions. The novelty and challenge in this project lies in synthesising (i) the design of nuclear interactions, (ii) ab initio calculations of nuclei, and (iii) statistical inference in the confrontation between theory and experimental data. This alignment of methods, harboured within the same project, will create a clear scientific advantage and allow me to tackle the following big research question: How well can atomic nuclei and nuclear matter be described in chiral effective field theories of quantum chromo dynamics?
In this project I will develop a novel methodology to use experimental information from heavy atomic nuclei in the construction of nuclear interactions from chiral effective field theory. I expect this approach to enable me and my team to make precise ab initio predictions of various nuclear observables in a wide mass-range from hydrogen to lead as well as infinite nuclear matter. I will apply Bayesian regression and methods from machine learning to quantify the statistical and systematic uncertainties of the theoretical predictions. The novelty and challenge in this project lies in synthesising (i) the design of nuclear interactions, (ii) ab initio calculations of nuclei, and (iii) statistical inference in the confrontation between theory and experimental data. This alignment of methods, harboured within the same project, will create a clear scientific advantage and allow me to tackle the following big research question: How well can atomic nuclei and nuclear matter be described in chiral effective field theories of quantum chromo dynamics?
This project focused on the development of novel methodologies for using experimental information from light-mass and heavy-mass atomic nuclei in the construction of nuclear interactions from effective field theory. Significant progress was made on all three fundamental questions that drive this project forward: (i) the design of nuclear interactions, (ii) ab initio predictions of nuclear observables, and (iii) statistical analysis of theoretical predictions.
The most important results include the development of eigenvector continuation for emulation and a novel technology called Subspace-Projected Coupled Cluster (SPCC). This is a game changer for the theoretical analyses of nuclei. In particular for performing Bayesian inference studies of nuclear properties to reveal patterns of the underlying nuclear interaction and to correlate experimental information across the nuclear chart. This project made key contributions to a breakthrough calculation of the neutron skin thickness in 208-Lead and aa novel statistical statistical analysis linking this result to details of chiral nuclear forces and the properties of infinite nuclear matter. These results were disseminated at international conferences and in high-impact scientific journals (Physical Review Letters and Nature Physics).
The most important results include the development of eigenvector continuation for emulation and a novel technology called Subspace-Projected Coupled Cluster (SPCC). This is a game changer for the theoretical analyses of nuclei. In particular for performing Bayesian inference studies of nuclear properties to reveal patterns of the underlying nuclear interaction and to correlate experimental information across the nuclear chart. This project made key contributions to a breakthrough calculation of the neutron skin thickness in 208-Lead and aa novel statistical statistical analysis linking this result to details of chiral nuclear forces and the properties of infinite nuclear matter. These results were disseminated at international conferences and in high-impact scientific journals (Physical Review Letters and Nature Physics).
With the developments made in this project we now have access to novel emulation methods for evaluating Bayesian posterior predictive distributions of key nuclear observables in light- and heavy-mass nuclei as well as posterior probability distributions, or non-implausible parameter regions, for the low-energy constants in effective field theory. The suite of methods developed in this project lay the foundation for carrying out experimental design studies and statistical inference analyses to generate new knowledge about the interaction between nucleons inside atomic nuclei and infinite nuclear matter. The latter has a direct bearing on improving our understanding of neutron stars.