Thermal and Nonequilibrium Field Theory
Thermal and nonequilibrium field theory asks how macroscopic matter and real-time response emerge from quantum fields in states other than the vacuum. It spans equilibrium thermodynamics, spectral functions, kinetic theory, hydrodynamics, open systems, quenches, heavy-ion and cosmological applications, and thermalization. “Finite-temperature QFT” covers only the equilibrium corner; nonequilibrium QFT additionally requires an initial state and a real-time contour or equivalent evolution law.
Evidence cutoff. 11 August 2026.
Required background. Thermal density operators and the KMS condition states equilibrium analyticity, closed-time-path functionals organize causal real-time observables, and conservation laws and hydrodynamic fields define the infrared variables.
Helpful background. Spectral reconstruction exposes the Euclidean inverse problem, linearized kinetics states the quasiparticle closure, QCD crossover thermodynamics supplies a principal evidence domain, and cross-method observable translation prevents false comparisons.
Thermal and nonequilibrium descriptions
Section titled “Thermal and nonequilibrium descriptions”| Description | Required separation or input | Characteristic observables | Breakdown signal |
|---|---|---|---|
| equilibrium Euclidean QFT | stationary KMS state | pressure, susceptibilities, static screening | real-time detail requires ill-posed continuation |
| Schwinger–Keldysh QFT | initial density operator and contour | causal response, noise, unequal-time correlators | memory and secular growth defeat a chosen closure |
| kinetic theory | quasiparticles and scale-separated collisions | distribution functions, rates, transport | broad excitations or strong coherence invalidate on-shell transport |
| hydrodynamics | long wavelengths and conserved slow modes | flow, diffusion, sound, constitutive response | gradients/frequencies leave the effective regime or modes become unstable |
| open-system EFT | declared system/environment split | channels, influence functionals, decoherence | non-Markovian memory or uncontrolled factorized initial state |
These descriptions overlap but are not a ladder of universal accuracy. Hydrodynamics can work before local isotropy, while kinetic theory can retain nonhydrodynamic modes that a gradient expansion omits. The closed-time contour organizes causal response and correlations for a specified initial state Keldysh 1965, foundational method. A Schwinger–Keldysh effective action can encode fluctuation–dissipation through dynamical KMS symmetry, but its ultraviolet completion and higher-gradient sector remain independent data Jain and Kovtun 2024, method.
Nonequilibrium evidence and uncertainty cultures
Section titled “Nonequilibrium evidence and uncertainty cultures”Equilibrium lattice calculations demand continuum, thermodynamic-limit, mass-tuning, and scale-setting control. Spectral reconstructions must report the kernel, temporal resolution, covariance, positivity assumptions, prior or regularization, mock-data resolution, and which features survive alternative reconstructions. Heavy-ion inference adds initial-condition, pre-equilibrium, hydrodynamic, hadronic-transport, and detector layers with shared parameters; a narrow posterior can still be conditional on a misspecified model family.
Real-time analytic approximations track expansion parameters: coupling, occupation, gradients, , or proximity to equilibrium. Self-consistent -derivable/nPI approximations can preserve chosen conservation laws but do not automatically preserve gauge identities or crossing Baym 1962, foundational method. A systematic account of the resulting 2PI/PI nonequilibrium hierarchy is given in Berges 2004, orientation. Hydrodynamic fluctuation loops generate nonanalytic long-time tails and limit how small higher-order transport corrections can be made Kovtun, Moore, and Romatschke 2011, obstruction.
Benchmarks should match actual observables: equilibrium equations of state; spectral sum rules; weak-coupling transport; exactly solvable quenches; kinetic-to-hydrodynamic matching; and the same initial-value problem across classical-statistical, tensor-network, and quantum-simulation methods. Thermal resummations should recover the free and known weak-coupling limits and display scale and prescription dependence. Spectral reconstructions should pass injection tests in which a known spectrum is propagated through the same kernel and covariance; collision solvers should conserve the required charges and reproduce detailed balance and known relaxation rates. Fluctuating-hydrodynamic calculations should recover equilibrium fluctuation–dissipation relations and the expected long-time tails before being used away from equilibrium.
Durable limitations
Section titled “Durable limitations”There is no general nonperturbative real-time Monte Carlo method for finite-density QCD because the sign problem destroys ordinary importance sampling. Euclidean data do not uniquely determine a finely resolved spectrum at realistic noise without additional information. Hydrodynamic gradient expansions are often asymptotic; an attractor can reorganize predictivity in a specified observable without proving rapid thermalization or universality across theories. First-order relativistic constitutive equations in conventional frames can be acausal or unstable; causal formulations add structure and a domain of validity rather than making hydrodynamics microscopic.
Thermalization itself has several meanings—local equilibration of selected observables, eigenstate thermalization, kinetic equilibration, or approach to a KMS state. Integrable systems and many-body localization supply important counterexamples to naive universal equilibration, although their stability depends on dimension and perturbations.
Entering the field
Section titled “Entering the field”Specify the state, contour, observable, hierarchy of scales, and time window before choosing a method. Reproduce an equilibrium or linear-response limit, then vary the closure or reconstruction prior on held-out synthetic data. The particle and nuclear pathway or quantum-matter pathway supplies domain preparation; uncertainty and disagreement supplies the evidence discipline.
This guide omits general statistical mechanics when no QFT or continuum-field question is at stake, and it does not treat heavy-ion phenomenology as a direct measurement of microscopic transport without the full inference chain.
Evidence scope and related assessments
Section titled “Evidence scope and related assessments”The finite search used arXiv, INSPIRE, journal/DOI records, lattice-collaboration sources, and targeted searches for inverse-problem, hydrodynamic, and thermalization counterexamples. Sources public through 11 August 2026 were eligible. Reassess when a continuum real-time benchmark becomes available, a reconstruction claim fails mock-data validation, or a causal/stochastic formulation changes an extracted transport result.
Continue to the QCD critical point, hydrodynamic attractors, the real-time/kinetic/hydrodynamic method map, or the finite-temperature QCD and attractor evidence briefs.
References
Section titled “References”- G. Baym, “Self-Consistent Approximations in Many-Body Systems,” Physical Review 127 (1962) 1391–1401. DOI.
- J. Berges, “Introduction to Nonequilibrium Quantum Field Theory,” AIP Conference Proceedings 739 (2004) 3–62. arXiv.
- A. Jain and P. Kovtun, “Schwinger–Keldysh Effective Field Theory for Stable and Causal Relativistic Hydrodynamics,” JHEP 01 (2024) 162. DOI.
- L. V. Keldysh, “Diagram Technique for Nonequilibrium Processes,” Soviet Physics JETP 20 (1965) 1018–1026. Journal archive.
- P. Kovtun, G. D. Moore, and P. Romatschke, “The Stickiness of Sound: an Absolute Lower Limit on Viscosity and the Breakdown of Second-Order Relativistic Hydrodynamics,” JHEP 07 (2011) 123. DOI.