---
title: Continuous-Time Decoherence Model
url: https://www.emergentmind.com/topics/continuous-time-decoherence-model
type: topic
---

# Continuous-Time Decoherence Model

A continuous-time decoherence model describes the irreversible suppression of quantum coherences under the action of environmental noise or measurement, formulated in the time domain via realistic system-plus-environment dynamics, master equations, or stochastic unravellings. Such models are foundational across quantum open systems, quantum information, condensed matter, and quantum technologies, enabling the quantitative analysis of decoherence rates, quantum-to-classical transition, entropy generation, and the limits of quantum sensing. Approaches span from “microscopic” system–bath Hamiltonians (path-integral, master-equation, and correlation-function methods), to phenomenological Lindblad-type descriptions, to stochastic process formalisms and measurement-based “continuous monitoring” schemes.

## 1. Real-Time Path Integral and the Caldeira–Leggett Model

The Caldeira–Leggett (CL) model provides a canonical framework for continuous-time decoherence in open quantum systems, describing a system coordinate $x$ coupled linearly to an environment of harmonic oscillators:
\[
H = \frac{p^2}{2M}+V(x) + \sum_{k=1}^{N_\mathcal{E}} \left[\frac{p_k^2}{2m}+\frac12\,m\omega_k^2\,q_k^2\right] - x\sum_{k=1}^{N_\mathcal{E}} c_k\,q_k\,.
\]
For quadratic $V(x)$, the path integral over the environment is Gaussian and tractable by deformation to a Lefschetz thimble in complexified field space. The reduced density matrix $\rho_\mathcal{S}(x_f, x_f'; t)$, traced over environmental degrees, is given as a double path integral weighed by the Feynman–Vernon influence functional
\[
\mathcal{F}[x, \tilde{x}] = \exp\left\{ - \int_0^t d\tau \int_0^\tau ds\, [x(\tau)-\tilde{x}(\tau)][\alpha(\tau-s) x(s) - \alpha^*(\tau-s)\tilde{x}(s)] \right\}\,,
\]
with the bath kernel $\alpha(\tau)=\sum_k \frac{c_k^2}{2m\omega_k}\left(\coth\frac{\beta\omega_k}{2} \cos \omega_k\tau - i\sin \omega_k \tau \right)$.

Under Gaussian initial conditions, the reduced dynamics are characterized by time-dependent quadratic forms in $(x_f\pm x_f')$; the decay rate of the off-diagonal components, $\Gamma_{\rm off}(t)$, determines the decoherence time. In the Markovian, high-$T$, weak-coupling regime,
\[
\Gamma_{\rm off}(t) \simeq \frac{8\gamma}{\beta} t, \qquad \tau_D(\Delta x) = \frac{\beta}{8\gamma \Delta x^2} = \frac{\hbar^2}{2M\gamma k_B T\,\Delta x^2}\,.
\]
This matches predictions from master-equation treatments [2503.20699]. Numerical simulations confirm convergence to Ohmic/Markovian limits at large bath size and deviations at stronger coupling or lower temperature.

## 2. Markovian Master Equation and Spectral Density Dependence

For a generic qubit or pointer-type system linearly coupled to a bosonic bath, the off-diagonal density matrix evolves as
\[
\varrho_{10}(t) = \varrho_{10}(0) e^{-\Gamma(t)},
\]
with the decoherence functional determined by the bath spectral density $J(\omega)$:
\[
\Gamma(t) = \int_0^\infty d\omega\, J(\omega)\coth{\frac{\beta\omega}{2}}\frac{1-\cos(\omega t)}{\omega^2}.
\]
The low-frequency scaling of $J(\omega)\sim c\,\omega^\gamma$ (sub-Ohmic, Ohmic, super-Ohmic) governs the long-time decay:
- Super-Ohmic ($\gamma >2$): partial, saturated decoherence
- Ohmic ($\gamma=1$): exponential with log correction
- Sub-Ohmic ($\gamma<1$): superexponential (faster than exponential)

A GKSL semigroup master equation is only strictly attainable for pure exponential (Ohmic, $\alpha=0$) case; non-Markovianity is generic elsewhere [2311.16010].

## 3. Memory Effects and Non-Markovian Functionals

Finite-memory environments generate time-nonlocal decoherence with functionals of the form
\[
\mathcal{F}(t) = \frac{a^2}{\hbar^2}\int_0^t ds \int_0^t ds' C(s-s'),
\]
where $C(\tau)$ is the bath autocorrelation. Short-time, model-independent quadratic growth (for $t\ll\tau_c$):
\[
\mathcal{F}(t)\approx \frac{a^2}{\hbar^2}C(0) t^2,
\]
yields an operational decoherence time $T_d \sim 1/\sqrt{A} \propto \sqrt{\tau_c}$ (with $A=a^2 C(0)/\hbar^2$). For stationary Ornstein–Uhlenbeck noise, the decoherence ODE closes exactly:
\[
\ddot{C}(t) + \frac{1}{\tau_c}\dot{C}(t) + \frac{2a^2 D}{\hbar^2 \tau_c}C(t) = 0\,.
\]
Exponential suppression emerges only for vanishing $\tau_c$ (memoryless limit). Quantum signatures (purity, entropy) vanish on a distinct, typically longer, timescale than the decay of a single coherence element [2601.17394].

## 4. Lindblad and Measurement-Based Approaches

Phenomenological continuous-time decoherence is often treated within Lindblad master equations:
- **Pure dephasing (site-basis)**: $d\rho/dt = -i[H,\rho] + \gamma \sum_k (P_k \rho P_k - \frac{1}{2}\{P_k,\rho\})$
- **Intrinsic decoherence (energy-basis)**: $d\rho/dt = -i[H,\rho] - \frac{\gamma}{2} [H, [H, \rho]]$
- **Quantum Stochastic Walk Interpolation**: $d\rho/dt = -(1-p)i[H,\rho] + p\mathcal{D}_{\rm classical}(\rho)$

Quantum walks on graphs subject to different decoherence channels demonstrate qualitative differences: site-basis noise leads to full classicalization and vanishing quantum-classical distance, while energy-basis decoherence leaves residual nonclassicality determined by Laplacian degeneracies [2204.01836, 2507.17880].

Measurement-based, or “universal” models (e.g., continuous tomographic measurement), give rise to isotropic Lindblad decoherence driving the state to the maximally mixed state on a time scale $\tau_d = \frac{(\sigma+1)\ln(N+1)}{4\gamma N}$, with faster classicalization of larger systems [2511.07369]. The associated Fokker–Planck equation for the quasiprobability distribution exhibits purely local exponential decay toward uniformity.

## 5. Quantum Stochastic Unravellings and Noise Models

Stochastic unravellings—such as the “quantum drift” (QD) model—simulate continuous-time decoherence by stochastic phase evolution in a pointer basis:
\[
d|\psi\rangle = -\frac{i}{\hbar}H_0|\psi\rangle dt + \sum_n |n\rangle\langle n|\psi\rangle \left[-\frac{i}{\hbar}dW_n(t)\right],
\]
where $dW_n$ are independent Wiener increments. The corresponding ensemble-averaged master equation reproduces the Lindblad pure-dephasing form. The QD scheme efficiently simulates transport or dynamics in large many-body systems, with advantageous convergence properties compared to quantum-jump unraveilings [1501.06610].

Noise models based on random telegraph processes (RTP) accurately describe both $1/f$ and Lorentzian noise spectra relevant in superconducting, spin, and donor qubits. The effective decoherence function is generically non-exponential, and exact analytical fitting to experiments in these platforms shows that both slow and fast fluctuators must be included [1201.2316].

## 6. Decoherence and Dissipation: Structure and Timescales

Continuous-time decoherence generically leads to a master equation of the form
\[
\partial_t \rho = \frac{1}{i\hbar}[H_S, \rho] - D[x,[x,\rho]] + \text{(friction, momentum diffusion terms)}.
\]
For the canonical quantum Brownian motion or measurement-induced decoherence, the pointer basis corresponds to minimum-uncertainty Gaussians in position (and, for the Stern–Gerlach system, spin-correlated wavepackets). The decoherence time for superpositions of separation $\Delta x$ is
\[
\tau_D(\Delta x) = \frac{\hbar^2}{2 m \gamma k_B T (\Delta x)^2}.
\]
Decoherence is typically much faster than relaxation in momentum (dissipation), and is maximally efficient in fully classical (high-$T$) environments; quantum (low-$T$) baths suppress decoherence relative to diffusion, as the environmental quantum coherence inhibits effective “which-path” marking [1605.00834, 1109.2412].

## 7. Generalizations: Nonlinear, Measurement-Controlled, and Relativistic Models

Beyond standard models, continuous-time decoherence under nonlinear stochastic control, as in atomic magnetometry and spin squeezing, is analyzed using conditional stochastic master equations (SMEs) including measurement backaction and dissipative channels. Gaussian or Kalman-filtered approximations enable scalable simulation for large $N$:
\[
\partial_t \rho_c = -i[H(t), \rho_c] + \gamma_{\rm rel}D[J_-]\rho_c + \gamma_\phi D[J_z]\rho_c + \kappa D[J_z]\rho_c + \sqrt{\kappa\eta}\mathcal{H}[J_z]\rho_c\,\xi(t).
\]
Quantum-limited precision scales linearly with both sensing time and particle number, precluding Heisenberg-type scaling once decoherence is included [2512.05265].

General relativistic decoherence models, such as time-dilation decoherence, yield master equations involving dynamical couplings (e.g., proper time differences)
\[
\mathcal{D}[\rho_x]=-\gamma[T,[T,\rho_x]],
\]
with decoherence rates proportional to the variance of “clock” energy and the spread in the relevant dynamical generator, incorporating both special and general relativistic effects [1503.05488].

---

These continuous-time models, spanning exact and effective treatments, underpin the current theoretical understanding of decoherence and its role in quantum measurement, quantum information, and the emergence of classicality [2503.20699, 2601.17394, 2204.01836, 2311.16010, 1201.2316, 1605.00834, 1109.2412, 2512.05265, 1503.05488, 1405.7666, 2511.07369, 2507.17880, 2311.08508, 1501.06610, 1001.1034, 2006.01121].

Source: https://www.emergentmind.com/topics/continuous-time-decoherence-model