---
title: Dynamic Quantum Fisher Information
url: https://www.emergentmind.com/topics/dynamic-quantum-fisher-information
type: topic
---

# Dynamic Quantum Fisher Information

Dynamic quantum Fisher information (QFI) quantifies the instantaneous sensitivity of a quantum system—pure or mixed, closed or open—to changes in physically relevant parameters as the system evolves in time. It provides a rigorous operational measure of distinguishability in parameter space, bounds quantum-enhanced estimation error, and directly diagnoses dynamical entanglement, coherence, non-Markovianity, and phase transitions in unitary and dissipative many-body evolution. The full dynamical QFI framework encompasses time-dependent state QFI, channel QFI, their generalization to multi-parameter and open system settings, and their functional role in quantum metrology bounds, resource theory, quantum thermodynamics, and critical phenomena.

## 1. Fundamental Definitions and Mathematical Structure

Dynamic quantum Fisher information is constructed from a family of parameter-dependent quantum states $\rho(\theta, t)$ or channels $\mathcal{E}_{\theta, t}$. For a mixed state evolving under any protocol (unitary, CPTP, Lindbladian), the symmetric logarithmic derivative (SLD) $L_\theta(t)$ is defined as the Hermitian solution to:
\[
\partial_\theta \rho(\theta, t) = \frac{1}{2}(L_\theta(t) \rho(\theta, t) + \rho(\theta, t) L_\theta(t))
\]
The (single-parameter) dynamic QFI is given by:
\[
F_Q[\rho(\theta, t)] = \mathrm{Tr}[\rho(\theta, t)\,L_\theta^2(t)]
\]
For pure states, it reduces to $F_Q = 4\,\mathrm{Var}_{\Psi(t)}[O]$ for a generator $O$. For a vector of parameters $\vec{x}$, the quantum Fisher information matrix (QFIM) generalizes as:
\[
F_{\mu\nu}(t) = \frac{1}{2}\,\mathrm{Tr}\left[\rho(t)\{L_\mu(t), L_\nu(t)\}\right]
\]
where $L_\mu(t)$ solves $\partial_{x_\mu}\rho(t) = \frac{1}{2}(L_\mu(t)\rho(t) + \rho(t)L_\mu(t))$ [2508.16413].

For channels parameterized by $\theta$, dynamic QFI is:
\[
F_Q(\mathcal{E}_\theta) = \sup_{\rho}\,F_Q[\mathcal{E}_\theta(\rho)]
\]
i.e., the maximal achievable QFI over all probe states [2104.01763].

## 2. Dynamical Evolution: Closed and Open Systems

In closed systems, the evolution is unitary: $\rho(t) = U(t)\rho_0 U^\dagger(t)$; QFI dynamics track spread in the generator basis. For a time-dependent unitary family $U(\theta, t)=e^{-i t H(\theta)}$ (e.g. SU(2) processes), the optimal QFI splits into quadratic-in-time and oscillatory parts:
\[
F_{\mathrm{max}}(t) = 4\left\| (d r/d\theta)\,\mathbf{e}_r \right\|^2 t^2 + \frac{16}{r^2}\left\| r\,d\mathbf{e}_r/d\theta \right\|^2 \sin^2\left(\frac{r t}{2}\right)
\]
This decomposition has clear geometric meaning: the time-square term measures generator norm variation; the oscillatory term, generator axis rotation [1503.05697].

For open systems governed by a Lindblad or Liouvillian master equation:
\[
\frac{d}{dt}\rho(\theta, t) = \mathcal L_\theta[\rho(\theta, t)], \qquad \rho(\theta, t) = e^{t\,\mathcal L_\theta}[\rho_0]
\]
the dynamic QFI is upper-bounded by closed-form expressions involving the non-Hermitian SLD, e.g. for Hamiltonian estimation,
\[
F_{\mathrm{ext}}[\rho(\theta_0, t)] = 4 t^2 \,\mathrm{Var}_{\rho(t)}(H)
\]
For dissipative parameters, it involves traces over Lindblad operators and their action on $\rho(t)$, yielding rich time dependence (quadratic, exponential, oscillatory) set by the Liouvillian spectrum [1403.8033, 2308.10183].

## 3. Scaling, System Size, and Critical Behavior

Dynamic QFI provides critical scaling information:

- **Heisenberg limit:** For multipartite systems, $F_Q\sim N^2$ witnesses genuine multipartite entanglement and sets the ultimate quantum metrological bound [1812.01013, 2507.05579].
- **Standard quantum limit/shot noise:** Linear scaling $F_Q\sim N$ arises in separable or decohered regimes.
- **Crossover:** In open systems, dynamic QFI interpolates between quantum ($N^2$) and classical ($N$) regimes, with the threshold set by decoherence strength and interrogation time—see $N^*=c_1/c_2$ in [1403.8033].

At phase transitions, time-averaged QFI displays universal nonanalytic behavior, serving as an order parameter for dynamical quantum phase transitions (DQPTs) [2507.05579]. In dissipative many-body systems, dynamical QFI also reveals transitions from Markovian (monotonic loss) to non-Markovian (information backflow, oscillatory or even revival) behavior [2508.16413, 2102.09586, 2304.14984].

## 4. Entanglement, Metrology, and Resource Quantification

Dynamic QFI detects and quantifies multipartite entanglement—$F_Q > N$ (for N qubits) implies at least $k$-partite entanglement, with $k \sim F_Q/N$ [1812.01013]. QFI underlies the quantum Cramér–Rao bound for parameter estimation:
\[
\delta\theta \geq 1/\sqrt{M F_Q}
\]
with $M$ repetitions. In dynamic protocols (e.g., Dicke, Lipkin–Meshkov–Glick, spin-boson models), ergodic or chaotic evolution can exponentially accelerate the growth of useful entanglement, lowering the time scale for reaching the Heisenberg limit from $\sim\sqrt{N}$ to $\sim\log N$ [1812.01013].

In resource theories, dynamic QFI (channel or state) quantitatively distinguishes resourceful channels and states from free ones, establishing Fisher information as a universal resource witness [2104.01763].

## 5. Markovianity, Non-Markovianity, and Information Flow

The instantaneous time derivative of dynamic QFI, or its geometric generalization (intrinsic density flow, IDF), is a sharp witness of non-Markovianity:
- For CP-divisible (Markovian) evolution, $\partial_t F_Q(t) \leq 0$.
- Temporary $\partial_t F_Q(t) > 0$ signals non-Markovian information backflow from environment to system.

IDF generalizes to multi-parameter manifolds; its sign and magnitude express quantifiable loss or regaining of local distinguishability in time [2102.09586, 2304.14984]. These concepts are analytically and numerically verified in models such as spin-boson, Yang-Baxterized spin chains, and qubit dephasing [2508.16413, 2008.11443].

## 6. Decoherence Channels and Analytic QFI Evolution

Explicit dynamic QFI formulas have been derived for memoryless decoherence channels (phase-damping, depolarizing, amplitude-damping) for both single-qubit and multi-qubit (X-state, GHZ, Werner) inputs [1212.0917, 2412.01473, 2209.15593]. Key findings:
- Phase-damping and phase-flip noise preserve residual QFI even after entanglement vanishes ("entanglement sudden death"), maintaining metrological advantage.
- Depolarizing channels eradicate both entanglement and QFI, restoring classical scaling.
- Time-dependent block-decomposition and affine Bloch representations facilitate analytic expressions for general qudit systems [1212.0917, 2412.01473].

## 7. Dynamical QFI in Many-Body and Nonequilibrium Scenarios

Recent work demonstrates that local nonequilibrium "kicking" (transient local perturbations) amplifies subsystem QFI from $O(L)$ (linear in block length $L$) to $O(L^2)$ (quadratic scaling), transiently producing macroscopic entangled states within compact blocks. These enhancements are tunable via protocol structure (single/local/global kicks, periodic driving), subject to a characteristic timescale for entanglement generation and subsequent decay [2503.21905]. The connection between QFI and Wigner–Yanase–Dyson skew information gives computational access to dynamical multipartite entanglement in large subsystems.

## Table: Dynamic QFI Scaling and Metrological Bounds

| Regime                     | QFI Scaling               | Parameter Estimation Bound               |
|----------------------------|---------------------------|------------------------------------------|
| Separable/shot noise       | $F_Q \lesssim N$          | $\delta\theta \sim 1/\sqrt{N}$           |
| Maximally entangled        | $F_Q \sim N^2$            | $\delta\theta \sim 1/N$                  |
| Decohered/open (large $t$) | $F_Q \sim N$ (eventually) | Standard classical limit                  |
| Dynamic enhancement (local kicks, chaotic) | $F_Q \sim L^2$ (for block size $L$ in optimal window) | Macroscopic multipartite entanglement for transient time $t^* \sim L/v_M$ |

Critical points, ergodicity, and non-Markovianity yield nontrivial time and system-size scaling for QFI in many-body protocols.

## Applications and Experimental Realizations

Dynamic QFI protocols inform quantum metrology, sensing, thermodynamics, and quantum control. Criticality (e.g., near quantum phase transitions) enhances sensitivity (divergent QFI) [2508.16413, 2507.05579]. Decoherence-resilient metrological strategies leverage channels and time windows where QFI remains robust. Adaptive measurement routines can exploit analytic expressions for QFI under noise to maximize achievable precision and optimize timing [1403.8033, 2412.01473]. Non-Markovianity witnesses facilitate experimental discrimination of environmental memory and retrodiction [2304.14984, 2102.09586].

## Open Problems and Future Directions

Challenges include exact computation of maximal multi-parameter QFIMs in noisy, strongly interacting systems; extension of dynamic QFI beyond quantum mechanics to generalized probabilistic theories; and experimental implementation of protocols that exploit dynamical enhancement, criticality, and non-Markovianity for robust quantum sensing [1705.08649, 2304.14984]. The relation of QFI to generalized skew information and recovery maps continues to unify operational and geometric perspectives.

Dynamic quantum Fisher information serves as a unifying metric for real-time quantum estimation, entanglement dynamics, resource identification, non-Markovianity detection, and criticality witnessing, with rigorous analytic and numerical foundations in diverse quantum many-body contexts.

Source: https://www.emergentmind.com/topics/dynamic-quantum-fisher-information