🔬 physics

Statistics of non-conserved observables in Lindblad master equations

This paper derives analytical expressions for the time evolution of the expectation values and second moments of observables that are conserved under Hamiltonian dynamics but become non-conserved due to Markovian environmental coupling, illustrating how collapse operators break conservation laws and generate fluctuations through various quantum models.

Original authors: G. Modanese

Published 2026-06-29

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: G. Modanese

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: When "Rules" Break Down

Imagine you are playing a game of billiards in a perfectly sealed, frictionless room. In this closed world, the laws of physics are strict: if you hit a ball, its total energy and momentum are perfectly conserved. No matter how the balls bounce off each other, the total amount of "movement" in the room never changes. In quantum physics, this is like a closed system governed by a Hamiltonian (the rulebook). If a quantity (like energy or particle count) fits the rules, it stays exactly the same forever.

Now, imagine opening a window in that room. Wind starts blowing in, or maybe a vacuum cleaner starts sucking air out. The room is no longer closed; it's an open system interacting with an environment. Suddenly, the old rules don't work the same way. The wind might push a ball, or the vacuum might steal its momentum.

This paper asks a simple question: What happens to a quantity that was perfectly safe in the closed room, once we open the window?

The author, G. Modanese, shows that just because a quantity is "safe" under the original rules (the Hamiltonian), it doesn't mean it stays safe when the environment interferes. The environment can act like a thief or a generator, changing not just the average amount of that quantity, but also how much it fluctuates (how much it jumps around).

The Two Ways Things Change: The Drift and The Wiggle

The paper introduces a way to measure two different things happening at once:

The Drift (The Average): This is like the average water level in a bucket. If a hole is punched in the bottom, the water level slowly drops. In quantum terms, the "expectation value" (the average) of a conserved quantity might start to decrease or increase because the environment is draining or adding to it.
The Wiggle (The Variance): This is the most interesting part. Imagine the water in the bucket isn't just draining; it's also being shaken violently. Even if the average water level stays the same (maybe water is leaking out at the same rate it's being poured in), the water is sloshing around wildly. The "variance" measures this sloshing.

The paper's main discovery is that the environment can do these things independently:

It can change the average but leave the sloshing alone.
It can leave the average alone but make the sloshing get worse.
It can do both.

The "Source" Analogy

The author uses a clever analogy from plumbing. In a closed pipe, water flows from one end to the other, but the total amount of water in the pipe stays constant.

When you add the environment (the Lindblad equation), it's like someone secretly attaching a hose to the pipe.

If the hose adds water, the pipe fills up (the average goes up).
If the hose sucks water out, the pipe empties (the average goes down).
Crucially: Even if the hose adds and removes water at the exact same rate (so the average stays the same), the act of adding and removing creates turbulence. The water starts splashing and bubbling. The "variance" (the splashing) grows, even though the "average level" looks stable.

The paper provides a mathematical formula to calculate exactly how fast this "splashing" (variance) grows based on how the environment interacts with the system.

Real-World Examples from the Paper

The author uses four specific examples to prove this point:

1. The Leaky Battery (One-Qubit Amplitude Damping)
Imagine a battery that is supposed to hold a charge. In a perfect world, it holds it forever. But if you connect it to a leaky wire (the environment), the charge drains away. The paper shows that as the charge drains, the uncertainty about exactly how much charge is left changes over time. It starts with zero uncertainty (we know it's full), then becomes uncertain as it drains, and finally becomes certain again when it's empty.

2. The Two-Bucket Exchange (Two-Qubit Model)
Imagine two buckets connected by a pipe. Water flows back and forth between them perfectly. The total amount of water in both buckets combined is conserved.
Now, imagine someone starts poking holes in the bottom of each bucket individually.

The total water level will drop (the average changes).
But here's the kicker: Even if you start with a perfectly known amount of water, the holes make the water level in the combined system "fuzzy." You can't predict exactly how much water is left at any given second because the holes are leaking randomly. The "variance" (the fuzziness) grows from zero to a peak and then settles down.

3. The Windy Road (Momentum Diffusion)
Imagine a car driving on a perfectly flat road. Its speed is constant. Now, imagine a strong wind starts blowing from the side.

The wind might push the car left and right equally, so the average speed stays exactly the same.
However, the car is now being jostled violently. Its speed is fluctuating wildly.
The paper shows a case where the average momentum stays perfectly conserved (the car doesn't speed up or slow down on average), but the variance (the jostling) grows linearly. The car is getting more and more "wobbly" even though its average speed is fine.

4. The Light Switch (Jaynes-Cummings Model)
This is a famous model of an atom and a light beam.

Scenario A: If the environment interacts with the atom in a way that respects the total number of light particles (photons), the "count" of photons stays perfectly stable. No drift, no wobble.
Scenario B: If the environment lets photons leak out of the room (like a broken lightbulb), the total count drops, and the uncertainty about the count increases.
The Twist: The paper shows that you can have a situation where the environment causes "dephasing" (messing up the timing of the light waves) without changing the count of photons at all. The "statistics" of the count remain perfect, even though the system is interacting with the environment.

The "Wave-Function" Connection

Finally, the paper touches on a philosophical point. In some theories, the universe "collapses" instantly when we look at it, breaking local conservation laws for a split second before fixing them up again.

The author suggests that the math used for open quantum systems (Lindblad equations) acts like a "soft" version of this collapse. The environment acts like a constant, gentle "source" or "sink" that breaks conservation laws statistically.

If you look at the average of many experiments, the conservation law might look like it's holding (the source and sink cancel out).
But if you look at the fluctuations (the variance), you can see the "scars" of those conservation-breaking events. The math allows us to see the "noise" left behind by the environment, even when the "signal" (the average) looks perfect.

Summary

In short, this paper provides a simple toolkit to measure how the environment messes up the "perfect rules" of quantum mechanics. It teaches us that conservation is fragile. Just because a quantity is conserved in a perfect, isolated world doesn't mean it stays conserved in the real, noisy world. The environment can change the average, or it can just make the quantity jittery, or both. The paper gives us the math to calculate exactly how much "jitter" (variance) is created by the environment.

Technical Summary: Statistics of Non-Conserved Observables in Lindblad Master Equations

Problem Statement
In closed quantum systems, conservation laws are strictly governed by the commutation of an observable $O$ with the Hamiltonian $H$ ( $[H, O] = 0$ ), ensuring that the expectation value $\langle O \rangle$ remains constant for all initial states. However, in open quantum systems described by Markovian environments via the Lindblad master equation, this condition is insufficient. Dissipative terms can break Hamiltonian conservation laws, causing formerly conserved quantities to evolve. While the loss of conservation in the expectation value is well-understood, the impact on the full statistical distribution—specifically the second moment $\langle O^2 \rangle$ and the variance $\text{Var}(O)$ —requires a more nuanced analysis. The paper addresses how collapse operators modify the statistics of observables that are conserved by the Hamiltonian but not by the full open-system dynamics.

Methodology
The author employs the adjoint Lindblad formalism to derive elementary expressions for the time evolution of the first and second moments of an observable $O$ .

Adjoint Generator: Starting from the standard Lindblad master equation for the density matrix $\rho$ , the time derivative of the expectation value is expressed via the adjoint generator $\mathcal{L}^\dagger$ :
$\frac{d}{dt}\langle O \rangle = \langle \mathcal{L}^\dagger(O) \rangle$
where $\mathcal{L}^\dagger(O) = i[H, O] + \mathcal{D}^\dagger(O)$ , and $\mathcal{D}^\dagger$ represents the dissipative contribution.
Moment Equations: The formalism is extended to the second moment $\langle O^2 \rangle$ and the variance $\text{Var}(O) = \langle O^2 \rangle - \langle O \rangle^2$ . The paper derives the variance-production rate $\Pi_O(t)$ , defined as:
$\Pi_O(t) = \langle \mathcal{L}^\dagger(O^2) \rangle - 2\langle O \rangle \langle \mathcal{L}^\dagger(O) \rangle$
For Hamiltonian-conserved observables ( $[H, O]=0$ ), this rate is generated entirely by the dissipative part $\mathcal{D}^\dagger$ .
Analytic Models: The derived formulae are applied to four specific analytic models to illustrate distinct dynamical behaviors:
- One-qubit amplitude damping.
- A two-qubit exchange model with local decay.
- A momentum-diffusion model.
- The Jaynes–Cummings model with different coupling types.

Key Contributions and Results

Distinction Between Mean and Variance Conservation: The paper establishes that the conservation of an expectation value does not imply the conservation of the observable's full statistics. Three distinct cases are identified:
1. Full Conservation: $\mathcal{L}^\dagger(O) = 0$ and $\mathcal{L}^\dagger(O^2) = 0$ . Both mean and variance are constant.
2. Mean Drift: $\mathcal{L}^\dagger(O) \neq 0$ . The mean value changes, typically accompanied by variance changes.
3. Variance Generation with Mean Conservation: $\mathcal{L}^\dagger(O) = 0$ but $\mathcal{L}^\dagger(O^2) \neq 0$ . The mean remains constant while fluctuations grow.
Analytic Examples:
- One-Qubit Amplitude Damping: For $O=\sigma_z$ and a collapse operator $L=\sigma_-$ , the mean $\langle \sigma_z \rangle$ decays to the ground state. Since $\sigma_z^2 = I$ , the second moment is trivial, and the variance evolves solely due to the decay of the mean.
- Two-Qubit Exchange Model: With a Hamiltonian conserving the total excitation number $N$ , but local decay operators $L_j = \sigma_-^{(j)}$ , the mean $\langle N \rangle$ decays. Crucially, even if the system starts in an eigenstate of $N$ (zero variance), the dissipative dynamics dynamically generates a non-zero variance as the system transitions between excitation sectors ( $|11\rangle \to |10\rangle/|01\rangle \to |00\rangle$ ).
- Momentum Diffusion: For a particle with $H=p^2/2m$ and a position-coupled dissipator $L \propto x$ , the mean momentum is conserved ( $\mathcal{L}^\dagger(p)=0$ ), but the variance grows linearly ( $\mathcal{L}^\dagger(p^2) \neq 0$ ). This explicitly demonstrates a case where the first moment is preserved while the distribution broadens.
- Jaynes–Cummings Model: The paper contrasts two scenarios for the total excitation number $N$ $N$ .
  - Dephasing Coupling: If the reservoir couples via $N$ itself, $N$ and $N^2$ are conserved ( $\mathcal{L}^\dagger(N)=\mathcal{L}^\dagger(N^2)=0$ ). The reservoir destroys coherences between $N$ -sectors but leaves the statistics of $N$ unchanged.
  - Dissipative Coupling: Standard photon loss ( $L_a=a$ ) and atomic decay ( $L_\sigma=\sigma_-$ ) break the conservation of both the mean and the second moment of $N$ .
Relation to Fluctuation-Dissipation: The variance-production term $\Pi_O(t)$ is identified as the natural object linking dissipative drift and fluctuation generation. In thermal models, this formalism recovers the standard fluctuation-dissipation balance, where the same environmental coupling that causes damping also fixes the equilibrium variance.

Significance and Claims
The paper claims to provide a "simple and operational" diagnostic for the transient statistics of Hamiltonian-conserved observables in open systems. Its primary significance lies in isolating the role of the dissipative part of the adjoint generator in generating fluctuations, independent of the drift of the mean.

The author emphasizes that this work is not a new classification of invariant operators or stationary subspaces, but rather a complement to rigorous theories of Lindblad symmetries, focusing on analytically transparent models. Furthermore, the paper draws a parallel between the Lindblad source terms and idealized wave-function reduction models. It suggests that while local conservation laws may be violated event-by-event in reduction models (or by Lindblad operators), they may hold statistically after averaging. In this context, the second moment and variance serve as carriers of information about the underlying stochastic non-conservation events, even when the mean value appears conserved.