Reference

The reference chapter is a navigation and checking layer for measurement theory, decoherence, quantum channels, noise, and open-system dynamics. Use it to locate a definition, recognize a standard equation, compare conventions, choose a model, or find a fuller derivation.

A reference entry is not a license to transplant a formula without its assumptions. The compact pages deliberately link back to canonical articles that explain where a result comes from, what it predicts, and when it fails.

Choose a Reference Page

If you need to…	Start here
identify a term or distinguish nearby terms	Glossary
recall a standard equation or state update	Formula Sheet
compare dephasing, damping, depolarizing, erasure, or Gaussian maps	Common Channels
recognize a dephasing, damping, thermal, optical, Redfield, or rate equation	Common Master Equations
interpret a commonly used collapse operator	Common Lindblad Operators
compare white, Ohmic, Lorentzian, low-frequency, thermal, or vacuum spectra	Common Noise Spectra
audit Born, Markov, secular, positivity, and steady-state assumptions	Approximation Checklist
find the canonical home of a named model	Model Index
choose a computational notebook contract	Notebook Index
choose a textbook, classic paper, review, or software reference	Reading List

The fastest reliable workflow is:

$$\begin{gathered} \text{identify the object} \\ \downarrow \\ \text{check the convention} \\ \downarrow \\ \text{check the regime} \\ \downarrow \\ \text{follow the canonical derivation}. \end{gathered}$$

Reading Path

First encounter

Begin with the Glossary and Formula Sheet. Use them to establish vocabulary and locate the relevant chapter. Do not try to learn an unfamiliar subject entirely from compact entries.

Solving a standard problem

Use Common Channels, Common Master Equations, Common Lindblad Operators, and Common Noise Spectra as comparison tables. Then read the linked canonical page for the model you actually use.

Deriving or fitting an effective model

Use the Approximation Checklist before accepting a weak-coupling or Markovian reduction. The Model Index helps distinguish models that share similar equations but assume different system boundaries or parameter regimes.

Computing and researching

Use the Notebook Index to choose a validated computational baseline and the Reading List to move into the literature.

Identify the Mathematical Object

Several objects that look similar answer different questions.

Object	Input and output	What it supplies
POVM	state to outcome probabilities	measurement statistics
instrument	state to outcome-resolved output states	statistics and backaction
channel	input state to unconditional output state	finite-step physical evolution
dynamical map	initial state to state at time $t$	time-indexed reduced evolution
generator	current state to time derivative	infinitesimal evolution law
noise spectrum	correlation function to frequency-domain weight	environmental fluctuation content
trajectory equation	state and record increment to conditional update	monitored single-record evolution
rate equation	populations to population derivatives	incoherent transition dynamics

Using the right formula begins by naming the right object. A POVM does not determine a unique state update. A channel at one fixed time does not determine a unique continuous-time generator. A Lindblad generator does not determine a unique monitored record.

Compact Structural Formulas

These formulas are signposts. Their linked pages own the definitions, derivations, and caveats.

Measurement statistics

For POVM effects $\{E_i\}$,

$$\begin{aligned} p(i) &= \operatorname{Tr}(\rho E_i), \\ E_i &\ge0, \\ \sum_iE_i &=I. \end{aligned}$$

An instrument $\mathcal I_i$ additionally gives the conditional output

$$\rho_i = \frac{\mathcal I_i(\rho)} {\operatorname{Tr}[\mathcal I_i(\rho)]}.$$

See Generalized Measurements and Instruments.

Quantum channels

A finite-dimensional channel can be written

$$\Phi(\rho) = \sum_\alpha K_\alpha\rho K_\alpha^\dagger,$$

with trace preservation when

$$\sum_\alpha K_\alpha^\dagger K_\alpha =I.$$

The Kraus representation is not unique. See Quantum Channels and Noise.

Markovian generators

A time-independent Lindblad–GKSL equation has the structure

$$\begin{aligned} \dot\rho = {}& -\frac{i}{\hbar}[H,\rho] \\ &+ \sum_\mu \left( L_\mu\rho L_\mu^\dagger - \frac{1}{2} \{L_\mu^\dagger L_\mu,\rho\} \right). \end{aligned}$$

This form guarantees a completely positive trace-preserving semigroup under its standard finite-dimensional assumptions. It does not prove that the Markov or secular approximations fit a given experiment. See Markovian Master Equations.

Noise spectra

For a stationary bath operator $B(t)$, one two-sided ordered convention is

$$S_{BB}(\omega) = \int_{-\infty}^{\infty} dt\, e^{i\omega t} \langle B(t)B(0)\rangle.$$

Another source may use a symmetrized spectrum, a one-sided transform, the opposite Fourier sign, or cyclic rather than angular frequency. See Noise Spectra before comparing numerical values.

Convention Checks

A formula copied from a trustworthy source can still be wrong in a new calculation if conventions do not match.

Rates and linewidths

Check whether a quoted rate describes amplitude decay, energy decay, coherence decay, a half-width, or a full width. If an amplitude obeys

$$\langle a(t)\rangle = \langle a(0)\rangle e^{-\kappa t/2},$$

then the occupation obeys

$$\langle a^\dagger a\rangle_t = \langle a^\dagger a\rangle_0 e^{-\kappa t}$$

for the corresponding simple loss model. Calling both exponents “the linewidth” without qualification creates factors-of-two errors.

Dephasing conventions

A collapse term written as

$$\gamma_\phi \mathcal D[\sigma_z]\rho$$

and one written as

$$\frac{\Gamma_\phi}{2} \mathcal D[\sigma_z]\rho$$

use different parameter names and may encode the same coherence-decay rate. Compute the off-diagonal equation directly before identifying the parameter with $1/T_\phi$.

Frequency units

Angular frequency and cyclic frequency satisfy

$$\omega=2\pi f.$$

Ratios are unchanged only when all quantities use the same convention. Mixed quoted units can corrupt detunings, spectral densities, cooperativities, and solver time scales.

Tensor and basis order

State whether a composite basis is ordered as system–environment or environment–system, and whether the computational basis begins with the ground or excited state. Matrix formulas do not reveal this choice.

Stochastic calculus

An Itō stochastic equation and its Stratonovich form have different drift terms. A reference formula must be used with the numerical method and noise convention for which it was derived.

Channels, Generators, and Models

A channel describes a finite transformation. A generator describes an infinitesimal law. A microscopic model proposes why either arises.

If

$$\Phi_t=e^{t\mathcal L},$$

then $\mathcal L$ generates a time-homogeneous semigroup. Not every channel belongs to a semigroup, and not every family of channels is divisible into completely positive intermediate maps. Taking a matrix logarithm of one channel does not automatically produce a physically valid Lindblad generator.

Use the reference pages in this order:

1. identify the observed finite-time behavior with Common Channels;
2. identify a candidate differential model with Common Master Equations;
3. interpret operator choices with Common Lindblad Operators;
4. audit the derivation with the Approximation Checklist;
5. locate nearby microscopic models in the Model Index.

This sequence avoids treating a phenomenological fit as a microscopic explanation.

Noise Spectra and Detailed Balance

Classical stationary noise often has a symmetric two-sided power spectrum. Quantum ordered spectra need not satisfy

$$S_{BB}(-\omega) = S_{BB}(+\omega).$$

In thermal equilibrium, positive- and negative-frequency components are related by a detailed-balance condition whose exponential sign depends on the transform and operator-ordering convention. This asymmetry encodes the environment’s unequal ability to absorb and emit energy.

Before using a tabulated spectrum, record:

• whether it is classical, ordered quantum, or symmetrized;
• whether it is one-sided or two-sided;
• whether the argument is $f$ or $\omega$;
• the units of the fluctuating quantity and the spectrum;
• infrared and ultraviolet cutoffs;
• temperature and equilibrium assumptions;
• which system operator samples the spectrum.

The Common Noise Spectra page compares standard shapes without hiding these choices.

Approximation Audit

The Approximation Checklist is organized around a derivation chain rather than the visual form of the final equation.

At minimum, ask:

1. What is the retained system?
2. What bath state and initial correlations are assumed?
3. How weak is the coupling relative to relevant system and bath scales?
4. How short is the bath correlation time?
5. Which Bohr frequencies are resolved?
6. Is the secular approximation justified near degeneracies?
7. Does the reduced equation preserve trace and Hermiticity?
8. Is positivity or complete positivity required and satisfied?
9. Does the stationary state obey the intended detailed balance?
10. Do zero-coupling, zero-temperature, high-temperature, and short-time limits behave correctly?

A GKSL-looking equation can still use incorrect rates or the wrong stationary state. Conversely, a controlled short-time nonsecular equation may be useful even if it is not of semigroup form. The claim must match the approximation.

Model Index

The Model Index routes named models to their canonical homes. Use it when a phrase such as “spin–boson model,” “Caldeira–Leggett model,” “collision model,” or “damped Jaynes–Cummings model” appears without enough context.

Two models may share the same reduced equation in one limit while differing in:

• the system–environment boundary;
• bath statistics and spectral density;
• weak- or strong-coupling assumptions;
• rotating-wave or secular approximations;
• initial-state restrictions;
• observables that remain trustworthy;
• behavior outside the fitted regime.

The model name is therefore a starting point, not a complete specification.

Notebook and Evidence Status

The Notebook Index organizes computational contracts by skill and physical task. It links to channel, Lindblad, trajectory, non-Markovian, decoherence, control, and thermodynamic examples.

Notebook status must remain explicit:

Status	Meaning
specification	the intended model, workflow, and tests are documented
executable	the implementation runs in a recorded environment
validated	analytic, structural, and convergence tests pass
reproduced	an independent clean run regenerates accepted outputs
benchmarked	an independent method or trusted data agree within tolerance

The Computational Notebooks gateway owns the chapter-wide validation workflow. Site-wide promotion policy lives in Reproducibility Status.

Reading the Literature

The Reading List is organized by purpose rather than prestige. A graduate reading route usually benefits from three layers:

1. a textbook or pedagogical review for notation and conceptual structure;
2. a canonical paper or rigorous source for the result being used;
3. a platform or methods source that matches the actual approximation regime.

Software documentation explains an interface, not the validity of the physical model. A classic paper establishes historical priority and a specific result, not automatic suitability for every current application.

When sources disagree, first compare definitions, Fourier signs, rate conventions, basis choices, and approximation order. Many apparent contradictions disappear at that level. Genuine disagreements should be stated rather than averaged away.

Reference-to-Derivation Workflow

Suppose a calculation contains a damped qubit with a measured coherence time.

1. Use the Glossary to distinguish relaxation, pure dephasing, and total coherence decay.
2. Use the Formula Sheet to identify the standard $T_1$, $T_2$, and $T_\phi$ relation for the simple Markovian model.
3. Use Common Lindblad Operators to check the operator and factor convention.
4. Use Common Noise Spectra if the rate is derived from environmental fluctuations.
5. Use the Approximation Checklist to ask whether one exponential rate is justified.
6. Follow the links to the canonical dephasing and amplitude-damping derivations.
7. Use Decoherence Timescale Estimation if numerical fitting or filter functions are needed.

This pattern generalizes: references identify and check; canonical pages explain; notebooks validate computations.

Canonical Boundaries

This chapter owns compact lookup, routing, and audit pages for this volume.

• Measurement Theory owns projective measurement models and state-update rules.
• Generalized Measurements and Instruments owns POVMs, operations, instruments, and dilation results.
• Quantum Channels and Noise owns channel structure and standard channel derivations.
• Decoherence and the Classical Transition owns decoherence mechanisms and their interpretive limits.
• Open Quantum Systems owns reduced dynamics and system–bath approximation chains.
• Markovian Master Equations owns Lindblad–GKSL dynamics.
• Continuous Measurement and Quantum Trajectories owns record-conditioned evolution.
• Quantum Noise, Dissipation, and Baths owns spectra, correlations, and response.
• Computational Notebooks owns numerical admission contracts and validation workflow.

Reference pages should remain compact enough to scan and rich enough to prevent a category or convention error. Long derivations belong at their canonical homes.

Common Mistakes

• Learning an unfamiliar subject only from a formula sheet.
• Treating a POVM, instrument, channel, and generator as interchangeable.
• Assuming Kraus or Lindblad operators are unique.
• Copying a rate without checking amplitude, energy, and linewidth conventions.
• Comparing spectra with different Fourier signs, orderings, or units.
• Inferring a Markovian generator from one finite-time channel without an embeddability check.
• Treating a phenomenological equation as a microscopic derivation.
• Applying a named model without specifying the system boundary and parameter regime.
• Using a reference table while ignoring its linked canonical caveats.
• Calling a notebook reproduced merely because its code executed once.
• Citing a software manual for a physical approximation rather than for the implementation.
• Assuming a page reviewed recently has resolved every active research question.

Exercises

Route the question

For each task, choose the best first reference page: identifying whether $\sigma_z$ causes pure dephasing, comparing an Ohmic and Lorentzian environment, finding the canonical spin–boson page, and choosing a trajectory simulation.

Solution

Use Common Lindblad Operators for the usual role and convention of $\sigma_z$. Use Common Noise Spectra for Ohmic and Lorentzian shapes. Use the Model Index for the spin–boson model. Use the Notebook Index to choose between jump and diffusive trajectory contracts.

Each answer is only the first stop. The linked canonical page should be read before a formula is adopted.

Channel or generator?

An experiment reports only an input state and an output state after $10\,\mu\mathrm{s}$. Does that observation determine a unique Lindblad generator?

Solution

No. Even complete knowledge of one finite-time channel need not determine a unique physically valid continuous-time generator. A matrix logarithm has branch ambiguities, and the channel may not be embeddable in a time-homogeneous completely positive semigroup. With only one input and one output state, even the channel itself is underdetermined.

The appropriate first object is a finite-time channel or a restricted process model. Inferring a generator requires additional time-resolved data and structural assumptions.

Dephasing factor

For the convention

$$\dot\rho = \frac{\Gamma}{2} \mathcal D[\sigma_z]\rho,$$

find the decay equation for $\rho_{01}$.

Solution

Since $\sigma_z^\dagger\sigma_z=I$,

$$\mathcal D[\sigma_z]\rho = \sigma_z\rho\sigma_z-\rho.$$

The off-diagonal element changes sign under conjugation by $\sigma_z$, so

$$\left( \mathcal D[\sigma_z]\rho \right)_{01} = -2\rho_{01}.$$

Therefore

$$\dot\rho_{01} = -\Gamma\rho_{01}.$$

In this convention $\Gamma$ is the pure coherence-decay rate. If the prefactor were written differently, the symbol attached to the dissipator would not have the same meaning.

Spectrum convention

Why can a symmetrized quantum noise spectrum be insufficient for predicting upward and downward transition rates separately?

Solution

A symmetrized spectrum combines positive- and negative-frequency ordered correlations. Upward and downward transitions sample the environment’s emission and absorption capacities separately, which are encoded in the asymmetric ordered spectrum. Symmetrization can erase that distinction.

The reference must therefore specify operator ordering and Fourier convention before a spectral value is converted into a transition rate.

Reference versus validation

A notebook uses a standard amplitude-damping formula and produces the expected-looking population curve. Name three additional checks needed before calling the output validated.

Solution

Suitable checks include:

1. verify trace, Hermiticity, and positivity throughout the evolution;
2. compare with the exact exponential population law for a declared initial state;
3. refine the time step or solver tolerances and report the observable residual;
4. check the long-time ground-state limit;
5. compare an ODE solution with a matrix exponential or channel implementation.

The standard reference formula identifies the target behavior. Validation establishes that the implementation realizes it accurately.

Cross-Links

• Glossary
• Formula Sheet
• Common Channels
• Common Master Equations
• Common Lindblad Operators
• Common Noise Spectra
• Approximation Checklist
• Model Index
• Notebook Index
• Reading List
• Concept Map
• Applications and Experimental Platforms
• Computational Notebooks

References

• H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems, Oxford University Press (2002).
• C. W. Gardiner and P. Zoller, Quantum Noise, 3rd ed., Springer (2004).
• H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control, Cambridge University Press (2010).
• M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, 10th anniversary ed., Cambridge University Press (2010).
• M. M. Wilde, Quantum Information Theory, 2nd ed., Cambridge University Press (2017).
• R. Alicki and K. Lendi, Quantum Dynamical Semigroups and Applications, 2nd ed., Springer (2007).
• U. Weiss, Quantum Dissipative Systems, 4th ed., World Scientific (2012).
• Á. Rivas and S. F. Huelga, Open Quantum Systems: An Introduction, Springer (2012).