Measurement and Open-System Reference

Use this chapter to locate a definition, compare standard models, or translate a formula before applying it. Each compact entry links to the explanation of its assumptions. If the topic is new, begin with the linked article and return to the table once the physical meaning is clear.

Required background. The volume introduction distinguishes outcomes, channels, reduced dynamics, and monitored records.

Find the quantity you need

Need	Reference page
Translate rates, tensor order, noise spectra, and heat signs	Notation and Conventions
Check what probabilities, channels, and trajectories do and do not specify	Common Misconceptions
Distinguish nearby terms such as dephasing, relaxation, and thermalization	Glossary
Recall a state update, channel condition, or dynamical equation	Formula Sheet
Compare finite-time noise transformations	Common Channels
Compare differential equations and their regimes	Common Master Equations
Translate a jump operator into population and coherence changes	Common Lindblad Operators
Compare spectral shapes, ordering, and cutoffs	Common Noise Spectra
Test Born, Markov, secular, and related assumptions	Approximation Checklist
Find a named system–bath model and its detailed treatment	Model Index
Choose a numerical experiment	Notebook Index
Choose a textbook, review, or original paper for a question	Reading List

Match the object before matching the symbols

A POVM gives probabilities. An instrument adds the conditional output state. A channel gives an averaged finite transformation. A generator describes evolution in time. Their formulas can share symbols without supplying the same information.

For example, one input and output state do not determine a channel. Even complete knowledge of a channel at one time need not determine a unique generator, and a matrix logarithm need not have GKLS form. Similarly, a master equation does not determine which field observable a detector monitors. One Interaction, Different Measurement Records connects these distinctions through an ancilla experiment.

When a measurement outcome is retained, normalize its operation only after calculating the outcome probability. When it is ignored, sum the unnormalized operations. When several Kraus operators describe one reported outcome, sum their operations before normalizing; they are not necessarily separate observed events.

Translate conventions at the level of a prediction

The quickest way to resolve a factor-of-two ambiguity is often to calculate the same observable in both conventions.

For the simple vacuum-loss model $\dot\rho=\kappa\mathcal D[a]\rho$, in the interaction picture,

$$\frac{d}{dt}\langle a\rangle=-\frac\kappa2\langle a\rangle, \qquad \frac{d}{dt}\langle a^\dagger a\rangle=-\kappa\langle a^\dagger a\rangle.$$

Thus field amplitude and occupation have different decay rates. These are two consequences of the same specified generator; one cannot infer a general occupation law merely from the mean amplitude, since states with zero mean amplitude can have nonzero occupation.

The Conventions page also fixes the dissipator normalization, Choi tensor order and partial trace, column vectorization, stochastic increments, and heat signs. For noise spectra, compare the full Fourier definition, operator ordering, frequency units, and one-sided or two-sided normalization. The symbol $S(\omega)$ alone does not identify those choices.

Use a compact formula with a physical regime

Before interpreting an exponential rate, identify the state or observable that decays, the time interval used, and the assumed bath model. Before interpreting a thermal steady state, identify the Hamiltonian used in the rate balance. Before interpreting a trajectory, identify the measured record and unobserved channels.

The Approximation Checklist connects these questions to timescale and coupling tests. The Computational Notebooks distinguishes numerical integration, truncation, and sampling errors from errors in the physical approximation. A correct reference formula does not ensure that its implementation or its application to a new regime is correct.

For literature study, combine an explanatory textbook with the source of the particular result and a methods or platform treatment using comparable assumptions. Use software documentation for interface and implementation details, and physics references for the model’s validity.

Exercises

One channel, multiple generators. At a fixed time $T>0$, compare the qubit channels generated by $H_1=0$ and $H_2=\pi\hbar Z/T$. Do they agree at $T$? Do they agree at intermediate times?

Solution

At $T$, $U_1=I$ and $U_2=e^{-i\pi Z}=-I$. Both produce the identity channel because a global phase cancels in $U\rho U^\dagger$. At intermediate times the second Hamiltonian rotates transverse Bloch components while the first does nothing. Even an exactly known finite-time channel need not select a unique Hamiltonian generator.

Zero amplitude does not mean vacuum. A cavity starts in the one-photon number state and follows the vacuum-loss generator above. Find its mean amplitude and mean occupation.

Solution

The mean amplitude is initially zero and remains zero. The mean occupation begins at one and is $e^{-\kappa t}$. A detector can therefore register photons even though the mean field amplitude vanishes. Mean amplitude, intensity, and counting statistics are distinct observables.

References

• Breuer, H.-P., and F. Petruccione. The Theory of Open Quantum Systems. Oxford University Press, 2002. Models, approximations, and master equations.
• Gardiner, C. W., and P. Zoller. Quantum Noise. 3rd edition, Springer, 2004. Field moments, spectra, and output signals.
• Nielsen, M. A., and I. L. Chuang. Quantum Computation and Quantum Information. 10th anniversary edition, Cambridge University Press, 2010. Channels and measurement representations.
• Wiseman, H. M., and G. J. Milburn. Quantum Measurement and Control. Cambridge University Press, 2010. Conditional records and feedback.