Concept Map

This map shows how the main objects in measurement, decoherence, and open systems fit together. It is a routing page: use it to decide whether a problem calls for a projector, a POVM, an instrument, a channel, a reduced system-environment model, a master equation, or a trajectory description.

An arrow means “is naturally modeled by” or “leads to after a modeling choice.” It does not mean logical equivalence. For example, a channel can arise by ignoring a measurement outcome, by tracing out an environment, or by using a phenomenological noise model; those are different physical stories that can produce the same map $\Phi$.

At a Glance

Question	Object to use	What it answers	Existing entry point
What probabilities does an ideal sharp measurement assign?	projectors $\{P_a\}$	outcome probabilities and ideal update	Projective Measurement
What if the detector is noisy, inefficient, or coarse grained?	POVM effects $\{E_i\}$	outcome probabilities only	POVMs: First Encounter
What happens to the state after each outcome?	instrument maps $\mathcal I_i$	outcome probability plus conditional output state	Generalized Measurements Overview
What happens if outcomes are ignored?	quantum channel $\Phi$	unconditional state transformation	Density Operators
What happens if an environment is ignored?	partial trace over $E$	reduced dynamics for $S$	Partial Trace
Why do interference terms disappear locally?	decoherence model	suppression of reduced-state coherences	Decoherence Preview
What is the continuous-time reduced evolution?	master equation	generator of open-system dynamics	Liouville–von Neumann Equation

Measurement Chain

The measurement side begins with an outcome set. In the projective idealization, outcomes are represented by orthogonal projectors:

$$P_aP_b=\delta_{ab}P_a, \qquad \sum_aP_a=I.$$

The probability and selective update are

$$p(a)=\operatorname{Tr}(\rho P_a), \qquad \rho_a = \frac{P_a\rho P_a}{p(a)}.$$

When the measurement is not sharp, the probability data are represented by POVM effects:

$$E_i\ge0, \qquad \sum_iE_i=I, \qquad p(i)=\operatorname{Tr}(\rho E_i).$$

The effects do not determine the post-measurement state. That is the conceptual reason instruments are needed. An instrument is a family of completely positive maps $\{\mathcal I_i\}$ whose traces give the outcome probabilities:

$$p(i) = \operatorname{Tr}\mathcal I_i(\rho), \qquad \rho_i = \frac{\mathcal I_i(\rho)} {\operatorname{Tr}\mathcal I_i(\rho)}.$$

The nonselective operation is obtained by summing over outcomes:

$$\Phi(\rho) = \sum_i \mathcal I_i(\rho).$$

So the measurement chain is:

projectors
  -> POVM effects for probabilities
  -> instruments for outcome-conditioned states
  -> channel after outcomes are ignored

Open-System Chain

The open-system side begins with a larger Hilbert space:

$$\mathcal H_{SE} = \mathcal H_S\otimes\mathcal H_E.$$

If the total state evolves unitarily but only $S$ is retained, the reduced state is

$$\rho_S(t) = \operatorname{Tr}_E\rho_{SE}(t).$$

For an initially uncorrelated state,

$$\rho_{SE}(0) = \rho_S(0)\otimes\eta_E,$$

the reduced dynamics defines a channel:

$$\Phi_t(\rho_S) = \operatorname{Tr}_E \left[ U_{SE}(t) (\rho_S\otimes\eta_E) U_{SE}^\dagger(t) \right].$$

Thus the open-system chain is:

system plus environment
  -> unitary joint dynamics
  -> trace out ignored degrees of freedom
  -> channel for the system
  -> master equation if a time-local generator is valid

The assumptions matter. Initial correlations, strong coupling, long bath memory, and coarse-graining choices can change whether a simple channel family or Markovian master equation is appropriate.

Decoherence Chain

Decoherence is not a separate postulate. It is a consequence of system-environment correlations in which different system alternatives become correlated with distinguishable environmental states:

$$\left(c_0\lvert0\rangle+c_1\lvert1\rangle\right) \lvert E_0\rangle \longrightarrow c_0\lvert0\rangle\lvert E_0'\rangle + c_1\lvert1\rangle\lvert E_1'\rangle.$$

After tracing over the environment, the off-diagonal terms in the system density matrix are multiplied by environmental overlaps:

$$\rho_{01} \longmapsto \rho_{01}\langle E_1'|E_0'\rangle.$$

When $\langle E_1'|E_0'\rangle\approx0$, interference between $\lvert0\rangle$ and $\lvert1\rangle$ is suppressed for measurements on $S$ alone. The total state may still be coherent. The lost interference has moved into correlations with degrees of freedom that the reduced description ignores.

The decoherence chain is:

system alternatives
  -> environment records distinguishing information
  -> reduced off-diagonal terms shrink
  -> pointer basis becomes dynamically preferred
  -> classical-looking records become stable in practice

The final step is a physical stability statement, not a proof that an interpretation problem has vanished.

Continuous-Time Chain

Closed density-operator dynamics is governed by the Liouville–von Neumann equation:

$$\frac{d\rho}{dt} = -\frac{i}{\hbar}[H,\rho].$$

Open-system dynamics adds nonunitary terms. In the Markovian completely positive case, the standard generator has the Lindblad–Gorini–Kossakowski–Sudarshan form:

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

The continuous-time chain is:

channel family Phi_t
  -> time-local generator if appropriate
  -> Hamiltonian part plus dissipative part
  -> jump operators encode noise, loss, or monitoring channels

This chain is often the right one for relaxation, dephasing, spontaneous-emission models, and Markovian noise. It is the wrong chain when bath memory, strong coupling, initial correlations, or nonsecular terms are central to the question.

Conditional and Unconditional Descriptions

Many confusions come from mixing conditional and unconditional states.

Situation	State description
outcome $i$ is known	conditional state $\rho_i$
outcome exists but is ignored	nonselective state $\sum_i\mathcal I_i(\rho)$
continuous record is monitored	stochastic conditional state
record is averaged over	deterministic master equation
environment is included	joint state $\rho_{SE}$
environment is ignored	reduced state $\rho_S=\operatorname{Tr}_E\rho_{SE}$

The same laboratory setup may admit several descriptions. The correct one depends on what information is available and what prediction is being made.

Boundary Map

This volume uses, but does not duplicate, several canonical homes:

• Core Formalism owns the minimal Born rule, state language, observables, and postulate-level update rules.
• Composite Systems and Entanglement owns tensor products, partial trace, reduced states, and entanglement as composition structure.
• Quantum Dynamics owns closed-system pictures of motion and the density-operator equation for unitary dynamics.
• This volume owns measurement instruments, channels, decoherence, reduced dynamics, master equations, trajectories, and open-system modeling assumptions.

The boundary is not bureaucratic; it prevents duplicate derivations from drifting apart.

Practice Check

A detector has two macroscopic outcomes, but each outcome can arise from several microscopic detector histories that disturb the system differently. Is a POVM enough to predict the post-measurement state?

Solution

No. The POVM effects determine the probabilities of the macroscopic outcomes, but not the output state for each outcome. To predict the post-measurement state, one needs an instrument, or equivalently a specified set of outcome-resolved completely positive maps. Different instruments can have the same POVM and different backaction.

References

• H.-P. Breuer and F. Petruccione, The Theory of Open Quantum Systems, Oxford University Press, 2002.
• M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, Cambridge University Press, 2010.
• H. M. Wiseman and G. J. Milburn, Quantum Measurement and Control, Cambridge University Press, 2010.
• M. Schlosshauer, Decoherence and the Quantum-to-Classical Transition, Springer, 2007.