Magnetic fields, turbulence and model limits

Width is evidence that needs a model

An interstellar spectral line can be broadened by thermal particle motion, bulk flows, unresolved nonthermal motions, opacity, blended components and instrumental response. A measured width is not automatically a thermometer or a direct measure of turbulent support.

For a classical gas at temperature T, the one-dimensional thermal velocity variance of particles of mass m is k_B T/m, where k_B is Boltzmann’s constant. Heavier particles at the same temperature have a narrower thermal velocity distribution. This is a statement about one velocity component, not a three-dimensional root-mean-square speed.

If independent Gaussian broadening contributions apply in calibrated nonrelativistic velocity coordinates, their variances add. The standard deviations themselves do not add. Optical-depth effects or multiple non-Gaussian components can invalidate a simple Gaussian decomposition.

Comparing lines from different species can help only when their emitting material and response assumptions are compatible. Species that occupy different regions need not share one temperature or one nonthermal distribution. The next calculation stipulates those conditions instead of claiming they hold in every cloud.

A declared two-line experiment

Use two hypothetical co-spatial, optically thin species of masses m0 and 4m0, where m0 is an arbitrary positive mass unit. Let u0 be an arbitrary positive speed unit. Both species share one temperature and the same zero-mean Gaussian nonthermal line-of-sight velocity component.

Suppose both lines have the same calibrated Gaussian instrumental standard deviation 2u0, so instrumental variance is 4u0². Supply observed one-dimensional velocity variances 29u0² for the lighter species and 17u0² for the heavier one. These are invented values, not real species measurements.

Define a = k_B T/(m0u0²) and q = σ_nt²/u0², where σ_nt is the nonthermal standard deviation. The two variance equations are 29 = a + q + 4 and 17 = a/4 + q + 4. Subtracting gives 12 = 3a/4, so a = 16; either equation then gives q = 9.

Thus T = 16m0u0²/k_B and σ_nt = 3u0 under the stipulated model. The thermal variances are 16u0² and 4u0², and the intrinsic thermal-plus-nonthermal variances are 25u0² and 13u0². No numerical temperature in kelvin is implied without physical values for m0 and u0.

Look for correlations across positions and scales

The recovered nonthermal term is what this supplied model assigns after thermal and instrumental contributions are separated. It is not automatically proof of a turbulent cascade. An unresolved ordered flow, multiple emitting structures or an incorrect response model can also complicate a line profile.

Turbulence involves irregular motions with spatial and scale-dependent correlations. Maps of velocity structure and comparisons across separations add information beyond the width of one line. McKee and Ostriker discuss the projection and emissivity complications involved in extracting such information.

Nonthermal motions can redistribute material and can also compress it. Treating every broad line as an isotropic supporting pressure discards the direction, scale and time dependence of the flow. A dynamical interpretation needs the relation between these motions, boundaries and gravity.

The same caution applies to a fitted model: a successful variance decomposition establishes a conditional inference, not a complete cloud simulation or a prediction that collapse is prevented. Its uncertainty must include the validity of shared temperature, shared region, optical thinness and Gaussian independence.

Equal line-of-sight values can hide different magnitudes

Magnetic observations also depend on projection and the measurement regime. In the usual weak-splitting regime for extended non-masing gas, Zeeman inference constrains a tracer-weighted line-of-sight magnetic component. Resolved splitting and some maser regimes require different treatment; the limited statement is not a universal description of every Zeeman observation.

For a separate geometric toy, use uniform magnetic vectors with magnitude B and angle θ to the line of sight. Define positive B_LOS toward the observer for this exercise, without adopting an instrumental polarization-sign convention. The geometric relation is B_LOS = B cosθ.

Model A has B = 5B0 and θ = 60°, giving B_LOS = 2.5B0. Model B has B = 2.5B0 and θ = 0°, giving the same component. Here B0 is an arbitrary positive magnetic-field unit. Their squared magnitudes are 25B0² and 6.25B0², differing by a factor of four.

This comparison does not provide a magnetic force, energy density in physical units, field topology or collapse verdict. Dust polarization can add projected orientation information under alignment and transfer assumptions, but does not by itself supply an exact full three-dimensional field vector.

A field is a physical constraint, not an automatic verdict

Magnetic effects include directional tension and pressure-like stresses. The resulting force depends on spatial structure, so a scalar field magnitude is not a complete force model. A magnetic-field detection alone does not prove that gravity is balanced or that all directions of contraction are prevented.

The ideal conducting-fluid approximation connects fluid motion to magnetic flux through induction, often described as flux freezing. It has conditions involving coupling, scales and conductivity. Field lines are a representation of a vector field rather than material wires, and nonideal effects can allow changes relative to the neutral material.

Interpreting a cloud therefore requires geometry, material distribution, kinematics, field constraints and a model whose approximations fit the regime. Crutcher and Kemball discuss how actual Zeeman measurements and instrumental limitations constrain proposed star-formation pictures.

This chapter has connected gas conditions, dust energy transfer, molecular tracers, formation indicators and feedback to their inference limits. The next chapter examines active galactic centres, where accretion, relativistic motion, imaging and host-galaxy evidence again require several compatible observations.

Worked examples

Try the reasoning

1. Can the listed one-dimensional sigma values be read directly as FWHM or three-dimensional rms speeds?

   Show the answer and reasoning

   No.

   Those are different width conventions; the supplied equations use one-dimensional standard deviations.

2. Is the two-line inversion justified if the species trace different temperatures and nonthermal regions?

   Show the answer and reasoning

   No.

   Its shared-temperature and shared-component assumptions would fail.

3. Does a nonthermal width alone prove an isotropic turbulent pressure that prevents collapse?

   Show the answer and reasoning

   No.

   Flow geometry, correlations, boundaries and gravity require additional evidence.

4. Does ideal flux freezing apply without conditions to every partially ionized cloud and scale?

   Show the answer and reasoning

   No.

   Coupling, conductivity, scale separation and nonideal effects determine whether that approximation is suitable.

What to carry forward

• A conditional two-mass Gaussian model separates variance terms only under its explicit shared-region assumptions.
• Projected magnetic measurements do not uniquely determine the full field or force.
• Cloud dynamics require compatible geometry, kinematics, material and magnetic evidence.