Escape Probability
Radiative line emission is a key mechanism regulating the thermal balance of astrophysical gas. In many interstellar environments, however, spectral lines become optically thick, and photons emitted in atomic or molecular transitions may undergo multiple absorption and re-emission events before escaping the medium. A full treatment of this problem requires solving the radiative transfer equation, which is computationally demanding, particularly in multi-level and multi-dimensional systems. This formalism provides a physically motivated approximation that captures the essential effects of radiative trapping while remaining tractable.
The escape probability, usually denoted by \(\beta\), represents the probability that a photon emitted in a given transition can leave the system without being reabsorbed. In optically thin conditions (i.e. where \(\tau \ll 1\)), photons escape freely and \(\beta \rightarrow 1\). In optically thick media (i.e. where \(\tau \gg 1\)), photons are trapped and \(\beta \ll 1\), reducing the effective cooling rate.
Statistical Equilibrium and Radiation Field
Consider a parcel of gas that emits and absorbs radiation through bound–bound transitions between discrete energy levels. Under the assumption of statistical equilibrium, the population of each level is governed by the balance between all upward and downward radiative and collisional processes,
where \(n_i\) and \(n_j\) are the populations of levels \(i\) and \(j\), respectively, and \(R_{ij}\) denotes the total transition rate from level \(i\) to \(j\).
For a given transition, the rate coefficients include spontaneous emission, stimulated emission or absorption, and collisional excitation or de-excitation. These rates depend on the local radiation field through the angle-averaged mean intensity \(\langle J_{ij} \rangle\), which couples the level populations to the surrounding medium.
Escape Probability Approximation
In the escape probability (or large velocity gradient, LVG; Sobolev 1960; Castor 1970; de Jong et al. 1975; Poelman & Spaans 2005) approximation, the mean radiation field is expressed as a weighted sum of locally produced line radiation and an external background field,
where \(S_{ij}\) is the line source function, \(B(\nu_{ij})\) represents the background radiation at the transition frequency \(\nu_{ij}\), and \(\beta_{ij}\) is the escape probability.
The source function for a transition between levels \(i\) and \(j\) is determined by the local level populations and statistical weights,
where \(g_i\) and \(g_j\) are the statistical weights of the upper and lower levels, respectively.
The escape probability \(\beta_{ij}\) represents the probability that a photon emitted in the transition \(i \rightarrow j\) leaves the system without being reabsorbed. In optically thin conditions, \(\beta_{ij} \rightarrow 1\), while in optically thick media radiative trapping reduces \(\beta_{ij}\) and suppresses the effective radiative cooling.
Optical Depth Dependence
The escape probability is primarily determined by the line optical depth \(\tau_{ij}\), which measures the opacity of the medium to photons of frequency \(\nu_{ij}\). For line radiation, and neglecting continuum absorption, the optical depth along a given direction is given by
where \(A_{ij}\) is the Einstein coefficient for spontaneous emission, \(u\) is the local velocity dispersion (including thermal and non-thermal contributions), and the integral is performed along the photon path.
Analytical Expression
The analytical expression of the escape probability (de Jong et al. 1975) is given by
For simple geometries, the above expression is further simplified forming the basis of many non-LTE excitation and cooling models used in studies of the ISM. For example, the expression
is frequently used for one-dimensional approximations. To solve the general expression in three-dimensions, a ray-tracing algorithm must be developed.