[Paper Review] The 2013 Release of Cloudy

G. J. Ferland, R. L. Porter+9 more|arXiv (Cornell University)|Feb 18, 2013

Atomic and Molecular PhysicsPhysics and Astronomy7 references443 citations

TL;DR

This paper presents the 2013 release of Cloudy, a plasma simulation code that models ionization, chemical, and thermal states in astrophysical environments under external radiation or heating. It advances modeling of dusty molecular regions, improves ionization/chemistry solvers, and integrates external atomic and molecular databases, enabling accurate prediction of emission and absorption spectra across extreme densities and temperatures.

ABSTRACT

This is a summary of the 2013 release of the plasma simulation code Cloudy. Cloudy models the ionization, chemical, and thermal state of material that may be exposed to an external radiation field or other source of heating, and predicts observables such as emission and absorption spectra. It works in terms of elementary processes, so is not limited to any particular temperature or density regime. This paper summarizes advances made since the last major review in 1998. Much of the recent development has emphasized dusty molecular environments, improvements to the ionization / chemistry solvers, and how atomic and molecular data are used. We present two types of simulations to demonstrate the capability of the code. We consider a molecular cloud irradiated by an X-ray source such as an Active Nucleus and show how treating EUV recombination lines and the full SED affects the observed spectrum. A second example illustrates the very wide range of particle and radiation density that can be considered.

Motivation & Objective

• To summarize major advancements in the Cloudy plasma simulation code since the 1998 review (F98), focusing on new physical capabilities and numerical improvements.
• To extend Cloudy's applicability to dusty molecular environments, which were not fully addressed in earlier versions.
• To improve the accuracy of ionization and chemical structure solvers by integrating modern atomic and molecular data from external databases.
• To enable robust, self-consistent simulations across extreme physical conditions—from ionized plasmas to cold molecular clouds—by enhancing numerical methods and parallelization.
• To support high-fidelity spectral predictions by incorporating full spectral energy distributions (SEDs), EUV recombination lines, and improved radiation transport.

Proposed method

• Cloudy models non-equilibrium plasma conditions using first-principles microphysical processes, including ionization, recombination, collisional excitation, and radiative transitions.
• The code self-consistently solves coupled equations for ionization structure, level populations, molecular abundances, and thermal balance under external radiation or heating sources.
• External databases (Chianti and Stout) are parsed at initialization to provide up-to-date atomic and molecular data, including collisional and radiative processes.
• The code uses iterative solvers to handle interdependencies between radiation field and ionization structure, ensuring self-consistency in complex environments.
• Parallelization via MPI enables large-scale, embarrassingly parallel simulations across distributed memory systems, improving performance for high-resolution models.
• Grain physics is enhanced with temperature- and charge-dependent opacities, surface reactions, and spinning dust emission models, informed by infrared space mission data.

Experimental results

Research questions

• RQ1 How does the 2013 release of Cloudy improve the simulation of dusty molecular regions compared to previous versions?
• RQ2 To what extent do full spectral energy distributions (SEDs) and EUV recombination lines affect predicted emission spectra in X-ray irradiated molecular clouds?
• RQ3 How does the integration of external atomic and molecular databases (Chianti and Stout) improve the accuracy of ionization and chemical structure calculations?
• RQ4 What are the limitations of the current iso-sequence model in achieving local thermodynamic equilibrium (LTE) at high densities, and how can they be addressed?
• RQ5 How does the code’s parallelization strategy via MPI scale across modern multi-core and distributed systems, and what are the performance bottlenecks?

Key findings

• The 2013 release enables accurate modeling of X-ray irradiated molecular clouds, showing that including EUV recombination lines and full SEDs significantly alters the predicted emission spectrum.
• Cloudy now supports a wide dynamic range in particle and radiation density, from $10^{-10}$ cm⁻³ to $10^{15}$ cm⁻³, and temperatures from the CMB to $10^{10}$ K.
• The integration of external databases (Chianti and Stout) allows for more accurate collisional and radiative processes, with Stout including all collisions and enabling better high-density behavior.
• Despite a 28-year increase in computing power, the benchmark pn_paris test case still takes ~1 minute to compute, now with far greater physical fidelity and emission model accuracy.
• Grain physics has been enhanced with temperature- and charge-dependent opacities, especially for PAHs, and spinning dust emission is being developed for improved radio emission modeling.
• The code’s iterative solvers handle complex interdependencies between radiation field and ionization structure robustly, though cache efficiency and memory bandwidth remain performance bottlenecks on modern CPUs.