Predefined Networks

JAFF ships a set of ready-to-use reaction networks in the networks/ directory. They cover a range of physical regimes and serve as both already verified research networks and format demonstrations.

Overview

h_photoionization

File: networks/h_photoionization/h_photo.jet

A two-reaction JAFF native network illustrating the photochemistry pipeline. It contains one photoionisation reaction with a 13.6 eV threshold and one radiative recombination reaction, making it the smallest self-contained network for testing photochemistry.

H -> H+ + E          []         PHOTO, 13.60
H+ + E -> H          []         2.63e-13*(Tgas/1e4)**(-0.7)

demos

Files: networks/demos/demo1.jet, networks/demos/demo2.jet

Demonstration files, not intended for scientific use.

• demo1.jet — A mixed-format file containing reactions written in PRIZMO, KIDA, UDFA, and KROME syntax side-by-side. Useful for checking format auto-detection and as a format cheat-sheet.
• demo2.jet — A two-reaction JAFF native snippet used in the documentation examples.

GOW

File: networks/GOW/GOW.jet

The Gong, Ostriker & Wolfire (2017) diffuse-ISM chemistry network in KIDA format. It covers hydrogen, carbon, and oxygen chemistry relevant to the cold neutral medium and was originally distributed with the Athena++ code.

Reference: Gong, Ostriker & Wolfire, ApJ 843, 38 (2017)

The GOW directory ships three files:

GOW/
├── GOW.jet      reaction network (KIDA format)
├── GOW.jfunc    custom rate and heating/cooling function definitions
└── GOW.hdf5     interpolation tables consumed by those functions

GOW.hdf5 data tables

GOW.jfunc defines several heating/cooling functions that are not closed-form — they call interpolation functions whose grids live in GOW.hdf5. The file is organised into three top-level groups, one per physical process. All datasets are stored in log-scaled form following the Omukai+2010 / Gong+2017 fits.

GOW.hdf5
├── co/                  CO rotational-line cooling — feeds cooling_CO()
│   ├── TCO       (11,)  int64    gas-temperature grid axis [K], 10 → 2000
│   ├── NeffCO    (11,)  float64  log₁₀ LVG column parameter Ñ axis
│   │                             [cm⁻²/(km s⁻¹)], 14 → 19
│   ├── L0CO      (11,)  float64  optically-thin cooling coeff L₀, 1-D over TCO
│   ├── LLTECO    (121,) float64  LTE cooling coeff L_LTE, 2-D (TCO × NeffCO)
│   ├── nhalfCO   (121,) float64  half-cooling density n₁⁄₂, 2-D (TCO × NeffCO)
│   └── alphaCO   (121,) float64  fit exponent α (dimensionless), 2-D
│       └─ group attr  nd_order = ['TCO', 'NeffCO']   ← axis order of 2-D tables
│
├── dust/                gas–dust collisional cooling Ψ_GD — feeds cooling_dust_coll()
│   ├── lognH     (150,) float64  log₁₀ H-nucleus density [cm⁻³], 0 → 6
│   ├── logTg     (150,) float64  log₁₀ gas temperature [K], 0.5 → 4
│   └── logps     (150,) float64  log₁₀ Ψ_GD cooling-rate coeff at (lognH, logTg)
│
└── radiative_cooling/   radiative cooling vs. radiation temperature
    ├── log_Trad       (110,) float64  log₁₀ radiation temperature [K], 3.8 → 8.16
    ├── log_gamma_H_He (110,) float64  log₁₀ H+He cooling coeff over log_Trad
    └── log_gamma_Z    (110,) float64  log₁₀ metal (Z) cooling coeff over log_Trad

• co/ — the four datasets L0CO, LLTECO, nhalfCO, alphaCO are the pieces of the Neufeld-style CO cooling fit, looked up by L0_CO_interp1d, LLTE_CO_interp2d, nHalf_CO_interp2d, and alpha_CO_interp2d against the TCO (temperature) and NeffCO (LVG column) axes. The 121-element tables are the 11 × 11 grids flattened in the order given by the nd_order attribute.
• dust/ — (lognH, logTg) form a 15 × 10 grid; logps is the tabulated Ψ_GD value at each node, used by PsiGD_coll_interp2d to set the dust temperature implicitly in the gas–dust collisional cooling term.
• radiative_cooling/ — radiative cooling coefficients for the H+He and metal (Z) components as a function of radiation temperature, indexed by log_Trad.

COthin

File: networks/COthin/react_COthin.jet

A CO-chemistry network in KROME format based on Glover+2010 and additional literature sources. It is designed for optically thin environments and is well-suited for simulations of diffuse molecular clouds where CO formation and destruction rates are needed without full grain-surface chemistry.

Reference: Glover et al., MNRAS 404, 2 (2010)

popsicle_semenov

File: networks/popsicle_semenov/react_popsicle_semenov.jet

A KROME-format network built for the POPSICLE simulation code (Sharda & Menon 2024). It combines a primordial chemistry network (including deuterium species) with metal-line cooling reactions from Omukai (2000), Omukai+2005, Glover & Jappsen (2007), and Glover+2010. Dust opacity follows Semenov+2003. Cosmic-ray and photochemistry reactions are not included.

Reference: Sharda & Menon (2024)

uclchem_small_gas

File: networks/uclchem_small_gas/uclchem_small_gas_network.jet

A small gas-phase development network generated by UCLCHEM v3.5.1. It includes a minimal set of ice-surface species alongside UMIST Rate22 gas-phase reactions. The network uses the UCLCHEM species notation: @ for bulk ice, # for surface ice, and no prefix for gas phase.

Generated by: Gijs Vermariën using UCLCHEM v3.5.1

kida_uva_2024

File: networks/kida_uva_2024/gas_reactions_kida.uva.2024.jet

The complete KIDA UVA 2024 gas-phase reaction database in KIDA format. It covers a broad range of ion-neutral, neutral-neutral, photodissociation, and cosmic-ray reactions and is one of the primary reference databases in astrochemical modelling.

Reference: Wakelam et al., A&A 689, A63 (2024)

Source: kida.astrochem-tools.org

rate22_final

File: networks/rate22_final/rate22_final.rates.jet

The complete UMIST Database for Astrochemistry Rate22 release in UDFA format. It is the largest network included and covers gas-phase reactions for over 500 species across a wide temperature range.

Reference: Millar et al., A&A 682, A109 (2024)

Source: umistdatabase.uk