`atmospheric_profile` block

Vertical profile of the model atmosphere: temperature, pressure, specific humidity, and (optionally) a level-reduction request. Profile direction is TOA → BOA (top of atmosphere first); GEOSChem-style data that arrives as BOA → TOA is flipped at parse time by geoschem_to_dict, the reader behind GeosChemSource.

Required fields

T — Vector{Real}. Temperature [K] at full levels (layer centers). Length = number of layers.
p — Vector{Real}. Pressure [hPa] at half levels (layer boundaries). Length = number of layers + 1. Must be monotonic in pressure (TOA = lowest p first).
profile_reduction — Union{Integer, Nothing}. When an integer N > 0, model construction reframes the atmosphere onto at most N layers, interpolating intensive profiles and recomputing integrated layer fields. null (YAML) or -1 keeps the full layering; TOML has no null, so use -1 there. Observer interfaces requested through geometry.obs_alt are always retained. If K distinct interior observer heights require more interfaces than the requested reduction permits, the effective count is raised to K + 1 layers and a warning reports the adjustment.

Optional fields

q — Vector{Real}. Specific humidity (water-vapor mass fraction, kg kg⁻¹). It may contain N layer-centred values (the preferred form) or N+1 values on the pressure interfaces; interface values are interpolated to layer centers in log pressure. Default: zeros (dry atmosphere). Used by the absorption pipeline to derive the H₂O VMR automatically; do not also list H₂O in the absorption molecule lists. A legacy H₂O key in the vmr dictionary is ignored.

Examples

Single-layer Rayleigh scene (Natraj test fixtures)

yaml

atmospheric_profile:
  T: [231.62]
  p: [0.14, 0.22]            # 1-layer column, ~0.08 hPa thick
  profile_reduction: null

Multi-layer with profile reduction

yaml

atmospheric_profile:
  T: [220.0, 230.0, 250.0, 270.0, 280.0, 290.0]
  p: [0.01, 1.0, 10.0, 100.0, 500.0, 800.0, 1013.25]
  q: [1e-6, 1e-5, 1e-4, 1e-3, 5e-3, 1e-2]
  profile_reduction: 5       # collapse 6 layers → 5

Reading from external sources

For real GEOSChem outputs, use the high-level reader rather than YAML:

julia

using vSmartMOM
src = vSmartMOM.GeosChemSource("/path/to/geoschem.nc4", 10, 20, 1) # x, y, face indices
config = vSmartMOM.geoschem_to_dict(src)
ap = config["atmospheric_profile"]
profile = vSmartMOM.read_atmos_profile_dict(Dict(
    "T" => ap["T"], "p_half" => ap["p"], "q" => ap["q"]))

geoschem_to_dict returns a complete configuration dictionary. read_atmos_profile_dict converts a standalone profile mapping to an AtmosphericProfile. That mapping names the pressure interfaces p_half (or gives hybrid ak, bk and p_surf), whereas the scene block above calls them p, hence the renaming. Use read_parameters(src) when building a complete RT scene.

atmospheric_profile block ​

Required fields ​

Optional fields ​

Examples ​

Single-layer Rayleigh scene (Natraj test fixtures) ​

Multi-layer with profile reduction ​

Reading from external sources ​

See also ​