global-aux-climate-chemistry

Download, screen, and map non-North-America auxiliary EcoSIM climate variables from global or regional atmospheric-chemistry data sources. Use when a non-US/non-NLDAS site needs annual EcoSIM precipit

SKILL.md

GitHub

Global Auxiliary Climate Chemistry

Scope

Use this skill for grids outside the US/NADP/NLDAS workflow when EcoSIM climate forcing needs annual auxiliary variables beyond ERA5/NLDAS meteorology.

This skill covers:

  • Annual precipitation chemistry variables: PHRG, CN4RIG, CNORIG, CSORG, CCARG, CMGRG, CNARG, CKARG, CCLRG, and supported template defaults.
  • Annual non-chemistry variables: Z0G, IFLGW, and ZNOONG.
  • Source selection among CAMS, EBAS, EANET, and MERRA-2.

Do not use NADP or tDEP outside their valid US/North-America domain unless the user explicitly supplies already-processed compatible rasters and asks to force them.

Source Priority

  1. CAMS global atmospheric composition forecasts: first automated gridded choice for non-North-America wet-deposition chemistry. Use it for a small test subset before downloading target years.
  2. Regional station networks: use as preferred pH/base-cation observations or as validation/override near the site.
    • Use EANET in East and Southeast Asia.
    • Use EBAS where European or other public precipitation-chemistry stations are available.
  3. MERRA-2: fallback aerosol/deposition context only. Do not treat it as a complete precipitation-chemistry replacement without an explicit unit and species-mapping review.
  4. Defaults: when no defensible source exists, use EcoSIM defaults with provenance. Use neutral PHRG=7 only when pH cannot be derived.

Workflow

  1. Confirm site coordinates, simulation years, and existing climate forcing NetCDF.
  2. Compute simple variables without downloading chemistry data:
    • ZNOONG: compute from longitude.
    • Z0G: use the forcing wind reference height. ERA5/CAMS 10 m winds imply Z0G=10 m; if a forcing source uses another measurement height, use that value and record it.
    • IFLGW: use site/tower/vegetation metadata if available; otherwise use 0 with a note.
  3. For chemistry, run a small CAMS smoke test first:
    • Request one month or one year around the site.
    • Include precipitation and candidate wet-deposition variables.
    • Inspect returned metadata for units, accumulation period, species basis, and missing values.
    • Do not build a full converter until this smoke test passes.
  4. Download only the target years needed after the smoke test succeeds.
  5. Check for nearby station data:
    • Use EBAS/EANET observations to validate CAMS annual concentrations where possible.
    • Prefer station pH and base cations over model-derived estimates when the station is representative and close enough for the study purpose.
  6. Convert annual wet-deposition fluxes to precipitation concentrations only after confirming source units.
  7. Write a derivation JSON under result/<site_id>/ and update the EcoSIM NetCDF only after all validity checks pass.

Unit Policy

EcoSIM precipitation chemistry variables are concentrations in precipitation, not deposition fluxes.

For gridded wet-deposition sources:

concentration_g_m3 = annual_wet_deposition_kg_m2 * 1000 / annual_precipitation_m

Apply elemental mass fractions only when the source deposition is stored as compound/ion mass rather than elemental mass:

  • NH4 to CN4RIG: multiply by 14.0067 / 18.0385 = 0.7765.
  • NO3 to CNORIG: multiply by 14.0067 / 62.0049 = 0.2259.
  • SO4 to CSORG: multiply by 32.065 / 96.06 = 0.3338.
  • If the source already reports N or S mass, do not apply the ion-to-element factor.

For station concentrations, preserve the native concentration basis and convert explicitly:

  • 1 mg/L = 1 g/m3.
  • 1 ug/L = 0.001 g/m3.
  • ueq/L requires ion charge and molar mass; do not convert by a scalar shortcut.

Validity Checks

Reject or flag:

  • Non-finite values, source fill values, negative concentrations, or negative precipitation.
  • PHRG outside 0-14.
  • Annual precipitation totals too small to safely divide wet deposition into concentration.
  • CAMS or MERRA-2 variables whose metadata does not clarify whether the mass basis is compound/ion/elemental.
  • Full-year EcoSIM outputs that contain partial-year chemistry without a documented interpolation or climatology policy.

Gap policy:

  • Interpolate short annual gaps with linear interpolation and nearest-edge fill only when neighboring years are from the same data source and variable.
  • Use neutral PHRG=7 only when no valid pH source exists.
  • Do not synthesize base cations from dust or sea-salt fields without labeling the result low confidence.

Source-Specific Details

For download snippets, source access notes, and suggested source-to-variable mappings, read references/source_downloads.md only when needed.

Reporting

Always report:

  • Site lon/lat, years, and bounding box or station-distance criteria.
  • Dataset names and access method.
  • Exact source variables used and units read from metadata.
  • Conversion factors, elemental-basis assumptions, and precipitation divisor.
  • Whether values came from CAMS, EBAS, EANET, MERRA-2, station override, interpolation, or defaults.
  • Quality flags and any variables left absent/defaulted.