ameriflux-site-info

Extract AmeriFlux site metadata and map it to EcoSIM JSON variables. Use when a task needs site latitude, longitude, elevation, mean annual temperature, Koppen climate code, or vegetation type for an

SKILL.md

GitHub

AmeriFlux Site Info Extractor

Use When

  • You need EcoSIM site metadata for an AmeriFlux site ID such as US-Ha1.
  • You need to map flux-site attributes into ALATG, ALONG, ALTIG, ATCAG, IETYPG, or IXTYP1.
  • You are building site-level inputs before climate, atmospheric chemistry, or soil extraction.

Constraints

  • NEVER use it extract climate data.

Purpose

Automate the identification and derivation of site-specific attributes (e.g., location, vegetation) from AmeriFlux site pages and map the extracted values to EcoSIM variables. The bundled script uses cached EcoSIM site JSON first, then structured AmeriFlux page text or rendered DOM extraction when a refresh is needed.

Workflow

  1. Resolve the AmeriFlux site ID or site name.
  2. Load existing result/<SITE_ID>/<SITE_ID>_ecosim_site.json metadata when present.
  3. Extract latitude, longitude, elevation, mean annual temperature, Koppen-Geiger code, and IGBP vegetation type.
  4. Map the extracted values to EcoSIM JSON variables and write under result/<SITE_ID>/.
  5. Check units: latitude/longitude are decimal degrees, elevation is meters, and temperature is degrees Celsius.

Metadata Extraction Policy

Use deterministic sources in this order:

  1. Existing local EcoSIM site metadata JSON.
  2. AmeriFlux siteinfo HTML fetched with Python standard library URL tools, or requests when available.
  3. Rendered AmeriFlux DOM text fetched with Playwright if the static HTML is incomplete.

Always inspect the extracted values for physical and ecological plausibility before using them in EcoSIM inputs.

Location Cross-Checks

AmeriFlux pages put each value on its own line under a label (Lat, Long: then -16.4980, -56.4120). Line cleanup must never strip a leading minus sign: an earlier version did, turning southern latitudes (BR-Npw, CL-SDF), sub-zero MAT (US-Brw -12.6, US-Prr -2), and below-sea-level elevation (US-Myb -4) positive. Every extraction is now cross-checked, and the result is stored in _location_validation:

Check Evidence On failure
page_signs Each parsed latitude, longitude, elevation, and MAT appears with the same sign in the raw page text under its label error
country_bounds Coordinates fall inside the envelope of the country in the site-ID prefix (BR-, CL-, US-, ...); the message says when the opposite sign would fit error
description_region Region words in the site Description or name (South America, Pantanal, Amazon, Patagonia, Alaska, Arctic, ...) agree with the latitude warning
koppen_consistency MAT lies in a plausible range for the Koppen group (A 15 to 35, C -3 to 30, D -25 to 16, E -60 to 8 degC); tropical classes within 35 deg latitude warning
koppen_from_coordinates The page Koppen class appears in the 3x3 neighbourhood of the parsed coordinates on a gridded Beck et al. (2018) Koppen-Geiger map; when it fits the opposite-sign latitude or longitude instead, the warning says to suspect a lost minus sign warning

The Koppen map is a NetCDF with ids(lat, lon) indexing a labels list, such as the ILAMB regions/Koppen.nc 0.5-degree product (Beck et al. 2018, Sci. Data 5:180214). Pass --koppen-raster PATH, set ECOSIM_KOPPEN_RASTER, or place it at data/Koppen.nc; without it the check reports not_available. A plain mismatch (e.g. US-MMS page Cfa, map Dfa) usually reflects a class boundary or a different climatology, so it is a warning, not a coordinate error.

Errors stop the extraction without writing JSON. Cached JSON is validated too; a cached file that fails is refetched automatically. Use --skip-location-check only after confirming the coordinates by hand. Warnings are printed and kept in the JSON; review them before building forcing. The checks are offline and catch sign and gross-location errors, not small coordinate errors inside the right country.

1. Site Metadata Extraction (Flux Network)

Given an AmeriFlux site name or ID (e.g., "Blodgett Forest" or "US-Blo"), read cached metadata or parse the rendered AmeriFlux site page text, then map the extracted values to the required EcoSIM JSON variables.

JSON Variable Source Attribute Description
ALATG Site Latitude Decimal degrees north.
ALONG Site Longitude Decimal degrees east.
ALTIG Elevation Meters above sea level.
ATCAG MAT Mean Annual Temperature (°C).
IETYPG Climate Class Koppen-Geiger climate zone code.
IXTYP1 IGBP Type Dominant vegetation type (mapped to plant litter flags).

Logic for Vegetation Mapping:

  • If IGBP is ENF (Evergreen Needleleaf) → Set IXTYP1 to 9 or 11 (Coniferous).
  • If IGBP is DBF (Deciduous Broadleaf) → Set IXTYP1 to 8 or 10 (Deciduous).

Koppen climate classification mapping: Using the koppenDict mapping, convert the site's Koppen-Geiger code (e.g., "Csa") to the corresponding integer code for IETYPG. This will allow the model to apply appropriate climate-specific parameters during simulations.

koppenDict = { "Af": 11, "Am": 12, "As": 13, "Aw": 14, "BWk": 21, "BWh": 22, "BSk": 26, "BSh": 27, "Cfa": 31, "Cfb": 32, "Cfc": 33, "Csa": 34, "Csb": 35, "Csc": 36, "Cwa": 37, "Cwb": 38, "Cwc": 39, "Dfa": 41, "Dfb": 42, "Dfc": 43, "Dfd": 44, "Dsa": 45, "Dsb": 46, "Dsc": 47, "Dsd": 48, "Dwa": 49, "Dwb": 50, "Dwc": 51, "Dwd": 52, "ET": 61, "EF": 62 }

2. Implementation & Execution

Prerequisites

  • Python 3.8+
  • requests: Optional HTTP fetch helper; the script also has a standard-library HTTP fallback.
  • Playwright: Optional fallback for rendered DOM extraction when static HTML is incomplete.

Setup for the Bundled Local Script

pip install playwright requests
playwright install chromium

No local vision service is required.

Usage

To execute the skill, run the following command from the project root. The resulting JSON will be saved under ./result/<SITE_ID>/ by default:

python .agents/skills/ameriflux-site-info/extract_ameriflux_site_data.py <SITE_ID>

Example:

python .agents/skills/ameriflux-site-info/extract_ameriflux_site_data.py US-Ha1

To bypass cached metadata and refresh from AmeriFlux:

python .agents/skills/ameriflux-site-info/extract_ameriflux_site_data.py US-Ha1 --force-refresh

Output

The script creates a JSON file named result/<SITE_ID>/<site_name>_ecosim_site.json with the following structure:

{
  "site_name": "US-Ha1",
  "ALATG": 40.0,      # Latitude (decimal degrees north)
  "ALONG": -120.0,    # Longitude (decimal degrees east)
  "ALTIG": 1000.0,    # Elevation (meters above sea level)
  "ATCAG": 10.0,      # Mean Annual Temperature (°C)
  "IETYPG": 34,       # Koppen climate zone code
  "IXTYP1": 10        # Vegetation type code
}