Column Schema

Every adapter and feature function in climagrid produces DataFrames that conform to a shared column contract. Adding a new column to any output requires defining it in climagrid.schema first.

The schema is machine-readable via climagrid.schema_summary() and exportable as JSON via climagrid.outputs.to_json_schema().

---

Index columns

Always present, never null. Uniquely identify each observation.

Column	Type	Units	Description
asset_id	str	n/a	Utility asset identifier from AssetRegistry
timestamp	datetime64[ns, UTC]	n/a	UTC timestamp of the observation hour
lat	float64	degrees	Asset latitude (WGS-84)
lon	float64	degrees	Asset longitude (WGS-84)

---

NOAA HRRR columns

3 km CONUS hourly NWP analysis fields. Require pip install "climagrid[noaa-nwp]".

Column	Type	Units	Description
hrrr_temperature_2m	float64	°C	2-metre air temperature
hrrr_wind_speed_10m	float64	m/s	10-metre wind speed (magnitude)
hrrr_wind_direction_10m	float64	degrees	10-metre wind direction (meteorological convention)
hrrr_relative_humidity_2m	float64	%	2-metre relative humidity
hrrr_precipitation_rate	float64	mm/hr	Hourly accumulated precipitation
hrrr_solar_irradiance_ghi	float64	W/m²	Downward short-wave radiation at surface
hrrr_snow_depth	float64	m	Snow depth at surface

---

NASA POWER columns

MERRA-2-based hourly surface meteorology. Global coverage, no API key.

Column	Type	Units	Description
nasa_temperature_2m	float64	°C	2-metre air temperature (MERRA-2 based)
nasa_wind_speed_10m	float64	m/s	10-metre wind speed
nasa_solar_irradiance_ghi	float64	W/m²	Global horizontal irradiance
nasa_relative_humidity_2m	float64	%	Relative humidity at 2 m
nasa_precipitation	float64	mm	Hourly precipitation

---

NOAA NCEI columns

Surface station observations from nearest GHCN station.

Column	Type	Units	Description
ncei_temperature_max	float64	°C	Daily maximum temperature
ncei_temperature_min	float64	°C	Daily minimum temperature
ncei_wind_speed	float64	m/s	Average daily wind speed
ncei_precipitation_daily	float64	mm	Daily precipitation
ncei_relative_humidity	float64	%	Average daily relative humidity

---

USDA NRCS columns

SCAN / SNOTEL soil and snowpack sensors.

Column	Type	Units	Description
nrcs_soil_moisture_pct	float64	%	Volumetric soil moisture (nearest SCAN station)
nrcs_soil_temperature	float64	°C	Soil temperature at 2-inch depth
nrcs_snow_water_equivalent	float64	mm	Snow water equivalent (nearest SNOTEL)
nrcs_station_distance_km	float64	km	Distance from asset to nearest NRCS station

---

USFS WFIGS columns

Active wildfire perimeter proximity.

Column	Type	Units	Description
wfigs_nearest_fire_km	float64	km	Distance to nearest active fire perimeter edge
wfigs_fire_active	bool	n/a	True if any active fire within 50 km
wfigs_fire_area_ha	float64	ha	Area of nearest active fire in hectares

---

Feature (stress index) columns

Computed by climagrid.features.*. Values are dimensionless indices or physical counts suitable for direct use as ML model inputs.

Column	Type	Units	Standard	Description
feat_thermal_aging_factor	float64	per-unit	IEEE C57.91	Arrhenius FAA relative to 110°C reference. Values >1 indicate accelerated aging.
feat_heat_hours_above_35c	float64	hours	n/a	Cumulative hours with temperature >35°C in rolling 168-hour window
feat_freeze_thaw_cycles	float64	count	n/a	Freeze-thaw transition count over the rolling window
feat_ice_loading_risk	float64	0-1	ASCE 7-22 (simplified)	Normalized ice accretion risk (temperature × precipitation × wind composite)
feat_soil_saturation_index	float64	0-1	n/a	Normalized soil saturation proxy (0=dry, 1=saturated)
feat_wildfire_proximity	float64	0-1	n/a	Normalized fire proximity score (0=no risk, 1=adjacent to active fire)
feat_conductor_sag_index	float64	0-1	IEEE 738 (simplified)	Normalized thermal sag index (0=no sag risk, 1=at design maximum)

---

Programmatic access

import climagrid

# Human-readable summary
print(climagrid.schema_summary())

# Machine-readable JSON (useful for SCADA integration)
from climagrid.outputs import to_json_schema
schema = to_json_schema("schema.json")