Weather🔗

📖 Overview🔗

This service returns historical and forecast weather at one location, for the weather model you choose. The location is resolved to its ~5 km grid cell, identified by a geohash (1), and the result is one JSON row per local day, or one row per time step with hourly aggregation.

Geohash🔗
Short string that encodes a geographic grid cell. Nearby locations share the same geohash prefix.

Use it to:

Retrieve the weather history of a field, back to 1940.
Get the daily or hourly forecast for the next 5 to 15 days.
Feed weather variables into crop analytics such as growth stage, disease risk or yield models.

🗂️ Baseline data🔗

Model	Source	Coverage	Historical	Forecast
`era5`	ECMWF reanalysis	Global	From 1940 to about seven days ago — daily, hourly	—
`ifs`	ECMWF forecast model	Global	Last 15 complete days, from archived short-range forecasts — daily, hourly	About 15 days ahead — daily, hourly (3 h steps)
`icon_eu`	DWD forecast model	Europe only (latitude 29.5 to 70.5, longitude -23.5 to 62.5)	Since 2023-05-24, from archived short-range forecasts — daily	5 days ahead — hourly
`seamless`	`era5`, then `ifs` up to yesterday	Global	From 1940 to yesterday — daily, hourly	—

Historical weather comes from the ERA5 reanalysis (1) and from archived forecasts of the IFS (2) and ICON-EU (3) models. ERA5 and IFS are produced by ECMWF (4); ICON-EU by DWD (5).

Reanalysis🔗
Historical weather dataset produced by running a fixed weather model over past observations, giving a consistent record in space and time.
IFS (Integrated Forecasting System)🔗
ECMWF's global numerical weather prediction model, used for forecasts up to about 15 days ahead.
ICON-EU (ICOsahedral Nonhydrostatic model, European domain)🔗
The European regional configuration of the ICON weather model run by DWD, with forecasts up to 5 days ahead over Europe.
ECMWF (European Centre for Medium-Range Weather Forecasts)🔗
Intergovernmental organization that produces global numerical weather forecasts and climate reanalyses, including the IFS forecast model and the ERA5 reanalysis.
DWD (Deutscher Wetterdienst)🔗
The German national weather service, which runs the ICON weather models.

Seamless history

Each seamless row's model field says which model served it. An ifs day can change slightly once era5 covers it.

⚙️API🔗

The full API reference is available at weatherflow.aws.geosys.com/docs.

Endpoint	Description
`GET /models`	Lists every model, with the datasets and aggregations it serves.
`GET /models/{model}`	Describes one model: for each dataset and aggregation, the time step, the columns of a row with their units, and the time convention.
`GET /models/{model}/historical`	Past weather at one location.
`GET /models/{model}/forecast`	Forecast at one location, from the latest model run.

Base URL: https://weatherflow.aws.geosys.com/prod
Authentication: API key passed in the X-API-Key header on every request.

Example request🔗

Daily ERA5 history for a field in Ames, Iowa, for the first week of June 2025:

curl -H "X-API-Key: <your key>" "https://weatherflow.aws.geosys.com/prod/models/era5/historical?latitude=42.03&longitude=-93.63&start=2025-06-01&end=2025-06-07"

import requests

r = requests.get(
    "https://weatherflow.aws.geosys.com/prod/models/era5/historical",
    headers={"X-API-Key": "<your key>"},
    params={
        "latitude": 42.03,
        "longitude": -93.63,
        "start": "2025-06-01",
        "end": "2025-06-07",
    },
)
if r.status_code == 204:
    rows = []  # no row in the range, for example era5 days not yet published
else:
    r.raise_for_status()
    rows = r.json()

Response (first row):

[
  {
    "date": "2025-06-01",
    "geohash": "9zmrj",
    "aggregation_timezone": "America/Chicago",
    "temperature_2m_mean": 21.3354,
    "temperature_2m_max": 27.9,
    "temperature_2m_min": 14.0,
    "precipitation_sum": 0.0,
    "precipitation_hours": 0.0,
    "snowfall_sum": 0.0,
    "wind_speed_10m_max": 4.3709,
    "wind_gusts_10m_max": 6.4,
    "wind_direction_10m_dominant": 140.0,
    "shortwave_radiation_sum": 29.988,
    "swvl1": 0.2452,
    "swvl2": 0.2691,
    "swvl3": 0.2383,
    "swvl4": 0.3086,
    "model": "era5"
  }
]

⚙️ Parameters & Variables🔗

Input Parameters🔗

/models/{model}/historical and /models/{model}/forecast take the same parameters.

Parameter	Variable Name	Description	Type
model	`model` (path)	Weather model: `era5`, `ifs`, `icon_eu` or `seamless` for historical; `ifs` or `icon_eu` for forecast	string
latitude	`latitude`	Latitude of the location, decimal degrees (WGS84), -90 to 90	number
longitude	`longitude`	Longitude of the location, decimal degrees (WGS84), -180 to 180	number
start	`start`	Start date, `YYYY-MM-DD`, local date at the location	date
end	`end`	End date, `YYYY-MM-DD`, local date at the location (inclusive)	date
aggregation	`aggregation` (optional)	`daily`: one row per local day. `hourly`: one row per time step (3 h for the IFS forecast). Default is `daily`	string

Available combinations

Model	Dataset	Aggregation	Time step
`era5`	historical	daily, hourly	24 h, 1 h
`ifs`	historical	daily, hourly	24 h, 1 h
`ifs`	forecast	daily, hourly	24 h, 3 h
`icon_eu`	historical	daily	24 h
`icon_eu`	forecast	hourly	1 h
`seamless`	historical	daily, hourly	24 h, 1 h

Request limits

Historical: end is today at the location at the latest. One call covers at most 366 days; hourly era5 and seamless at most 10 local days. Split longer ranges into several calls.
ifs history serves the last 15 complete days only. icon_eu history starts on 2023-05-24.
Forecast: start is today at the location at the earliest.
era5 days are published with a delay of about seven days. A range that lies entirely in that delay returns 204. A range that overlaps it returns 200 with only the published days, with no error and no padding: check the last date returned. Use seamless to fill the gap with ifs.

Output Variables🔗

The response body is the list of rows, in time order, with no envelope. Each row carries these columns:

Parameter	Variable Name	Description	Type
Date	`date`	Local date (daily) or local time with UTC offset (hourly), for example `2026-09-18T15:00:00-07:00`	string
Geohash	`geohash`	Identifier of the ~5 km grid cell the location resolves to	string
Model	`model`	Model that served the row (`era5` or `ifs` for `seamless`)	string
Aggregation Timezone	`aggregation_timezone`	Time zone used to define local days, for example `America/Chicago`. Over the sea, a fixed-offset zone with the POSIX sign: `Etc/GMT+1` is UTC-1	string

The weather variables depend on the model, dataset and aggregation. In the Available in column, ERA5 stands for era5 and seamless history, IFS hist. for ifs history, IFS fcst. for ifs forecast and ICON-EU for icon_eu.

Units vary by model

The ifs forecast gives wind speed and gusts in km/h and snowfall in cm of fresh snow, where the historical datasets give m/s and mm of water equivalent. Always check the units returned by GET /models/{model}.

Daily variables🔗

Parameter	Variable Name	Description	Unit	Available in
Mean Temperature	`temperature_2m_mean`	Mean air temperature at 2 m	°C	All
Maximum Temperature	`temperature_2m_max`	Maximum air temperature at 2 m	°C	All
Minimum Temperature	`temperature_2m_min`	Minimum air temperature at 2 m	°C	All
Precipitation	`precipitation_sum`	Total precipitation (rain, showers and snow)	mm	All
Precipitation Hours	`precipitation_hours`	Number of hours with precipitation. Always `null` on `ifs` history rows, including the `ifs` rows of `seamless`: IFS 3-hour steps are not comparable to the ERA5 hourly count. In 3-hour steps for the `ifs` forecast	h	All
Snowfall	`snowfall_sum`	Total snowfall	mm water equivalent; cm fresh snow (IFS fcst.)	ERA5, IFS hist., IFS fcst.
Maximum Wind Speed	`wind_speed_10m_max`	Maximum wind speed at 10 m	m/s; km/h (IFS fcst.)	All
Maximum Wind Gusts	`wind_gusts_10m_max`	Maximum wind gusts at 10 m	m/s; km/h (IFS fcst.)	All
Dominant Wind Direction	`wind_direction_10m_dominant`	Dominant wind direction at 10 m	degrees (meteorological)	All
Solar Radiation	`shortwave_radiation_sum`	Total incoming shortwave solar radiation	MJ/m²	All
Soil Water, Layer 1	`swvl1`	Volumetric soil water, 0–7 cm	m³/m³	ERA5, IFS hist.
Soil Water, Layer 2	`swvl2`	Volumetric soil water, 7–28 cm	m³/m³	ERA5, IFS hist.
Soil Water, Layer 3	`swvl3`	Volumetric soil water, 28–100 cm	m³/m³	ERA5, IFS hist.
Soil Water, Layer 4	`swvl4`	Volumetric soil water, 100–289 cm	m³/m³	ERA5, IFS hist.
Relative Humidity	`relative_humidity_2m_mean`, `_max`, `_min`	Mean, maximum and minimum relative humidity at 2 m	%	IFS fcst., ICON-EU
Cloud Cover	`cloud_cover_mean`, `_max`, `_min`	Mean, maximum and minimum total cloud cover	%	IFS fcst., ICON-EU
Wind Speed	`wind_speed_10m_mean`, `_min`	Mean and minimum wind speed at 10 m	km/h	IFS fcst.
Wind Gusts	`wind_gusts_10m_mean`, `_min`	Mean and minimum wind gusts at 10 m	km/h	IFS fcst.
Rain	`rain_sum`	Total large-scale rain	mm	ICON-EU
Showers	`showers_sum`	Total convective precipitation	mm	ICON-EU
Snowfall	`snowfall_water_equivalent_sum`	Total snowfall	mm water equivalent	ICON-EU
Snow Depth	`snow_depth_max`	Maximum snow depth	m	ICON-EU
Direct Radiation	`direct_radiation_sum`	Total direct solar radiation	MJ/m²	ICON-EU
Diffuse Radiation	`diffuse_radiation_sum`	Total diffuse solar radiation	MJ/m²	ICON-EU
Soil Temperature	`soil_temperature_0cm_mean`	Mean soil surface temperature	°C	ICON-EU
Soil Moisture	`soil_moisture_0_to_1cm_mean`	Mean volumetric soil moisture, 0–1 cm	m³/m³	ICON-EU
Visibility	`visibility_min`	Minimum visibility	m	ICON-EU
Weather Code	`weather_code_max`	Most severe weather condition of the day	WMO code	ICON-EU

Soil water columns swvl1 to swvl4 are expressed as volumetric soil water (1); weather codes follow the WMO weather code table (2).

Volumetric Soil Water🔗
Volume of water per volume of soil, in m³/m³. For example, 0.30 means water fills 30% of the soil volume.
WMO (World Meteorological Organization) weather code🔗
Numeric code from the World Meteorological Organization describing the weather condition, from 0 (clear sky) to 99 (thunderstorm with heavy hail).

Hourly variables🔗

Parameter	Variable Name	Description	Unit	Available in
Temperature	`temperature_2m`	Air temperature at 2 m	°C	All
Dew Point	`dew_point_2m`	Dew point temperature at 2 m	°C	ERA5, IFS hist.
Relative Humidity	`relative_humidity_2m`	Relative humidity at 2 m	%	All
Precipitation	`precipitation`	Total precipitation over the time step	mm	All
Snowfall	`snowfall_water_equivalent`	Snowfall over the time step	mm water equivalent	ERA5, IFS hist., ICON-EU
Snowfall	`snowfall`	Snowfall over the time step	cm fresh snow	IFS fcst.
Snow Depth	`snow_depth`	Snow depth	m	IFS fcst., ICON-EU
Wind Speed	`wind_speed_10m`	Wind speed at 10 m	m/s; km/h (IFS fcst.)	ERA5, IFS hist., IFS fcst.
Wind Direction	`wind_direction_10m`	Wind direction at 10 m	degrees (meteorological)	ERA5, IFS hist., IFS fcst.
Wind Components	`wind_u_component_10m`, `wind_v_component_10m`	Eastward and northward wind components at 10 m	m/s	ICON-EU
Wind Gusts	`wind_gusts_10m`	Wind gusts at 10 m	m/s; km/h (IFS fcst.)	All
Solar Radiation	`shortwave_radiation`	Incoming shortwave solar radiation	W/m²	ERA5, IFS hist., IFS fcst.
Direct Radiation	`direct_radiation`	Direct solar radiation	W/m²	ICON-EU
Diffuse Radiation	`diffuse_radiation`	Diffuse solar radiation	W/m²	ICON-EU
Cloud Cover	`cloud_cover`	Total cloud cover	%	IFS fcst., ICON-EU
Cloud Cover by Level	`cloud_cover_low`, `_mid`, `_high`	Low, mid and high cloud cover	%	IFS fcst.
Rain	`rain`	Large-scale rain over the time step	mm	ICON-EU
Showers	`showers`	Convective precipitation over the time step	mm	ICON-EU
Soil Temperature	`soil_temperature_0cm`	Soil surface temperature	°C	ICON-EU
Soil Moisture	`soil_moisture_0_to_1cm`	Volumetric soil moisture, 0–1 cm	m³/m³	ICON-EU
Visibility	`visibility`	Visibility	m	ICON-EU
Weather Code	`weather_code`	Weather condition	WMO code	ICON-EU

Time convention🔗

Days are local at the location: start and end are local dates, and aggregation_timezone names the zone.
Hourly rows are local time with an offset (2026-09-18T15:00:00-07:00). Sort on the instant, not the string: the offset changes on daylight-saving nights, when a day has 23 or 25 hours.
The 3-hourly grid of ifs is anchored on UTC, so the first step of a local day is rarely 00:00 local (00:30 in Tehran, 02:00 in Phoenix).
Days are served only once they are complete.

Errors🔗

Status	Meaning
`200`	Rows returned.
`204`	No row in the requested range, for example `era5` days not yet published. The response has no body: check the status before parsing JSON.
`400`	The request is valid but outside what the model serves: range too long, location or dates not covered, dates in the future or the past. Also returned for a valid aggregation the model does not serve, for example `icon_eu does not serve historical/hourly`. The `detail` field says what to change, for example `start must be <= end`.
`401`	API key missing or invalid.
`422`	Malformed request: out-of-range value, unknown model or aggregation value, or a forecast asked from a history-only model.
`500`	Internal error while reading the data. Contact support if it persists.
`503`	The archive cannot be read for a moment. Retry after the number of seconds given in the `Retry-After` header.
`504`	The request timed out (29 s). Can happen on the first long request at a new location; retry.

📊 Performance and Accuracy🔗

Spatial resolution: location resolved to a ~5 km cell; native model grids: ERA5 and IFS 0.25° (~28 km), ICON-EU ~7 km
Coverage: Global (era5, ifs, seamless); Europe only (icon_eu)
Average Generation Time: typically 1 to 2 s for a short range; up to about 20 s on the first request for a full year at a new location

💼 Use Case and Product Integration🔗

This service is used in:

📚 Glossary🔗

Weather terms used on this page, such as ERA5, IFS, ICON-EU, reanalysis and geohash, are defined in the Glossary. Hover over an acronym to see its definition.

Ready to learn more? Our team is available to discuss how EarthDaily's platform can meet your specific needs. Contact us

Weather🔗

📖 Overview🔗

Geohash🔗

🗂️ Baseline data🔗

Reanalysis🔗

IFS (Integrated Forecasting System)🔗

ICON-EU (ICOsahedral Nonhydrostatic model, European domain)🔗

ECMWF (European Centre for Medium-Range Weather Forecasts)🔗

DWD (Deutscher Wetterdienst)🔗