Overview

Daily surface weather for agriculture, 1979 to present, on a global 0.1° grid. AgERA5 aggregates the hourly ECMWF ERA5 reanalysis to local-time days and corrects it to the finer topography, land use and coastlines of ECMWF’s 0.1° HRES model. Its 27 variables (air and dew point temperature, precipitation, solar radiation, humidity, vapour pressure, wind, cloud cover, snow, FAO-56 reference evapotranspiration) match the inputs of most crop and agro-ecological models. The primary access is ECMWF’s two analysis-ready Zarr stores, one chunked for time series at points and one for maps of a region; a Google Earth Engine copy and the CDS API are alternatives.

#AgERA5 #ERA5 #reanalysis #daily weather #agrometeorology #crop modelling #evapotranspiration #analysis-ready #zarr

Important note

Each day runs midnight to midnight local time at the grid cell, not UTC. Temperatures are kelvin and solar radiation is J m-2 d-1; convert as needed. The time axis trails real time by about a week. Values are reanalysis, not station observations: expect bias at single sites, especially for precipitation in mountains.

Quick start

Python

"""Read AgERA5 daily values for a few points from the ECMWF ARCO Zarr store.

Needs a free Copernicus CDS account and its Personal Access Token (see AGENTS.md):
  pip install xarray zarr dask fsspec aiohttp
  CDSAPI_KEY=<token> python example.py
"""

import os
from pathlib import Path

import xarray as xr

# "geoChunked" = small spatial tiles, long time runs: cheap for long time series at a few places.
# Swap in .../cadl-arco-time-001/.../timeChunked.zarr (one day per chunk) for maps on a few days.
URL = "https://arco.datastores.ecmwf.int/cadl-arco-geo-001/arco/sis_agrometeorological_indicators/all/geoChunked.zarr"


def cds_key() -> str:
    """CDSAPI_KEY env var, else the `key:` line of ~/.cdsapirc."""
    if key := os.environ.get("CDSAPI_KEY"):
        return key
    rc = Path.home() / ".cdsapirc"
    if rc.exists():
        for line in rc.read_text().splitlines():
            if line.startswith("key:"):
                return line.split(":", 1)[1].strip()
    raise SystemExit("No CDS key: set CDSAPI_KEY or create ~/.cdsapirc")


ds = xr.open_zarr(
    URL,
    consolidated=True,
    chunks={},
    storage_options={"headers": {"Authorization": f"Bearer {cds_key()}"}},
)

# Pick variables and dates first, then all points in one vectorised sel.
variables = ["Temperature_Air_2m_Max_24h", "Temperature_Air_2m_Min_24h", "Precipitation_Flux"]
lats = [-1.29, 9.03]  # Nairobi, Addis Ababa
lons = [36.82, 38.74]
points = dict(latitude=xr.DataArray(lats, dims="pt"), longitude=xr.DataArray(lons, dims="pt"))

sub = ds[variables].sel(time=slice("2023-01-01", "2023-12-31")).sel(**points, method="nearest").compute()

df = sub.to_dataframe().reset_index()
df["tmax_c"] = df["Temperature_Air_2m_Max_24h"] - 273.15  # kelvin to celsius
print(df.head())

Provided by the dataset authors: Python example

Full access details and endpoints

Intended uses

daily weather inputs for crop, livestock and agro-ecological models
weather time series for trial sites, farms and other point locations
growing-season summaries such as rainfall totals, heat days and evapotranspiration

Not recommended for

hourly or sub-daily analysis

Why: AgERA5 is a daily aggregate.

Use instead: ERA5 hourly data on single levels (DOI 10.24381/cds.adbb2d47).

station-accurate values at one site

Why: reanalysis on an 11 km grid, so local effects and extremes are smoothed.

Use instead: nearby weather station records, with AgERA5 to fill gaps.

Access & use

Files, endpoints, and worked examples for each distribution.

Access conditions

Free, but every route needs an account. The Zarr stores and the CDS API take a Copernicus Climate Data Store (CDS) Personal Access Token: register at https://cds.climate.copernicus.eu, copy the token from your profile page, then open the dataset page, go to the Download tab and accept the dataset licence at the bottom. Until the licence is accepted the stores answer 401 or 403 even with a valid token. Send the token as an HTTP header "Authorization: Bearer <token>"; example.py beside this record is a working call. The Earth Engine asset needs a Google Earth Engine account.

Data

zarr-geo-chunked

Primary access. Zarr store chunked 2112 days x 16 lat x 32 lon (about 5.8 years x 1.6° x 3.2° per chunk), so long time series at a few points or a small area read few chunks. Slow for a whole region on one day; use the time-chunked store for that.

Format: Zarr — best for cloud-native analysis, reading just the slices you need from the full cube

Variables: Temperature_Air_2m_Max_24h, Temperature_Air_2m_Min_24h, Temperature_Air_2m_Mean_24h, Temperature_Air_2m_Max_Day_Time, Temperature_Air_2m_Mean_Day_Time, Temperature_Air_2m_Min_Night_Time +21 more

Dimensions: cell × time

Data dictionary →

Access the data

https://arco.datastores.ecmwf.int/cadl-arco-geo-001/arco/sis_agrometeorological_indicators/all/geoChunked.zarr

zarr-time-chunked

Same data chunked 1 day x 1024 lat x 1024 lon, so a map of a region or the globe for a few days reads few chunks. Slow for long time series; use the geo-chunked store for that.

Format: Zarr — best for cloud-native analysis, reading just the slices you need from the full cube

Dimensions: cell × time

Data dictionary →

Access the data

https://arco.datastores.ecmwf.int/cadl-arco-time-001/arco/sis_agrometeorological_indicators/all/timeChunked.zarr

gee

Google Earth Engine image collection projects/climate-engine-pro/assets/ce-ag-era5-v2/daily, curated by Climate Engine. One image per day, same band names and units as the Zarr stores, but 21 of the 27 variables.

Dimensions: cell × time

Data dictionary →

Access the data

https://earthengine.googleapis.com/v1/projects/climate-engine-pro/assets/ce-ag-era5-v2/daily

Other access methods

Earth Engine:ee://projects/climate-engine-pro/assets/ce-ag-era5-v2/daily

cds-api

Backup route. The CDS retrieve API returns NetCDF per variable, statistic and period on request; slower than the Zarr stores and queued, but needs only the cdsapi package.

Format: application/x-netcdf

Dimensions: cell × time

Data dictionary →

Access the data

https://cds.climate.copernicus.eu/api/retrieve/v1/processes/sis-agrometeorological-indicators

Additional assets

exampleexample

Reads three variables for two points and one year from the geo-chunked store with the CDS token.

Format: text/x-python

Access the dataOpen

https://raw.githubusercontent.com/CGIAR-Climate-Data-Hub/cdh-catalog/main/records/agera5/example.py

View example

Python

"""Read AgERA5 daily values for a few points from the ECMWF ARCO Zarr store.

Needs a free Copernicus CDS account and its Personal Access Token (see AGENTS.md):
  pip install xarray zarr dask fsspec aiohttp
  CDSAPI_KEY=<token> python example.py
"""

import os
from pathlib import Path

import xarray as xr

# "geoChunked" = small spatial tiles, long time runs: cheap for long time series at a few places.
# Swap in .../cadl-arco-time-001/.../timeChunked.zarr (one day per chunk) for maps on a few days.
URL = "https://arco.datastores.ecmwf.int/cadl-arco-geo-001/arco/sis_agrometeorological_indicators/all/geoChunked.zarr"


def cds_key() -> str:
    """CDSAPI_KEY env var, else the `key:` line of ~/.cdsapirc."""
    if key := os.environ.get("CDSAPI_KEY"):
        return key
    rc = Path.home() / ".cdsapirc"
    if rc.exists():
        for line in rc.read_text().splitlines():
            if line.startswith("key:"):
                return line.split(":", 1)[1].strip()
    raise SystemExit("No CDS key: set CDSAPI_KEY or create ~/.cdsapirc")


ds = xr.open_zarr(
    URL,
    consolidated=True,
    chunks={},
    storage_options={"headers": {"Authorization": f"Bearer {cds_key()}"}},
)

# Pick variables and dates first, then all points in one vectorised sel.
variables = ["Temperature_Air_2m_Max_24h", "Temperature_Air_2m_Min_24h", "Precipitation_Flux"]
lats = [-1.29, 9.03]  # Nairobi, Addis Ababa
lons = [36.82, 38.74]
points = dict(latitude=xr.DataArray(lats, dims="pt"), longitude=xr.DataArray(lons, dims="pt"))

sub = ds[variables].sel(time=slice("2023-01-01", "2023-12-31")).sel(**points, method="nearest").compute()

df = sub.to_dataframe().reset_index()
df["tmax_c"] = df["Temperature_Air_2m_Max_24h"] - 273.15  # kelvin to celsius
print(df.head())

agent guideagents

Getting a key, choosing a store, variable units, and efficient point reads.

Format: text/markdown

Access the dataOpen

https://raw.githubusercontent.com/CGIAR-Climate-Data-Hub/cdh-catalog/main/records/agera5/AGENTS.md

Data dictionary

What each file holds. A layout describes how the data is stored in a file: its variables, the dimensions they vary over, and any keys.

File layouts

Files	Format	Variables	Dimensions	No data
zarr-geo-chunked, zarr-time-chunked, cds-api	Zarr, application/x-netcdf	Temperature_Air_2m_Max_24h, Temperature_Air_2m_Min_24h, Temperature_Air_2m_Mean_24h, Temperature_Air_2m_Max_Day_Time +23 more	cell × time	-9999
gee		Temperature_Air_2m_Max_24h, Temperature_Air_2m_Min_24h, Temperature_Air_2m_Mean_24h, Temperature_Air_2m_Max_Day_Time +17 more	cell × time	-9999

In every file

Dimensions

cellxy

0.1° (~11 km at equator). Latitude ascends from -90 to 90, longitude from -180 to 179.9.

Grid spacing: 0.1 degree

timetemporal

Day, midnight to midnight local time (00-00LT). Grows daily, about a week behind present.

Only in zarr-geo-chunked, zarr-time-chunked, cds-api

Variables

Variable	Type	Unit	Description
Temperature_Air_2m_Max_24h	float32	K	Maximum air temperature at 2 m over the local day.
Temperature_Air_2m_Min_24h	float32	K	Minimum air temperature at 2 m over the local day.
Temperature_Air_2m_Mean_24h	float32	K	Mean air temperature at 2 m over the local day.
Temperature_Air_2m_Max_Day_Time	float32	K	Maximum air temperature at 2 m between 06 and 18 local time.
Temperature_Air_2m_Mean_Day_Time	float32	K	Mean air temperature at 2 m between 06 and 18 local time.
Temperature_Air_2m_Min_Night_Time	float32	K	Minimum air temperature at 2 m between 18 and 06 local time.
Temperature_Air_2m_Mean_Night_Time	float32	K	Mean air temperature at 2 m between 18 and 06 local time.
Dew_Point_Temperature_2m_Mean_24h	float32	K	Mean dew point temperature at 2 m over the local day.
Precipitation_Flux	float32	mm d-1	Total precipitation over the local day.
Precipitation_Duration_Fraction	float32	1	Fraction of the day's hours with precipitation of at least 0.1 mm/h.
Precipitation_Rain_Duration_Fraction	float32	1	Fraction of the day's hours with rain.
Precipitation_Solid_Duration_Fraction	float32	1	Fraction of the day's hours with solid precipitation (freezing rain, snow, wet snow, rain-snow mix, ice pellets; no hail).
Solar_Radiation_Flux	float32	J m-2 d-1	Surface downward solar radiation summed over the local day.
ReferenceET_PenmanMonteith_FAO56	float32	mm d-1	Penman-Monteith reference evapotranspiration, FAO-56 method, over the local day.
Derived_Relative_Humidity_2m_Max_24h	float32	%	Maximum relative humidity at 2 m, derived from the daily minimum temperature and mean vapour pressure.
Derived_Relative_Humidity_2m_Min_24h	float32	%	Minimum relative humidity at 2 m, derived from the daily maximum temperature and mean vapour pressure.
Relative_Humidity_2m_06h	float32	%	Relative humidity at 2 m at 06 local time.
Relative_Humidity_2m_09h	float32	%	Relative humidity at 2 m at 09 local time.
Relative_Humidity_2m_12h	float32	%	Relative humidity at 2 m at 12 local time.
Relative_Humidity_2m_15h	float32	%	Relative humidity at 2 m at 15 local time.
Relative_Humidity_2m_18h	float32	%	Relative humidity at 2 m at 18 local time.
Vapour_Pressure_Mean_24h	float32	hPa	Mean vapour pressure over the local day.
Vapour_Pressure_Deficit_at_Maximum_Temperature	float32	hPa	Vapour pressure deficit at the daily maximum temperature: saturated vapour pressure at the maximum temperature minus the daily mean vapour pressure.
Wind_Speed_10m_Mean_24h	float32	m s-1	Mean wind speed at 10 m over the local day.
Cloud_Cover_Mean_24h	float32	1	Mean total cloud cover over the local day, 0 to 1.
Snow_Thickness_Mean_24h	float32	cm	Mean snow depth over the local day.
Snow_Thickness_LWE_Mean_24h	float32	cm	Mean snow liquid water equivalent over the local day.

Only in gee

Variables

Variable	Type	Unit	Description
Temperature_Air_2m_Max_24h	float32	K	Maximum air temperature at 2 m over the local day.
Temperature_Air_2m_Min_24h	float32	K	Minimum air temperature at 2 m over the local day.
Temperature_Air_2m_Mean_24h	float32	K	Mean air temperature at 2 m over the local day.
Temperature_Air_2m_Max_Day_Time	float32	K	Maximum air temperature at 2 m between 06 and 18 local time.
Temperature_Air_2m_Mean_Day_Time	float32	K	Mean air temperature at 2 m between 06 and 18 local time.
Temperature_Air_2m_Min_Night_Time	float32	K	Minimum air temperature at 2 m between 18 and 06 local time.
Temperature_Air_2m_Mean_Night_Time	float32	K	Mean air temperature at 2 m between 18 and 06 local time.
Dew_Point_Temperature_2m_Mean_24h	float32	K	Mean dew point temperature at 2 m over the local day.
Precipitation_Flux	float32	mm d-1	Total precipitation over the local day.
Precipitation_Rain_Duration_Fraction	float32	1	Fraction of the day's hours with rain.
Precipitation_Solid_Duration_Fraction	float32	1	Fraction of the day's hours with solid precipitation (freezing rain, snow, wet snow, rain-snow mix, ice pellets; no hail).
Solar_Radiation_Flux	float32	J m-2 d-1	Surface downward solar radiation summed over the local day.
ReferenceET_PenmanMonteith_FAO56	float32	mm d-1	Penman-Monteith reference evapotranspiration, FAO-56 method, over the local day.
Relative_Humidity_2m_06h	float32	%	Relative humidity at 2 m at 06 local time.
Relative_Humidity_2m_15h	float32	%	Relative humidity at 2 m at 15 local time.
Vapour_Pressure_Mean_24h	float32	hPa	Mean vapour pressure over the local day.
Vapour_Pressure_Deficit_at_Maximum_Temperature	float32	hPa	Vapour pressure deficit at the daily maximum temperature: saturated vapour pressure at the maximum temperature minus the daily mean vapour pressure.
Wind_Speed_10m_Mean_24h	float32	m s-1	Mean wind speed at 10 m over the local day.
Cloud_Cover_Mean_24h	float32	1	Mean total cloud cover over the local day, 0 to 1.
Snow_Thickness_Mean_24h	float32	cm	Mean snow depth over the local day.
Snow_Thickness_LWE_Mean_24h	float32	cm	Mean snow liquid water equivalent over the local day.

Technical metadata

The full spatial, temporal, and modeling detail behind the summaries elsewhere on this page.

Spatial

CRS: EPSG:4326
Extent (W S E N): [-180, -90, 180, 90]
Coverage: World
Grid spacing: 6 arc-minutes (~11 km at the equator)

Temporal

Start: 1979-01-01
End: open (present)
Updates: Daily

Climate methods

Downscaling: Grid- and variable-specific regression from ERA5 interpolated to 0.1°, trained on ECMWF HRES at 0.1°, so values follow the finer topography, land use and land-sea mask. · 0.1 degree

Provenance

Ordered processing steps, from source to publication.

source
Hourly ERA5 aggregated to local-time days and corrected to a 0.1° grid with regressions trained on ECMWF HRES, by Wageningen Environmental Research for the Copernicus Climate Change Service. Published on the Climate Data Store as NetCDF.
Derived from: ECMWF ERA5 reanalysis ↗

How to cite

Please cite the dataset and any associated peer-reviewed publications.

H. Boogaard, J. Schubert, A. De Wit, J. Lazebnik, R. Hutjes, G. Van der Grijn (2020) Agrometeorological indicators from 1979 to present derived from reanalysis. Version 2.0. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.6c68c9bb. Accessed through the CGIAR Climate Action Data Hub, https://cgiar-climate-data-hub.github.io/catalog/agera5_2.0/.

BibTeX

@misc{agera5_2020,
  title     = {Agrometeorological indicators from 1979 to present derived from reanalysis},
  author    = {Boogaard, H. and Schubert, J. and De Wit, A. and Lazebnik, J. and Hutjes, R. and Van der Grijn, G.},
  year      = {2020},
  version   = {2.0},
  publisher = {Copernicus Climate Change Service (C3S) Climate Data Store (CDS)},
  doi       = {10.24381/cds.6c68c9bb},
  note      = {Accessed through the CGIAR Climate Action Data Hub, https://cgiar-climate-data-hub.github.io/catalog/agera5_2.0/}
}

Attribution

Reproduce this credit line wherever you reuse the data.

Generated using Copernicus Climate Change Service information [year].

Contacts

Who produced this dataset, and who to reach about it.

Brayden YoungbergMaintainer

Alliance of Bioversity International and CIAT

b.youngberg@cgiar.org

Copernicus Climate Change Service (C3S), ECMWFLicensor

Producer

climate.copernicus.eu/ ↗

Wageningen Environmental Research

Producer

www.wur.nl/en/research-results/research-institutes/environmental-research.htm ↗

AgERA5 daily agrometeorological indicators

Important note

Quick start

Intended uses

Not recommended for

Access conditions

Data

Additional assets

File layouts

In every file

Dimensions

Only in zarr-geo-chunked, zarr-time-chunked, cds-api

Variables

Only in gee

Variables

Spatial

Temporal

Climate methods

How to cite

Attribution

Contacts