ERA5-Land datasets¶
ERA5-Land is a global reanalysis of land surface variables produced by ECMWF. It provides a consistent record from 1950 to the near-present at ~9 km resolution. The Open Climate Service ships six ERA5-Land dataset templates covering temperature and precipitation at daily and monthly resolution.
Setup¶
ERA5-Land data is fetched from two sources depending on the dataset. Both are free but require registration.
Earth Data Hub (hourly and daily datasets)¶
- Create a free account at earthdatahub.destine.eu.
- Go to Account settings → My API keys → Standard API keys and generate a key.
- Add it to
~/.netrc:
Windows: use %USERPROFILE%\_netrc (underscore prefix) with the same content.
Copernicus Climate Data Store (monthly datasets)¶
- Register at cds.climate.copernicus.eu and get your API key from your profile.
- Create
~/.ecmwfdatastoresrc:
Ingest your first ERA5-Land dataset¶
Once authenticated, ingest ERA5-Land daily temperature:
curl -s -X POST http://127.0.0.1:8000/ingestions \
-H "Content-Type: application/json" \
-d '{
"dataset_id": "era5land_temperature_daily",
"start": "2024-01-01",
"end": "2024-01-31",
"publish": true
}' | jq
Temperature datasets¶
Daily (pre-computed)¶
| Property | Value |
|---|---|
| Dataset ID | era5land_temperature_daily |
| Variable | t2m |
| Units | °C |
| Period | Daily |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | Earth Data Hub |
| Lag | ~4 weeks |
Daily mean temperature for the UTC calendar day (00:00–23:59 UTC). For most instances this is the right choice.
For instances where health data is reported using a local calendar that does not align with UTC (e.g. UTC+3 East Africa), use era5land_temperature_daily_from_hourly instead.
Sync — new days are appended incrementally. Earth Data Hub provides pre-computed daily means; this dataset uses those directly for efficiency and falls back to a CDS-derived daily mean for the most recent ~4 weeks not yet published by Earth Data Hub.
Daily (from hourly, timezone-aware)¶
| Property | Value |
|---|---|
| Dataset ID | era5land_temperature_daily_from_hourly |
| Variable | t2m |
| Units | °C |
| Period | Daily |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | Earth Data Hub |
| Lag | ~4 weeks |
Daily mean temperature computed by averaging the 24 hourly values that fall within the local calendar day. The local day is defined by utc_offset_hours in climate-service.yaml (default: 0 = UTC).
For a UTC+3 instance, "January 1" covers 21:00 UTC December 31 through 20:59 UTC January 1 — matching the calendar day as experienced locally.
When to use this instead of era5land_temperature_daily: when your health data is reported by local date and your instance has a non-zero utc_offset_hours.
Monthly¶
| Property | Value |
|---|---|
| Dataset ID | era5land_temperature_monthly |
| Variable | t2m |
| Units | °C |
| Period | Monthly |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | CDS reanalysis-era5-land-monthly-means |
| Lag | ~2 months |
Monthly mean 2 m temperature, downloaded directly from the Copernicus monthly averages product. Each value is the mean of all hourly temperature readings in the calendar month.
Precipitation datasets¶
Precipitation values represent total precipitation reaching the surface, including rain and snow as liquid-equivalent water. Units are always mm (millimetres of water depth).
About daily totals
Daily precipitation values represent the total water that fell during that calendar day. Summing the daily values over a month matches the monthly total dataset to within rounding.
Daily (UTC calendar day)¶
| Property | Value |
|---|---|
| Dataset ID | era5land_precipitation_daily |
| Variable | tp |
| Units | mm |
| Period | Daily |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | Earth Data Hub |
| Lag | ~4 weeks |
Total precipitation for the UTC calendar day. Each value is the sum of all precipitation that fell within 00:00–23:59 UTC.
For instances where health data is reported using a local calendar, use era5land_precipitation_daily_from_hourly instead.
Sync — Earth Data Hub covers history. For the most recent ~4 weeks not yet published by Earth Data Hub, daily totals are derived from an hourly download from CDS.
Daily (timezone-aware)¶
| Property | Value |
|---|---|
| Dataset ID | era5land_precipitation_daily_from_hourly |
| Variable | tp |
| Units | mm |
| Period | Daily |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | Earth Data Hub |
| Lag | ~4 weeks |
Total precipitation summed over the local calendar day, defined by utc_offset_hours in climate-service.yaml.
When to use this instead of era5land_precipitation_daily: when your health data is reported by local date and your instance has a non-zero utc_offset_hours.
Monthly¶
| Property | Value |
|---|---|
| Dataset ID | era5land_precipitation_monthly |
| Variable | tp |
| Units | mm |
| Period | Monthly |
| Coverage | Global, 1950–present |
| Resolution | ~9 km |
| Source | CDS reanalysis-era5-land-monthly-means |
| Lag | ~2 months |
Total precipitation for the calendar month in mm. The CDS monthly-means product provides the mean daily precipitation; this is multiplied by the number of days in the month at ingest time to give a directly comparable monthly total.
This is the recommended dataset for monthly precipitation analysis and climate–health modelling. It is produced by Copernicus and represents the most authoritative ERA5-Land precipitation product.
Choosing the right dataset¶
| Use case | Recommended dataset |
|---|---|
| Daily health modelling, UTC-aligned | era5land_temperature_daily / era5land_precipitation_daily |
| Daily health modelling, local calendar | era5land_temperature_daily_from_hourly / era5land_precipitation_daily_from_hourly |
| Monthly climate–health modelling | era5land_temperature_monthly / era5land_precipitation_monthly |
| CHAP integration (South Sudan, East Africa) | era5land_precipitation_monthly |
UTC offset configuration¶
For instances in non-UTC time zones, set utc_offset_hours in climate-service.yaml:
This affects all daily datasets — both the UTC and local-time variants use the offset when computing day boundaries.