ERA5 terminology: 'time' and 'steps', and instantaneous, accumulated and min/max parameters

This article applies to the ERA5 sub-daily datasets:

• ERA5 high resolution (HRES), sub-daily data (streams oper/wave)
• ERA5 ensemble (EDA), sub-daily data (streams enda/ewda).

This article does not apply to other ERA5 streams, for example monthly data. See what streams are available(ERA5 documentation, section 'Data organisation').

Instantaneous parameters vs accumulated parameters

Each parameter is classed as either 'instantaneous' or 'accumulated', depending on the temporal property of the parameter:

• Parameters that refer to a point in time are 'instantaneous', for example temperature at 12:00.
• Parameters that refer to a time period are 'accumulated', for example precipitation between 12:00 and 18:00.

Analysis and 'time'

In ERA5 two types of data are available, 'analysis' and 'forecast'

An analysis is computed for each full hour (i.e. for 00:00, 01:00, ... , 23:00). An analysis is a point-in-time snapshot of the atmospheric conditions at a specific time, hence an analysis can by definition only contain instantaneous parameters.

However, to compute the analysis a 30-minute window (+/-15 min) around the validity time is taken into account:

Forecast and 'time'/'step'

A forecast starts with the atmospheric conditions at a specific time (the forecast start time or initialization time). In ERA5 a new forecast is initialized every day at 06:00 and 18:00 UTC. The forecast then computes the future conditions (internally in 30-minute intervals or 'model steps'), and the output data is aggregated to 18 one-hour intervals (steps) for archiving. For example, an ERA5 forecast initialized at 06:00 produces ERA5 archive data in 18 one-hourly steps:

• data for instantaneous parameters is valid at the end of the step. For example, data from the forecast at time=06:00, step=1, represents the parameter value at 06:00 + (1 * 1 hour), i.e. at 07:00.
• data for accumulated parameters is aggregated into hourly steps, starting at the end of the previous step. For example, precipitation at time=06:00, step=4 represents precipitation up to 06:00 + 4 * 1h, starting from 06:00 + 3 * 1h, i.e. precipitation in the period 09:00 to 10:00.

The following table summarizes how instantaneous and accumulated parameters, from analysis and forecast are available in ERA5:

Retrieving data

Instantaneous parameters

All the analysed fields and many forecast fields (e.g. temperature) are referred to as "instantaneous" parameters, but are actually representative of time scales equal to the model time step (30min in ERA-Interim).

Let's say you want to extract forecast 2m-temperature at 3pm (15:00). Select as start time 12:00 (midday), with step 3 (+3 hours). This gives you the temperature at 15:00.

Accumulated parameters

In ERA-Interim the forecast accumulations (e.g. total precipitation and radiation parameters) are accumulated from the start of the forecast, ie. from T=00:00 or T=12:00.

For example, Snowfall with Time=12:00 and Step=9 gives the accumulated Snowfall in the time period 12:00 to 21:00 (12:00+9h).

See also ERA-Interim data documentation, table 9, and examples 1 to 3 below..

In ERA5, the short forecast accumulations are accumulated from the end of the previous step.

Accumulated parameters are not available from the analyses.

Examples

Example 1

Let's say you want to extract daily total precipitation. On the ECMWF data server you can select as start times 00:00 (midnight) and 12:00 (midday), both with step 12. This gives you two records:

* accumulated precipitation from 00:00 to 12:00 (midnight + 12 hours)

* accumulated precipitation from 12:00 to 24:00 (midday + 12 hours)

Then sum the two values to get the daily total.

(Note the units for total precipitation is meters.)

Example 2

You need the average precipitation per hour between 00:00 and 06:00.

To obtain the average of accumulated fields between two forecast steps (STEP1 and STEP2) it is necessary to retrieve the fields for the two steps: (Field(STEP1) and Field(STEP2)) then calculate the difference and divide by the time difference:

(Field(STEP2) - Field(STEP1)) / (STEP2 - STEP1).

To obtain average precipitation per hour between 00:00 and 06:00, download Total Precipitation with time 00:00 and steps 0 (tp0) and step 6 (tp6), then calculate:

( tp6 - tp0 ) / ( 6 - 0 ) 

(Note the units for total precipitation is meters.)

Example 3

You need total precipitation for every 6 hours.

Daily total precipitation (tp) is only available with a forecast base time 00:00 and 12:00, so to get tp for every 6 hours you will need to extract (and for the second and fourth period calculate):

tp(00-06) = (time 00, step 6)

tp(06-12) = (time 00, step 12) minus (time 00, step 6)

tp(12-18) = (time 12, step 6)

tp(18-24) = (time 12, step 12) minus (time 12, step 6)

(Note the units for total precipitation is meters.)

Example 4

You need the daily minimum and maximum of 2m temperature.

You can use the parameters 'Maximum temperature at 2 metres since previous post-processing' (mx2t, parameter id 201) and 'Minimum temperature at 2 metres since previous post-processing' (mn2t, parameter id 202). These two parameters are available from the forecasts initialized at 06:00 and 18:00 in one-hour steps:

mx2t(06-07) = (time 06:00, step 1)

mx2t(07-08) = (time 06:00, step 2)

mx2t(08-09) = (time 06:00, step 3)

...

mx2t(17-18) = (time 06:00, step 12)

mx2t(18-19) = (time 18:00, step 1)

mx2t(19-20) = (time 18:00, step 2)

mx2t(20-21) = (time 18:00, step 3)

...

mx2t(05-06) = (time 12:00, step 12)

Sample Python script for the ECMWF WebAPI to retrieve Min and Max 2m temperature since previous post-processing:

#!/usr/bin/env python
from ecmwfapi import ECMWFDataServer
server = ECMWFDataServer()
server.retrieve({
    "class": "ea",
    "dataset": "era5",
    "stream": "oper",
    "expver": "1",
    "date": "2016-01-01/to/2016-01-31",
    "type": "fc",
    "levtype": "sfc",
    "param": "201.128/202.128",             # 'Maximum temperature at 2 metres since previous post-processing' and 'Minimum temperature at 2 metres since previous post-processing'
    "time": "06:00:00/18:00:00",            # 2 forecasts per day, at T=06:00 and T=12:00
    "step": "1/2/3/4/5/6/7/8/9/10/11/12",   # For each forecast 18 hourly steps are available. Here we use only steps 1 to 12.
    "grid": "0.30/0.30",
    "area": "60/-10/20/50",
    "target": "output",                     # change this to your output file name
})

This script retrieves the maximum and minimum 2m temperature from 2016-01-01 06:00 to 2017-02-01 06:00, as one-hour periods. To identify the daily maximum/minimum 2m temperature, find the the maximum of these 3-hour maxima/minima.

Further information

About ERA-Interim: http://www.ecmwf.int/en/research/climate-reanalysis/era-interim

ERA-Interim documentation: http://www.ecmwf.int/en/elibrary/8174-era-interim-archive-version-20

Worth reading about spin-up in ERA-Interim: http://www.ecmwf.int/en/elibrary/10381-forecast-drift-era-interim

Related articles

• Page:
  C3S Documentation Centre
• Page:
  C3S Knowledge Base (CKB)
• Page:
  C3S Seasonal Forecast known issues
• Page:
  C3S User Support Journey
• Page:
  Climate Data Store (CDS) API Keywords