• ta and va have been back-extended to 1940 for the 5 pressure levels (and for 1hr, day, mon) to equate the other pressure level variables.

• Re-processed rlut and rlutcs, during the correction of 2022.09.21 there was an error in the timestep increment for some files (it went backwards).

• Back extended ps to 1940.

• Added orog invariant variable.
• Removed sic from invariant variables.

Added potential evapotranspiration evspsblpot at 1hr and mon time frequencies.

Due to the ERA5.1 E1 2000-20006 patch, variables esn snm wsgsmax lspf uvb cape clwvi clivi prlsprof prcprof prsn bld hfss hfls zmla rsds rlds rss rls rst rlut tauu tauv evspsbl xgwdparam ygwdparam gwd tasmax tasmin mrro rstcs rlutcs rsscs rlscs rsdt pr prsnc prlsns had some missing hours in that interval:

• some had 2000-01-01 starting at 07 (instead of 00 or 01 depending on whether the variable was Instantaneous or accumulated),
• and 2006-12-31 ending at 23 instead of 2007-01-01 00 (again, depending on the type of variable).

This missing timesteps are no longer there, the gaps have been filled with ERA5 E5 data, and the daily and monthly averages recalculated accordingly.

Variables rsds, rlds, rss, pr, prsn had 1H 2024.03.22 information missing in original grb file. This has been amended and re-run.

Back-extended 2 evapotraspiration variables esn and evspsbl (1hr, mon) from 1940 till now

Added two derived variables:

• hurs: or Near-Surface Relative Humidity. Derived from tdps and tas as the fraction of e_sat(tdps)/e_sat(tas) tpds and tas saturation vapour pressures.
• sfcwind: or Near-Surface Wind. Derived from uas and vas as sqrt(uas2 + vas2)

• all the variables have been updated for the whole 2023 year. NOTE: be aware that from 2023.12.31 till around 2023.09.01 the data corresponds to the ET stream, as the E5 has a delay of around 3 months.

• ERA5 rerun for 2000-2006 due to biases: ALL currently available variables and time-frequencies have been replaced with the improved ERA5 re-run for the period 2000 to 2006 (ERA5.1). This measure superseeds the one taken on 21.10.2021 where only ta was replaced.
• cvl, cvh tvl tvh sic variables have been re-formatted as fx variables

Bug in sf12 variables solved: the recently processed sf12 1hr variables had the first day duplicated at each file. The derived files (e.g. day and mon averages) have also been recomputed. This did not affect sf00 nor pl00 data.

NOTES: concerning sf12 data exclusively:

1. there is a problem with the source /pool/data/ERA5/ET/sf/fc/1H (hourly) grib files starting at 2023-04-18 (the files instead of going 00-23 go 07-06 next day, with a consequent gap of 7 hrs in April). Consequently 2023-04 and 2023-05 data have not been fully correctly processed. As soon as this is solved, 2023 year will be re-processed. This should not affect neither sf00nor pl00 data.
2. there is no 1940-01-01 file, 1940 starts with 2nd of Jan. Consecuently the monthly average of 1940-01 is incomplete.

All the variables have been updated to latest date: sf00, sf12 (including accumulated/max/min) and pl00.

All* the variables have been updated to latest date:

1. All sf00 and pl00 have been updated until beginnining of March 2023.
2. sf12 variables have been updated for the whole 2022 year (until 2022.12.31 23:00). There is a problem with the time axis in the original GRIB files at the /pool/data for 2023 files, as soon as this is solved they will be updated to latest date as well.

A selected group of variables have been back-extended until 1940:

1. pr tasmin tasmax wsgsmax tas uas vas psl tasminn tasmaxx
2. the following pl00 variables are available until 1959.01.01: ua va zg. But as soon as they are available at /pool/data they will also be back-extended until 1940 as well.

Additionally:

1. wsgsmax is being offered at 1hr, day, mon time frequencies now, instead of 1hr, mon.
2. land_area_fraction (sftlf ) variable (time_frequency=fx) has also beend added.

WARNING: due to the aforementioned problem sf12 variables of Accumulated, Maximum, Minimum type, whose day is defined from 01:00 to 00:00 (of the following day) lack the first hour of 2023 (2023.01.01 00:00). Affected variables are bld esn evspsbl gwd hfls hfss lspf mrro pr prcprof prlsns prlsprof prsn prsnc rls rlscs rlut rlutcs rsds rsscs rss rst rstcs snm rlds tasmax tasmin tauu tauv rsdt uvb wsgsmax xgwdparam ygwdparam.

The attributes of 16 have been modified to be more conformant with CF/CMIP6 anotation, including variable name for 7 radiation variables. In 2 cases the sign of the variable has also been inverted.

All the variables have been updated from 1979.01.01 to 2022.02.28 (both included).

The bias corrected data has been added: ta for years 2000-2006. See below for more info.

All the variables have been updated from 1979.01.01 to 2021.07.31 (both included).

All the variables have been updated from 1979.01.01 to 2021.01.31 (both included).

All the variables have been moved under ERA5. ERA5-FC no longer exists.

The following variables were modified and reprocessed (also reflected when needed in the tables that follow the above links):

1. var49: wsgmax10m is now wsgsmax && A (accumulated) is now I (instantaneous).
2. var169: ssrd is now rsds && I is now A && long_name has been modified. That means that ssrd no longer exists in XCES.
3. var176: rsds is now rss && long_name modified.
4. var180 (tauu): I is now A.
5. var181 (tauv): I is now A.
6. var195 (xgwdparam): I is now A.
7. var196 (xgwdparam): I is now A.
8. var201 (tasmax, tasmaxx): A is now I.
9. var202 (tasmin, tasminn): A is now I.

Whether a variable is I or A affects the daily and monthly aggregation. The “day” is considered from 01-00 for A while from 00-23 for I.

ECMWF started to provide ERA5 data from earlier than 1979 (as of 2020-07-08). However, the present data (as offered by ECMWF via CDS) are preliminary, only (see here). DKRZ will copy these data as soon as they are available in a final version, which is supposed to be around the end of 2021. Once the data is there we will proceed to CMORize and ingest it to the system.

Recently the ECMWF announced a rerun of ERA5 (referred as ERA5.1) for the 2000-2006 period to improve upon the cold bias in the lower stratosphere seen in ERA5 during this period (more info also here). Consequently ALL the variables having replaced by this stream.