If you experience problems or need assistance please send an email to
. A ticketing system will acknowlegde your enquiry and generally your request will be processed within one working day. When sending attachments (i.e. screenshots) please bear in mind that the ticketing system will only accept mails up to 3MB in size.
For an overview of the current status of the WDCC's services please visit http://wdcc-status.dkrz.de/
Accounts & Password
How do I get access to the CERA database?
Please send an email containing your name and institute affiliation to .
How can I reset a lost password?
Please contact to reset your password.
How can I change my password?
The password can be changed after login on the CERA UI main page.
What should I do if I think I found an error in the metadata?
with details of the metadata entry in question and the suggested correction(s).
How is the data licensed?
All data and pages available from WDCC are licensed under a Creative Commons License
as far as those conditions are not in any way modified by the following conditions or
by any conditions specific to data or pages. For more detail see 'Conditions WDCC Database':
How to cite data?
Citation can be found via webbrowser using the CERA user interface (http://cera-www.dkrz.de/WDCC/ui/Index.jsp):
- CERA WWW-Gateway >> 'Browse experiments' >> 'Experiment information' >> 'CERA UI Compact' >> 'Citation Elements'
- CERA WWW-Gateway >> 'Browse experiments' >> 'Experiment information' >> 'Extended Entry list' >> select 'dataset_group' >> 'CERA UI Compact' >> 'Citation Elements'
For more information see also: http://www.komfor.net/cite-data.html
Experiment citation on http://cera-www.dkrz.de/WDCC/ui/Compact.jsp?acronym=MODIS__Arctic__MPF_V02
Rösel, Anja; Kaleschke, Lars; Kern, Stefan (2015): Gridded Melt Pond Cover Fraction on Arctic Sea Ice derived from TERRA-MODIS 8-day composite Reflectance Data bias corrected Version 02. World Data Center for Climate. doi:10.1594/WDCC/MODISArcticMPF_V02
Dataset group citation on http://cera-www.dkrz.de/WDCC/ui/Compact.jsp?acronym=NCCNMhg
Bentsen,Mats; Bethke,Ingo; Debernard,Jens; Drange,Helge; Heinze,Christoph; Iversen,Trond; Kirkevåg,Alf; Seland,Øyvind; Tjiputra,Jerry (2012): cmip5 output1 NCC NorESM1-M historicalGHG, served by ESGF. World Data Center for Climate. doi:10.1594/WDCC/CMIP5.NCCNMhg
Data download in general
How can I access data stored in the CERA database?
CERA datasets can either be downloaded via webbrowser using the CERA user interface or with the java-based commandline tool Jblob that can be used on any OS supporting java (Oracle or OpenJDK).
Why does it take so long before the download actually starts?
A large part of the datasets available from the CERA database are stored on
tapes and have to be retrieved before the actual transfer. Only downloads for
datasets that are currently present in the system's disk-cache will start
immediately. It is not necessary to start a download again if the first request
ist still active but hasn't transferred anything yet.
Can the data format be changed during download (i.e. Grib => NetCDF)?
Currently it is not possible to change a dataset's format during download.
Is it possible to select a geographic region during data download?
Jblob offers a subset of cdo processing capabilities, including selection of a geographic region, please take a look at the Jblob documentation for more information. Regional selection from the interactive download is currently not possible.
Can wget be used to download data from CERA?
Wget access is currently not supported.
What is the maximum size of downloads?
Currently a maximum of 16TB can be downloaded or selected for processing in a single request.
Can I get more information about the datasets stored within the CERA database?
In addition to the standard metadata there are detailed header information available for many datasets, these can be accessed (if available) from the 'Entry information' page by clicking on the "Data headers" button in the 'General information' section.
CERA user interface
How many parallel downloads can I start?
Please limit the number of parallel downloads to 4. If you need to download
large numbers of datasets please take a look at our batch download tool
create a script for chain-downloads of datasets. Additionally there is a
feature available on many list page that allows you to download .csv lists for
script generation and/or pre-defined Jblob download scripts.
Are download managers (including browser add-ons) supported?
Download managers are not supported, if you need to use them regardless
please configure a retry interval of at least 60 seconds and limit the number
of parallel downloads to 4. If usage of a download manager results in too many
request in a very short time your account may be temporarily blocked!
How many parallel downloads can I start?
Please limit the number of parallel downloads to 4 and the number of processing (cdo) requests to 1 when
processing very large datasets.
How do I know if my Jblob download was successful?
Please check the return value of the Jblob call. A return code of 0 is a successful transfer, any other
value indicates a problem during data download. Please take a look at the Jblob documentation
for a description of the different exit values.
Is there a way to speed up processing during Jblob downloads?
If you use more than one cdo option, e.g. 'sellonlatbox' and 'selcode', the order of the options can have
a significant impact on the time needed for processing. The cdo processing evaluates options from right to
left, so it is preferable to put the option that results in the biggest amount of data reduction as the final
cdo option. For example, if a dataset has a large number of variables/parameters it will be most beneficial
to select the required parameter(s)/variable(s) first:
jblob --dataset <acronym> --cdo "sellonlatbox,0,180,0,90" --cdo "selcode,167"
What happens if a download of a processing request is interrupted?
The processing request will be finished regardless and the result will be cached for a certain amount of time.
If you restart an interrupted processing request and the background processing has been finished the download
of the result should start immediately, otherwise the new request will be queued until the processing is finished.
Common Jblob error messages
cdo(2) sellonlatbox (Abort): Unsupported grid type!
In most cases appending '--cdo "selgridname,gaussian"' to the processing request will avoid the error, e.g.
jblob --dataset <acronym> --cdo "sellonlatbox,0,180,0,90" --cdo "selgridname,gaussian"
If this doesn't work please contact
or download a single record completely and use 'cdo info' to find out the designation of the lonlat grid referenced in the dataset.