English version see below

CMIP5 Daten basierend auf dem MPI-ESM Modell : ESG Archiv

Die im CMIP5 Projekt für das ESG Archiv angeforderten Variablen sind in einer Exceltabelle aufgeführt, die Kurznamen, NetCDF/CF Standardnamen, und lange Namen sowie Einheiten und andere festgelegte Aspekte für die Daten enthält (xlsx). Nicht alle verlangten Variablen können abgeliefert werden. Die im Folgenden aufgeführten PDF-Dokumente enthalten Tabellen mit den Kurznamen der abgelieferten Variablen, sowie die Experimente und die Perioden, für die die Daten im ESG erwartet werden können. Dies weicht in einigen Details von den Festlegungen des CMIP5 Projekts ab (z.B.keine Variablen des Bereichs ocnBgchem für die dekadischen Experimente).





CMIP5 Daten basierend auf dem MPI-ESM Modell : GPFS der DKRZ HPC Platform 'blizzard' (IBM power-6)

Die für den ESG Datenknoten am WDCC aufbereiteten Daten aus CMIP5 Experimenten mit MPI-ESM sind für Inhaber von blizzard-Benutzernummern auch auf den GPFS Platten am DKRZ zugreifbar. Im GPFS Dateiverzeichnis gibt es allerdings keine Versionierung, so dass nicht offensichtlich ist, ob Parameter ersetzt wurden. Eine Kontrollmöglichkeit besteht darin, eine Wiki Seite mit Warnungen für Datennutzer zu konsultieren. Dort sind Löschungen, Ersetzungen und Ergänzungen im Dateiverzeichnis dokumentiert. Man beachte, dass immer ganze 'atomic data sets' ersetzt werden. Mit einiger Zeitverzögerung (Tage) finden die Änderungen Eingang ins ESG.

Liste der CMIP5 Experimente basierend auf den MPI-ESM Modellen durchgeführt am DKRZ auf der HPC Platform 'blizzard' (IBM power-6)




English version :

From the Past to the Future:

New Climate Simulations MPI-ESM for Science and Society

The Max Planck Institute for Meteorology (MPI-M) and the German Climate Computing Center (DKRZ) organized an informative meeting concluding the BMBF funded consortia simulations in the framework of CMIP5 (Coupled Model Intercomparison Project Phase 5). The results of these simulations are an essential contribution to the 5th assessment report (AR5) of the IPCC.


pdf (German only).


Thursday, 23 February 2012

Consortial Simulations with the MPI-M Earth System Model for CMIP5 -
Background, Experiments, Scenarios und Models pdf

Marco Giorgetta

Climate model data : Requirements and Usage pdf

Michael Lautenschlager

Friday, 24 February 2012 (Workshop for data users)

Survey of CMIP5 Archives at DKRZ pdf

Stephanie Legutke

EarthSystemGrid (ESG): Access, Search and Downloadpdfpptx

Estanislao Gonzalez

Long Time Archiving and DataCite Data publishingpdf

Martina Stockhause

WDCC/CERA: Forcing Data for RCMs pdfpptx

Peter Lenzen

Local Disk and Tape Archives at DKRZ pdf

Heinz-Dieter Hollweg

Data Processing pdf

Jörg Wegner


The German Contribution to the CMIP5 (IPCC/AR5) Data archive pdf

Ksenia Glushak, S. Legutke, H. Widmann, J. Wegner, H.-D. Hollweg, K. Fieg, M. Lautenschlager, M. Giorgetta, J. Jungclaus and C. Reick

CORDEX Experiments with the RCM 'CCLM'pdf

Karin Meier-Fleischer, M. Schubert-Frisius, P. Lenzen, K. Keuler, H-J. Panitz, M. Büchner and S. Legutke

CMIP5 Data Based on MPI-ESM : the ESG data node

The CMIP5 atomic data sets requested for the CMIP5 ESG archive are listed in an excel file, including short names, NetCDF standard names, and long names as well as units and other data attributes as specified in the CMIP5 data protocol (xlsx). Not all requested variables can be provided. The PDF documents below contain short names of the variables delivered, as well as the experiments and periods for which the data can be expected to appear in the ESG data node. Some details differ from the CMIP5 specifications (e.g. ocnBgchem variables are not provided for decadal experiments).





CMIP5 Data Based on MPI-ESM : GPFS archive on DKRZ's HPC platform 'blizzard' (IBM power-6)

The data from CMIP5 experiments with MPI-ESM, processed for the ESG data node at WDCC/DKRZ, can be used or downloaded by blizzard users from the GPFS disks at DKRZ directly. There is no versioning for the GPFS directories, however. Therefore, it is not obvious whether an atomic data set has been changed since the last download. A certain degree of controlling file versions is possible, however, by consulting a Wiki page containing warnings for data user. The page contains notes on file deletions, replacements, or additions. Note that always complete 'atomic data sets' are concerned. The updates find their way into the ESG at WDCC/DKRZ with some delay (days).

List of CMIP5 Experiments based on MPI-ESM and performed on the HPC DKRZ platform 'blizzard' (IBM power-6).




CMIP5: Infos, Tipps & Tools

Convert MPI-ESM Ocean Grids to Regular Grids

The Ocean Grids of the earth system models MPI-ESM-LR,MPI-ESM-P and MPI-ESM-MR are curve linear. The grid north pole of the MPI-ESM-LR and ESM-MPI-P models is moved to Greenland (GR15), the south pole remains at its geographical position. The MPI-ESM-MR model is running on a tripolar grid with the north pole degenerated to a line with end points over Canada and Sibiria (TP04). The grid of the southern hemisphere is the geographical one and is matched to the tripolar hemisphere at the equator.
Both grids also contain halos: the first and the last longitude lon(1,:) and lon(n,:) are mirrored on the opposite side: lon(n,:),lon(1,:)....,lon(n,:),lon(1,:) and these halos have to be removed before transformation.
All necessary informations are stored in the netcdf file. Use
ncdump -h ifile.nc
to get informations about dimensions, variables and attributes. To remove the halos and to convert the grids to regular grids use the CDOs (cdo-homepage).
Remove the halos:
cdo selindexbox,2,255,1,220 ifile.nc(MPI-ESM-LR/P) ofile.nc
cdo selindexbox,2,801,1,404 ifile.nc(MPI-ESM-MR) ofile.nc
and convert to regular grid using a bilinear interpolation:
cdo remapbil,r{mlon}x{mlat} ofile.nc rfile.nc
with {mlon} and {mlat} number of longitudes/latitudes of your choice (should be near to the original values). For example:
cdo remapbil,r240x220 ofile.nc(MPI-ESM-LR/P) rfile.nc
cdo remapbil,r800x400 ofile.nc(MPI-ESM-MR) rfile.nc

To rotate ocean vector variables use the not documented cdo operator mrotuvb. Example:
#--- rotate uo and vo (uo and vo results in 1 output file)
#--- uo6.nc and vo6.nc contain the 6m level of uo/vo
cdo mrotuvb uo6.nc vo6.nc uv6.nc
#--- Now use selindexbox to get rid of the halos
#--- get grid from a scalar variable
cdo selindexbox,2,255,1,220 tos_Omon_MPI-ESM-LR_historical_r1i1p1_185001-200512.nc tosh.nc
cdo griddes tosh.nc > grids.asc
#--- set grid to scalar grid
cdo setgrid,grids.asc uv6h.nc uv6hs.nc
#--- Go on with remapbil

Metadata errors in lgm/midholocene r1i1p2 zosga and zossga

The variables zosga and zossga were computed with a wrong startyear mean. It was the same as in the lgm(midHolocene)_r1i1p1-P experiments:
wrong values: zosga_mean = 0.63886; zossga_mean = 2.5415 ;
correct values: zosga_mean = 0.64122; zossga_mean = 2.5302.
wrong values: zosga_mean = -0.34760; zossga_mean = -0.36278 ;
correct values: zosga_mean = -0.35626; zossga_mean = -0.37334
with zosga = (zosga - zosga_mean) + (zossga - zossga_mean)
and zossga = zossga - zossga_mean.
The data will not be replaced.

Metadata errors in historical r2i1p1 umo,vmo,msftmyz

The umo,vmo, and msftmyz atomic data sets from historical_r2i1p1-P have been given a wrong parent_experiment (past1000 instead of piControl).
Accordingly, the branch time is wrong: 1850-850 in days, instead of zero.
The data will not be replaced.

Info on external forcing in the past1000 experiment

Info on stratospheric (volcanoe) aerosol external forcing in the experiment with MPI-ESM-LR/MR

The external forcing data sets of volcanic aerosols for the historic period from 1850 to 1999 is based on data provided by G. Stenchikov. It is an extended version of the Pinatubo aerosol data set (Stenchikov et al. (1998)) from stellite measurements of aerosol extinction and effective radii and was successfully applied in climate model studies (Stenchikov et al. (2004, 2009); Thomas et al. (2009a,b)). The data are available on 40 pressure levels and interpolated to the horizontal grid of ECHAM6 (it is the same for MPI-ESM-LR and MPI-ESM-MR). It contains monthly mean zonal averages of the aerosol extinction, the single scattering albedo, and the asymmetry factor as a function of altitude, wavelength, and time. Furthermore, the integral aerosol optical depth of a column is given as a function of wavelength and time. For the simulation of years after 1999, the 1999 data are used.

Sometimes the data set is referred to Sato et al. (1993) which is the original data source of the data set from Gera Stenchikov.

Sato, M., J.E. Hansen, M.P. Mc Cormick, and J.B. Pollack 1993. Stratospheric aerosol optical depth, 1850-1990. J. Geophys. Res. 98, 22987-22994.

Stenchikov, G. et al. (2004): Arctic oscillation response to the 1991 Pinatubo eruption in the SKYHI general circulation model with a realistic quasi–biennial oscillation. J. Geophys. Res., 109, doi:10.1029/2003JD003699.

Stenchikov, G. et al. (2009): Volcanic signals in oceans. J. Geophys. Res., 114,doi:10.1029/2008JD011673.

Stenchikov, G. L. et al. (1998): Radiative forcing from the 1991 Mount Pinatubo volcanic eruption. J. Geophys. Res., 103, no. D12, 3,837–13,857.

Schmidt, H., Rast, S., Bunzel, F., Esch, M., Giorgetta, M. A., Kinne, S., Krismer, T., Stenchikov, G., Timmreck, C., Tomassini, L., & Walz, M. (2013). The response of the middle atmosphere to anthropogenic and natural forcing in the CMIP5 simulations with the MPI-ESM. Journal of Advances in Modeling Earth Systems, 5, 98-116.

MadWiki: DatennutzerWorkshop (last edited 2014-08-20 11:38:38 by StephanieLegutke)