=== WARNINGS ===


The list below gives warnings for users of CMIP5 atomic data sets based on the MPI-ESM earth system model of the Max-Planck Institute (MPI-M) for Meteorology in Hamburg, Germany.

The issues are listed in chronological order with the date being the time of detection or correction according to the text. Note that the dates relate to the changes of the GPFS file system on the HPC platform 'blizzard' at DKRZ (/work/ik0555/cmip5/archive). There is some time delay between changes of this file system and the corresponding changes in the ESG data node at DKRZ.

Note that ocnBgchem atomic data sets are not delivered for amip*, sstClim*, or decadal* experiments, nor for any experiment based on the paleo version of the model (MPI-ESM-P).

The data in /work/ik0555/cmip5/archive is created with the help of the CMOR software provided by PCMDI. This implies that it has passed the quality check level1 (QCL1). Shortly after creation, and before the data is published in the ESG, it is checked again with the software developed at DKRZ (QCL2). The protocols of the QCL2 are available on the blizzard GPFS for review.


<!> 26 September 2012

The variables zosga and zossga of member r1i1p2 of experiments midHolocene and lgm performed with model MPI-ESM-P were computed with incorrect offset values.

The intent was to substract the annual mean of the first year (calendrical year 1850) of each simulation from all values. However, the same offset was used for both members r1i1p2 and r1i1p1 of lgm and midHolocene despite the fact that the two members have a different spinup history, and therefore different initial states. Note that the values used for member r1i1p2 are the correct values of member r1i1p1.

midHolocene offsets of member r1i1p2:

lgm offsets of member r1i1p2:

<!> 22 August 2012

The three aqua*-LR experiments have been rerun, and the data of these experiments are available at /work/ik0555/cmip6/archive (together with the old directories that have been renamed to aqua*-LR_DoNotUse). The difference between the old and new versions is a 1-day spinup (31 Dec 1978).

<!> 15 August 2012

The three aqua*-LR directories have been renamed to aqua*-LR_DoNotUse because there was a problem detected in the first output record of at least code 202 (tasmin atomic data set). The problem is already present in the raw model output and may be due to a wrong initialisation (the aqua experiments are the only CMIP5 experiments which are not restarted but initialized). The data can be used if the 1st record (each atomic data set in order to be on the save side) is ignored. The data will be regenerated, i.e. the experiments will be rerun.

<!> 01 August 2012

All areacella variables in all experiments are checked and contain the experiment specific attributes. No changes necessary.

<!> 30 July 2012

All basin variables in all experiments are checked and contain the experiment specific attributes. No changes necessary.

<!> 16 July 2012

The volcIn2010_r1i1p1-LR experiment has been run with a wrong experiment name (volcln2010 instead of volcIn2010). This will remain the name of the raw model output data files und the data files of the standard post-processing (ATM*, BOT*, etc.). The data in /work/ik0555/cmip5/archive, however, will be given the correct name.

<!> 16 July 2012

The umo,vmo, and msftmyz atomic data sets from historical_r2i1p1-P have been given a wrong parent_experiment (past1000 instead of piControl). The data will not been replaced.

<!> 19 June 2012

The amipFuture_r1i1p1-LR experiment has been run with wrong SST boundary conditions. The experiment has been rerun, and the data are replaced.

<!> 16 June 2012

All umo,vmo atomic data sets from all experiments have been deleted, because a slightly wrong algorithm has been used for their calculation (i.e. visually the data seem to be correct). The data is being replaced.

All msftmyz atomic data sets from the non-decadal experiments have been deleted. The reason is that at rare occasions (a single one is known) the basin indices are exchanged. The data is being replaced. Downloaded data may be used with caution. Incorrect values are detected by a change in bottom relief (or 1 year peaks in curves).

<!> 04 June 2012

All variables described below are recalculated and available again:

frn, spco2, dpco2, dpo2 since 24/25. 04. 2012 except 1pctCO2_r1i1p1-MR and abrupt4xCO2_r1i1p1-MR since 08. 05. 2012.

dcalc, intpcalcite, fvegLitter since 17. 04. 2012.

zosga, hfds, mfo, transifs, fco2nat, fLuc, fFire, nbp, burntarea since 17. 02. 2012 except decadals LR since 03. 05. 2012 and decadals MR since 31. 05. 2012.

areacella since 24. 04. 2012.

basin since 20. 04. 2012

<!> 17 April 2012

All spco2, dpco2, and dpo2 atomic data sets are too large by a factor of 0.101325 because of wrong units (µatm as in an old standard_output table version instead of Pa) and have been deleted. They will be newly provided a.s.a.p.

All frn atomic data sets contain wrong values and have been deleted. They will be newly provided a.s.a.p.

All fVegLitter atomic data sets are too large by a factor of 86400 (per day units instead of per second) and have been deleted. They will be newly provided a.s.a.p.

All intpcalc atomic data sets contain the values of the CMIP5 variable intpcalcite because the calculation was based on an old version of the standard_output table. The intpcalc atomic data sets have been deleted. The intpcalcite atomic data sets will be provided a.s.a.p.

All dcalc atomic data sets are too small by a factor of 1.e3 and have been deleted. They will be newly provided a.s.a.p.

<!> 30 March 2012

spco2, dpco2, and dpo2 are too large by a factor of 1/0.101325 in all experiments. They will be replaced with the right values a.s.a.p..

<!> 30 March 2012

frn is wrong and can not been used. It will be deleted and replaced a.s.a.p..

<!> 12 March 2012

mfstmyz in all experiments has been rebuilt in order to correct the order of regions in the region coordinate variable.

<!> 12 March 2012

chl is corrected in all experiments except piControl-MR (is was too low by 1.e6) .

<!> 09 March 2012

sim is corrected in all experiments except decadal* and piControl-MR.

<!> 02 March 2012

sim is corrected in all experiments except decadal*, esmControl-LR, and piControl-MR.

<!> 24 Feb 2012

sstClim4xCO2_r1i1p1-LR was rerun. All atomic data sets have been replaced. (Note for blizzard users: the model raw output of the old experiment (run with wrong restart-files) can be found at /work/ik0555/cmip5/experiments/sstClim_r1i1p1-LR_DoNotUse.)

<!> 23 Feb 2012

chl atomic data sets are too large by a factor 1.e-6 and will be corrected a.s.a.p..

<!> 21 Feb 2012

sim atomic data sets contain wrong values and will be corrected a.s.a.p. The wrong data sets are deleted.

<!> 1 Feb 2012

burntArea, fco2antt, fco2nat, fFire, fLuc, nbp, zosga, transifs, mfo, co3satcalc, and hfds are wrongly diagnosed and have been deleted. They will be replaced a.s.a.p.

<!> 25 Nov - 27 Nov 2011; 5 Dec (year 2006 only)

Due to technical problems all atomic data sets of historical_r3i1p1-MR have been regenerated; no changes of values.

<!> 23 Nov - 25 Nov 2011; 1 Dec 2011 (year 2006 only)

Due to technical problems all atomic data sets of historical_r2i1p1-MR have been repeated; no changes of values.

<!> 03 - 17 Nov 2011

umo, vmo have been regenerated for all decadal*-LR, all *-MR, and all *-P experiments. For the other *-LR experiments the corrections will be done together with the other necessary corrections.

<!> 13 Oct 2011

All umo and vmo data sets from all experiments have been deleted.

<!> 12 Oct 2011

Due to a time axis error the data sets ta_day_MPI-ESM-LR_rcp85_r3i1p1_20490101-20491231.nc and va_day_MPI-ESM-LR_rcp26_r3i1p1_20450101-20451231.nc had to be corrected for experimentsrcp85_r3i1p1-LR and rcp26_r3i1p1-LR.

<!> 06 Oct 2011

evspsbl (wrong flux direction), the lev dimension for cl, cli, clw, and orog are corrected in experiments piControl-P, 1pctCO2-P, abrupt4xCO2-P, lgm-P and midHolocene-P.

<!> 05 Oct 2011

The lev dimension for atomic data sets cl, cli, clw are corrected in the decadal*-LR experiments.

<!> 04 Oct 2011

evspsbl (wrong flux direction) is corrected in the decadal*-LR experiments.

<!> 19 Sep 2011

evspsbl (wrong flux direction) is corrected for experiments piControl-LR, historical*-LR, rcp26*-LR, rcp45*-LR, rcp85-LR, amip*-LR, sstClim*-LR, sstClim4xCO2*-LR,1pctCO2*-LR,abrupt4xCO2*-LR, esmFdbk1*-LR, and esmFixClim1.

<!> 19 Sep 2011

The lev dimension for the variables cl, cli, clw is corrected now for the experiments piControl*-LR, historical*-LR, rcp26*-LR, rcp45*-LR, rcp85*-LR, amip*-LR, sstClim*-LR, sstClim4xCO2*-LR, 1pctCO2*-LR, abrupt4xCO2*-LR, esmFdbk1*-LR and esmFixClim1*-LR.

<!> 16 Sep 2011

orog is replaced (by code 129 of ECHAM66 divided by 9.80665) in order to account for the spectral nature of ECHAM6's dynamical core for experiments piControl*-LR, historical*-LR, rcp26*-LR, rcp45*-LR, rcp85*-LR, amip*-LR, sstClim*-LR, sstClim4xCO2*-LR, 1pctCO2*-LR, abrupt4xCO2*-LR, esmFdbk1*-LR and esmFixClim1*-LR.

<!> 13 - 30 Sep

Experiment historical_r1i1p1-MR (wrong setting of YR_PERP in ECHAM6's namelist) is rerun. (The model output of the old experiment (with wrong YR_PERP=1850) can be found in blizzard:/work/ik0555/cmip5/experiments/historical_r1i1p1-MR_DoNotUse.)

<!> 02 Sep 2011

The Fram Strait value (index 10) of mfo is set to 'undef' in all *-LR experiments.

<!> 02 Sep 2011

All 6hr atomic data sets of all experiments have been renamed (no minutes end date).

<!> 22 - 23 Aug 2011

All atomic data sets of esmFdbk1_r1i1p1-LR are regenerated. co2mass is output as time series instead of as climatology.

<!> 9 - 16 Aug 2011 midHolocene_r1i1p1-P was rerun due to a problem with the orbit formulation. The calender time span is now 1850-1949.

<!> 4 - 13 Aug 2011

lgm_r1i1p1-P was rerun due to a problem with the orbit formulation. The calender time span is now 1850-1949.

<!> 13 - 14 Aug 2011

All atomic data sets of esmFixClim1_r1i1p1-LR are regenerated.

<!> 10 Aug 2011

The CMIP5 'references' metadata value is corrected for 1pctCO2-r1i1p1-P, abrupt4xCO2-r1i1p1-P and piControl-r1i1p1-P.

<!> 28 Jul 2011

The CMIP5 'references' metadata value is corrected for amip_r1i1p1-LR and historical_r1i1p1-LR

<!> 26 Jul 2011

The CMIP5 'references' metadata value (NetCDF attribute) is truncated. All data sets are concerned. This will be corrected asap.

<!> 24 - 25 Jul 2011

The values of the 'co2mass' atomic data sets were too small by a factor of 36 (-R, -P) or 48 (-MR). They have been recalculated.

<!> 21 Jul 2011

All atomic data sets amip_r3i1p1-LR have been recalculated since files are lost.

<!> 21 Jul 2011

baresoilFrac of amip_r[1,2]i1p1-LR has been recalculated since it should be provided for all years and all months instead of for the 1st year only (though it is invariable in time).

<!> 20 - 21 Jul 2011

cropFrac, pastureFrac, fLuc of experiments esmFdbk1_r1i1p1-LR and esmFixClim1_r1i1p1-LR is being recalculated since it should be provided for all years and all months of all coupled experiments instead of for the 1st year only.

<!> 20 Jul 2011

cropFrac, pastureFrac, fLuc of experiment piControl_r1i1p1-LR is being recalculated since it should be provided for all years and all months of all coupled experiments instead of for the 1st year only.

<!> 20 Jul 2011

The CMIP5 variables cropFrac, pastureFrac, fLuc of experiment abrup4xCO2_r1i1p1-LR is being recalculated since it should be provided for all years and all months of all coupled experiments instead of for the 1st year only.

<!> 19 - 20 Jul 2011 landCoverFrac, cropFrac, pastureFrac, treeFrac, fLuc, grassFrac, shrubFrac, c3PftFrac, c4PftFrac, baresoilFrac of experiment piControl_r1i1p1-P is recalculated since it should be provided for all years and all months of all coupled experiments instead of for the 1st year only.

<!> 19 Jul 2011 fLuc of experiment 1pctCO2_r1i1p1-LR is being recalculated since it should be provided for all years and all months of all coupled experiments instead of for the 1st year only.

<!> 9 - 19 Jul 2011

rcp45_r1i1p1-LR

started Sun. 17 Jul 16:48

finished Tue. 19 Jul 02:46

rcp85_r1i1p1-LR

started Thur. 14 Jul 18:20

finished Sun. 17 Jul 21:44

rcp26_r1i1p1-LR

started Thur. 17 Jul 16:57

finished Tue. 19 Jul 03:18

rcp45_r2i1p1-LR

started Sa. 9 Jul 10:54

(blizzard crashed) finished Mo. 11 Jul 10:01

rcp85_r2i1p1-LR

started Sa. 9 Jul 12:01

(blizzard crashed) finished Mo. 11 Jul 07:40

rcp26_r2i1p1-LR

started Sa. 9 Jul 12:27

(blizzard crashed) finished Mo. 11 Jul 07:14

rcp45_r3i1p1-LR

started Tues. 12 Jul 20:30

finished Wed. 13 Jul 10:23

rcp85_r3i1p1-LR

started Tues. 12 Jul 20:30

finished Wed. 13 Jul 10:19

rcp26_r3i1p1-LR

started Tues. 12 Jul 20:30

finished Wed. 13 Jul 10:10

The value of base time for the rcpXY_r?i1p1-LR experiments has been corrected to 1850 (from 2006). Values of the atomic data sets are not concerned.

<!> 24 Jun -16 Jul 2011

rcp45_r1i1p1-LR

started Fri. 24 Jun 17:32

year 2100 finished Fri. 1 Jul 16:34

year 2300 finished Fri. 15 Jul 10:42

rcp45_r1i1p1-LR

started Fri. 24 Jun 17:32

year 2100 finished Fri. 1 Jul 16:34

year 2300 finished Fri. 15 Jul 10:42

rcp85_r1i1p1-LR

started Fri. 24 Jun 17:33

year 2100 finished Thur. 30 Jun 13:00

year 2300 finished Thur. 14 Jul 07:42

rcp26_r1i1p1-LR

started Fri. 24 Jun 17:32

year 2100 finished Sa. 2 Jul 01:52

year 2300 finished Sa. 16 Jul 15:28

The branch_time values have been corrected for experiments historical_r1i1p1-LR, historical_r2i1p1-LR, historical_r3i1p1-LR, abrupt4xCO2_r1i1p1-LR, 1pctCO2_r1i1p1-LR, rcp26_r2i1p1-LR, rcp45_r2i1p1-LR, rcp85_r2i1p1-LR, esmFdbk1_r1i1p1-LR, and esmFixClim_r1i1p1. Values of the atomic data sets are not concerned.

MadWiki: WarningsDataUsers (last edited 2012-09-26 14:33:51 by StephanieLegutke)