How to relate the model time to our common calendar?

[wl@eimb_ru]

I need to compare observations and modeling results with making predictions about future changes.

Therefore I need to relate somehow the time variable in the output ncetCDF files with the time when the observations were made.

Comparing long-term climatological means doesn't fully satisfy my needs, because I would like to obtain trends from the model and apply them to the observational results.

I have read the discussion http://bb.cgd.ucar.edu/showthread.php?t=96 but it didn't give me the solution.

 

[rneale]

To compare observed dates with model runs you need to choose 20th century CCSM runs. Most CCSM 'control' simulations are do not involve real dates per se.
I believe there are 20th century runs available on the Earth System Grid which include observed trends in CO2, aerosols, volcanic forcing etc. The filename dates refer to real dates.

 

[wl@eimb_ru]

Thank you, Richard.

Am I correct in thinking, that the dates in the runs b30.030a-e, b30.036a-e, b30.040a-e, etc (other SRES, commitment, etc runs) were simply reassigned to 1870-1999, 2000-2099, and 2100-2199 to synchronize the model results with our common calendar?

To be more specific, consider high resolution runs.
I have found that runs b30.030a-e were branched from different years of the preindustrial control run (b30.020).
Suppose, I am a modeling technician, making such a branch in order to get the b30.030a run.
I take the b30.020 model restart files, created on the 360th simulation year, assign the time variable the value corresponding the 1870-01-01, and create a new branch case named b30.030a...
Or I make a new branch case, and rename the history files to have that dates in file names.
Then I consider those dates same way as our real calendar.

Am I right?

Also, could you point me to a description, why the branches from the b30.020 were made at those certain years (360, 380, 400, 420, 440)?

 

[wl@eimb_ru]

Was the CCSM history of the runs b30.030a-e, b30.036a-e, b30.040a-e, b30.041a-e, b30.042a-e, available from the EarthSystem Grid, subjected to any additional transformations?

In other words, is it possible to produce seamless consistent time series from history files of runs b30.030a and b30.036a?
b30.030a, b30.041a, and b30.045a?

 

[strandwg]

Was the CCSM history of the runs b30.030a-e, b30.036a-e, b30.040a-e, b30.041a-e, b30.042a-e, available from the EarthSystem Grid, subjected to any additional transformations?

In other words, is it possible to produce seamless consistent time series from history files of runs b30.030a and b30.036a?
b30.030a, b30.041a, and b30.045a?

Yes - the last date for b30.030a is 12-31-1999, the first date for the future runs is 01-01-2000.