We will frequently have the use case that we want to compare two simulations that do not cover the same time period. One example that I'm currently facing are Tidos tuning runs for AWI-CM3 TCO319-DART.
The first run is 40 years with one set of tuning parameters, the second run is branched off from the first and only 6 years long. Both runs start in 1990. I want to compare the last 3 years of the first simulation (years 2033-2035) with the last three years of the second simulation (years 1993-1995). Sea ice has a short memory, so a few years of spinup should already show results. However right now, I can only select one set of start_year and end_year which is then applied to all input_names.
Is there a general solution for this? I see that we are currently making extensive use of the dates and times that are contained within the NetCDF files themselves. This makes manipulation of the data later tricky. Is this a feature we need/want?
We will frequently have the use case that we want to compare two simulations that do not cover the same time period. One example that I'm currently facing are Tidos tuning runs for AWI-CM3 TCO319-DART.
The first run is 40 years with one set of tuning parameters, the second run is branched off from the first and only 6 years long. Both runs start in 1990. I want to compare the last 3 years of the first simulation (years 2033-2035) with the last three years of the second simulation (years 1993-1995). Sea ice has a short memory, so a few years of spinup should already show results. However right now, I can only select one set of
start_yearandend_yearwhich is then applied to allinput_names.Is there a general solution for this? I see that we are currently making extensive use of the dates and times that are contained within the NetCDF files themselves. This makes manipulation of the data later tricky. Is this a feature we need/want?