Project

General

Profile

using mergetime for files with more than one time dimension

Added by Mika Lanzky about 5 years ago

I still consider myself a cdo newbie: I have used it succesfully for different stuff, but I find the documentation hard to understand for me.
I realise CDO is climate data operators, and my data is not actually climate data, so it is quite probable cdo is not the right tool to use here.

I have netCDF data files that contain data in different timely resolutions (think 1second, 1min, 1hour).
The data is 1D data, as in, regular xy-data, no grid or lat/long or anything.
So the data files has different time dimensions, and of course also the variable data for this different time dimensions (think temperature in averaged to 1sec, 1min and 1hour).

These data files comes in one file pr day. For ease of further data analysis, I would like to merge these files into one file. Usually, for my purposes, the hourly data is also enough, so my ideal output would be one data file with all the hourly resolved variables.

I tried using mergetime, but this seem to do it on the last stated time dimension (stated as when I inspect the file using ncdump), which in this case was 1 second. It also seem to stop before it is done, possible owing to the 2GB maximum problem, though I did enable the options that should get around this.

So what I am asking, is there are way to tell mergetime which time dimension it should be merging by?
Does mergetime have a dimension parameter?


Replies (1)

RE: using mergetime for files with more than one time dimension - Added by Mika Lanzky about 5 years ago

Correction to above:
If I inspect the file with cdo sinfov, it clearly shows the "Time coordinate" is the one in seconds.
The other time coordinates/variables are listed under "Grid coordinates".

So I assume that mergetime acts on the given "Time coordinate"?
Then of course the follow-up question is how to make one of the other time dimensions the time coordinate, but I would think that is beyond cdo's intended capabilities.

    (1-1/1)