Hi - The purpose of the "timestep" technique, defining and adding a time-axis variable to your input variables, is to add time information to a variable that does not have a time axis already defined. If the variables in your datasets already have time as a dimension in their grids, then you don't need to do that step. If the variable in file1 has a time axis with t=1-jan-1980, and the one in file2 has data at 2-jan-1980, and if their time axes are consistent with each other (the same units and time origin), then you can just append each to your output file. use file1.cdf save/CLOBBER/file=concat.cdf temperature cancel data/all use file2.cdf save/file=concat.cdf/append temperature cancel data/all ... If this is not what you need to do, then we need more informatino to help you out. What does the grid of your variables look like: yes? use file1 yes? sho data yes? show grid temperature yes? show axis/t `temperature,return=taxis` and likewise for file2. Ansley On 2/1/2012 9:31 PM, Geetha R wrote: Hi Sir, |