How to read a netcdf file and convert it to time table

Hi guys,

I have a climate data file in netsdf format containing long, lat, time and temperature. The issue is that temperature has been shown for 9 lang and 5 lat during 3 years.

As can be observed. I need to extract t2m variable during the time for each lat,long(9*5=45 different point).

The question is how can I do that?

Thanks in advsnce

Answer 1

[Peter Perkins]

Hamed, it's not currently possible to create a timetable that is 3-D (lat-by-lon-by-time). One thing you can do is make a 45*8759x3 timetable, with variables for lat, lon, and temp. Another possibility is to make an 8759x45 timetable with one var for each grid point. Or an 8759x9 timetable each of whose variables is itself 8759x5.

Hard to say which of these might work for you, it depends on what you need to do next.

[Hamed Hedayatnia]

Thanks but how can I remove one dimension and keep the other? with reshape?

[Peter Perkins]

I think in all cases you first need permute to turn 9x5x8759 into 8759x9x5.

You would use ndgrid to create the repeated lat/lon for the 45*8759x3 option, and yes, reshape on the temperatures. Reshape and and array2table would get you the 8759x45 version. Subscripting and the table constructor would get you the 8759x9 version.

[Hamed Hedayatnia]

How can I create an 8759x45 timetable with one var for each grid point?