How to extract variable data for given set of coordinates?

I'm currently working on Argo data. I used ncdisp to check the required variables and I need to extract Argo Temperature "TEMP" variable data for the coordinates: longitudes 60 E to 110 E and latitudes 5N to 25N.

I used ncread an checked that I need to get the TEMP data for longitudes 60.5 to 100.5 and latitudes 5.5 TO 25.5 since that is how the values are recorded in the "LONGITUDE" and "LATITUDE" variables. I would like some help on how to go about this and what codes to use.

Here's the link to the files for reference: http://apdrc.soest.hawaii.edu/projects/Argo/data/gridded/On_standard_levels/Monthly_mean/1x1/2016/m02/index.html > click on the third link (netCDF Data file)

The dimensions of the file are:

dimensions:

           LONGITUDE = 360

           LATITUDE  = 180

           LEVEL     = 27

Answer 1

[KSSV]

ncfile = 'ArgoData.nc'; 
lon = ncread(ncfile,'LONGITUDE') ; nx = length(lon) ; dx = min(diff(lon)) ;
lat = ncread(ncfile,'LATITUDE') ; ny = length(lat) ; dy = min(diff(lat)) ;
[X,Y] = meshgrid(lon,lat) ;
level = 27 ;
T = ncread(ncfile,'TEMP',[1,1,level],[nx,ny,1]) ;

xi = 60.5:dx:100.5 ;
yi = 5.5:dy:25.5 ;
[Xi,Yi] = meshgrid(xi,yi) ;
Ti = interp2(X,Y,T',Xi,Yi)' ;

figure(1)
pcolor(X,Y,T') ; shading interp; 
figure(2)
pcolor(Xi,Yi,Ti') ; shading interp;

[Keegan Carvalho]

Thank you KSSV. I have a question:

T = ncread(ncfile,'TEMP',[1,1,level],[nx,ny,1])

What does [1,1,level] & [nx,ny,1] mean?

Is [1,1,level] mean the first lat and lon where temperature values are extracted for all levels?

I am new to using netcdf files so I'm a beginner here :) Hope you don't mind

[KSSV]

[1 1 level] means extract the chunk at given level i.e 27 here.

[nx ny 1] means extract the entire matrix...of size nx*ny.

Read about ncread in the documentation.

[Keegan Carvalho]

Thanks KSSV. I wanted to find the coordinates (LATITUDE,LONGITUDE) and LEVEL at which TEMP~20°C. So I thought I could set LEVEL=1, 2.. 27 and find 20°C individually since the Ti file (generated using your code above) is 41x21.

Is there a faster and efficient way to do this with codes? Or should I go ahead with the above plan?