Make a plt from a 4-D multi array variable from nc file

I have a nc file with:

Dimensions:

time = 8

depth = 48

latitude = 109

longitude = 25

and make this:

u='global-reanalysis-phy-001-030-daily_1591804259623.nc'
ncdisp(u)
east = ncread(u,'uo'); 
lat = ncread(u,'latitude'); 
lon = ncread(u,'longitude');
time=ncread(u, 'time');
depth=ncread(u, 'depth');
dtime = datetime(1950, 1, 1, time, 0, 0);
jd=juliandate(dtime);
plot(lat, east(:,:,15,1));

east is a 4-D 8371200 double

But I want a plot (surfc) in time, this 8 days, latitude, one depth (lets say 48, the last) and the east variable that is a velocity m/s.

Any idea? What I should do with east variable? squeeze, quiver, etc realy dont know or find solution

Thanks.

[KSSV]

OP commented:

the plot(lat, east(:,:,15,1)) is working but only for one day if I change 1 for : then the error

Data cannot have more than 2 dimensions.

Answer 1

[KSSV]

ncfile = "global-reanalysis-phy-001-030-daily_1591804259623.nc";

time = ncread(ncfile,'time') ; 
depth = ncread(ncfile,'depth') ; 
lat = ncread(ncfile,'latitude') ;
lon = ncread(ncfile,'longitude') ; 
u = ncread(ncfile,'uo') ; 
% plots 
for i = 1:length(time)  % time loop
  for j = 1:length(depth) % depth loop
    z = u(:,:,j,i)' ;  % extract the matrix here 
    pcolor(lon,lat,z) ;
    colorbar
    shading interp
    xlabel('lon')
    ylabel('lat')
    title(sprintf('time = %d, depth = %d',i,j))
    drawnow
 end
end

[Pepe Grillo]

Thanks for you answer, I try but this came:

>> surf(lon,lat,east(:,:,1))

Error using surf (line 74)

Data dimensions must agree.

>> pcolor(lon,lat,east(:,:,2))

Error using pcolor (line 59)

Matrix dimensions must agree.

I make the squeeze .

Any idea?

[KSSV]

try transposing the matrix..

surf(lon,lat,east(:,:,1)')

If not working share the dimensions of each variable here..

[Pepe Grillo]

Thanks!

Here is the file

https://drive.google.com/drive/folders/1P7d-Qnt2lvMpjzrk3Mo4FLLegQ7lI_rx?usp=sharing

What i want is a plot in 2D

x=time

y=latitude

z=east

at specific depth.

[KSSV]

Edited the answer.

[Pepe Grillo]

ok! I try it runs with out error but the plot is empty.

[Pepe Grillo]

the z matrix is NaN

[KSSV]

All the matrices are not NaN....when depth is increasing....data is having NaN's....see the plot, it is in a loop..it is plotting all the values..

[Pepe Grillo]

I add this to your answer:

nx = length(lon);

ny = length(lat);

nd=length(depth);

nt=length(time);

that I saw you make in other question and works but is like running the time and depths like animation

but still the cordinates are lat and long

[KSSV]

Okay..I got it..you want to plot for specific depth, all the time data...

For this case, you need to fix the range of lat and lon...

[Pepe Grillo]

ok! yes I understand so how can I make the plot time vs lat with u as 3 variable

[KSSV]

ncfile = "global-reanalysis-phy-001-030-daily_1591804259623.nc";

time = ncread(ncfile,'time') ; 
depth = ncread(ncfile,'depth') ; 
lat = ncread(ncfile,'latitude') ;
lon = ncread(ncfile,'longitude') ; 
u = ncread(ncfile,'uo') ; 

% plots 
for i = 1:length(lon)
  ud = squeeze(u(:,i,1,:)) ; 
  pcolor(lon,time,ud')
  shading interp ;
  colorbar
  title(sprintf("depth = 1, lat rows = %d",i)) ; 
  drawnow
  pause
end

[KSSV]

This is called Hovmoller diagrams ..am I right?

[Pepe Grillo]

[Pepe Grillo]

ok! thanks I think I can change to axe x time and to y latitude

great!

[Pepe Grillo]

Yes Hovmoller! I just start with this this week and try with few data now came the big things...

so what is doing: ud = squeeze(u(:,i,1,:))

thanks!

[KSSV]

You can try it your self......check the dimensions before and after squeeze.

squeeze : it will remove the unwanted extra dimensions

[Pepe Grillo]

yes! thanks a lot!