Plot netcdf datas through time

Hello,

I'm currently working on an Antarctica project and I have 40go of netcdf data. I need to plot those datas in differents graphs but i can't share the datas as the project is still running and anyway the files are too important in size to be shared. I will try my best to explain.

In my netcdf file I have a variable named 'sisali' representing the salinity evolution in Amundsen Sea every month of 1973. This variable is defined as :

"sisali

Size: 231x190x12

Dimensions: x,y,time_counter

Datatype: single

Attributes:

long_name = 'Mean ice salinity'

units = '1e-3'

online_operation = 'average'

interval_operation = '1800 s'

interval_write = '1 month'

cell_methods = 'time: mean (interval: 1800 s)'

_FillValue = 1.000000020040877e+20

missing_value = 1.000000020040877e+20

coordinates = 'time_centered nav_lon nav_lat' "

The problem is that, my coordinates are in 2 dimensions, and the salinity datas are in 3 dimension because of the time scale. When I try to plot it the first problem occuring is the size difference and then nothing happen.

Do you guys have any idea how to plot this salinity evolution during this year in the Amundsen Sea. Or how to separate the coordinates datas so I can plot just the salinity through time ?

Thanks for your help!

Answer 1

[Chad Greene]

The answer depends on exactly what you want to display. Do you want 12 different maps, to show salinity distribution for each month? If so, this would do it:

for k = 1:12 
    subplot(3,4,k)
    imagesc(x,y,sisali(:,:,k))
    axis xy 
    title(['Month ',num2str(k)])
    caxis([lowerlimit upperlimit]) % for consistency across all plots
    cmocean haline % optional colormap
end

Or do you simply want a single map showing average salinity for the year? If so,

imagesc(x,y,mean(sisali,3))
axis xy 
cmocean haline

You could similarly subtract mean salinity from each month to show monthly anomalies, like this.

for k = 1:12 
    subplot(3,4,k)
    imagesc(x,y,sisali(:,:,k)-mean(sisali,3))
    axis xy 
    title(['Month ',num2str(k)])
    caxis([-1 1]*limit) % for consistency across all plots
    cmocean balance % divergent colormap
end

Instead of 12 separate plots, you could make an animation with my gif function like this:

h = imagesc(x,y,sisali(:,:,k));
axis xy
caxis([lowerlimit upperlimit])
cmocean haline
title('Month 1')
gif('amundsen_sea_salinity.gif')
for k = 2:12
    h.CData = sisali(:,:,k);
    gif
end

Or if you simply want to plot a time series of the average salinity in a given region, you could create a mask for that region that is true for any grid cells you're interested in. An easy way to do that might be to grid up your 1D arrays of coordinates and then use my isopenocean function to determine which of those grid cells correspond to ocean:

[X,Y] = meshgrid(x,y); 
mask = isopenocean(X,Y);

Then use my local function to average all the grid cells in that region, for each month. Then you'll end up with a time series of 12 values, containing the mean salinity within the masked region for each month of the year:

sal = local(sisali,mask);
plot(1:12,sal)
xlabel 'month of year'
ylabel 'average salinity within region of interest'