Want to split a 4D matrix based on the different dimensions

I have a 4D matrix for temperature

lon x lat x level x time (144 x 73 x 17 x 886)

I want to split it into multiple time series such that level is 1000 (first value in the level array) for each lon x lat pair.

for lon[]

for lat[]

for time[]

timeseries[i,j] = temp['lon' 'lat' 0 'time'];

i = i + 1;

end

j = j + 1;

end

end

This is the code I tried to use.

I would also like find the correlation between each time series and convert it into a complex network where the nodes are each lon x lat pair with temperature as its value and connect them based on the correlation value. I would also like to obtain the link density, clusterring coefficient and correlation path length for the network.

The data is from a netcdf file.