Replace values in grid data (3D) to NaN, like another (3D) have NaN

Hi to all

I have grid data VWind=300*16*21 and

SST=300*16*21 have NaN values

How i can to the ​​replace NaN value at VWind data to the same location in SST data

Thanks in advance

Answer 1

[Sulaymon Eshkabilov]

It can be done very easily with a logical indexing opt: e.g.:

Index = isnan(SST);
VWind(Index)=NaN;

Now you can compare and verify the elements of SST and VWind using sum() operator, for instance.

Good luck.

[nada]

Thank you ..

[James Karden]

How could I get this to loop over each time step (let's say over 40 years) if I have a matrix of lat x lon x time?

[Walter Roberson]

If SST is 3D then you would not normally work by timestep: you would normally just do

VWind(isnan(SST)) = NaN;

If you need to work timestep by timestep for some reason, then see https://www.mathworks.com/matlabcentral/answers/731168-extract-specific-data-from-3d-matrix#answer_610468 for some ways in which you can extract information slice by slice.

Answer 2

[Sudheer Bhimireddy]

I don't know if I understood your question properly, but here goes:

If you have NaN values in SST matrix and would like to replace VWind values to NaN wherever SST(i,j,k)=NaN:

VWind(isnan(SST)) = nan;

Similarly if VWind already has NaNs, to make corresponding indices of SST to NaN:

SST(isnan(VWind)) = nan;

If you run both these commands, then both your matrices will have same number of NaN values at same locations.

Hope this helps.

[nada]

very thanks

[James Karden]

How could I get this to loop over each time step (let's say over 40 years) if I have a matrix of lat x lon x time?