Questions about data filtering and regridding

[JairoVS]

Hello! NCtoolkit, it's been a lifesaver for my analysis. I appreciate the efforts of the developers.

I am working with satellite data of column densities (mol/cm2) of air pollutants, and I have some questions about data processing using nctoolkit. I need to set as_missing all the negatives and all values above 1e17, but at the same time, I want to fill all the empty grids using fill_na.

1. Is it better to fill the na's first and then do the as_missing, or is it better to do it the opposite way?

2. How can I use the as_missing function to filter all the negatives? I tried using as_missing (<0), but it did not work

3. What would be a good number for the fill_na()? What happens if several contiguous grids are empty when using fill_na?

Thank you in advance for your guidance.

[robertjwilson]

Thanks.

1. Is it better to fill the na's first and then do the as_missing, or is it better to do it the opposite way?
   I would set the missing values first.

2. How can I use the as_missing function to filter all the negatives? I tried using as_missing (<0), but it did not work
   Set a suitable range to a missing value:
   ds.as_missing([-1e90, 0])

3. What would be a good number for the fill_na()? What happens if several contiguous grids are empty when using fill_na?
   This is somewhat data dependent. The method is distance weighted and n defines how many neighbours to use. Typically, 8 covers most use cases, as you have 8 cells around each cell. n=1 is a simple nearest-neighbour and often that is fine. It really depends on what kind of gaps you have. Sometimes I need to infill coasts, so in that case would just do n=5.

[JairoVS]

Thank you very much for your answer! One more question related to the same topic, if I have several variables in the same NetCDF and want to apply as_missing, is it possible to choose only one of the variables? Or can the method only be applied to all of the variables in the file?

[robertjwilson]

Right now you can only apply it to all variables. If you only want to apply it to one variable you would need to create a separate dataset including only it and then merge it back in.

[JairoVS]

Understand! I appreciate your guidance.

[JairoVS]

Hello Robert! Again, I want to ask you some questions regarding the vertical_mean function. Is it better to open a new discussion, or can I ask it here? Or is it better if I contact you through your email? Many thanks for considering my request.

[robertjwilson]

Just open a new discussion