Without knowing more about your study, or knowing anything about the stations themselves, I'll make a couple of suggestions:

- If one has a higher % data coverage than the others, that would be the easiest justification.

- If one, for any number of reasons, is a better representation of your entire domain geographically (e.g. station x is on a landcover more representative of the entire domain than station y or z).

- If one has a longer period of record than the others, that would also be an easy justification.

- If one station has more data types that may suggest that it is a "better" station than the others.

Again, these are all just ideas. If all three of them are truly acceptable and comparable stations, then having an exceptionally strong justification may not be 100% necessary. Just something along the lines of "All three stations would have been acceptable, however only station x was chosen to limit dataset size" or something to that effect.

Thank you Walker-Skeeter for the answer. This will be very helpful

When dealing with meteorological data, one way to justify it is according to what is established by the WMO, that is, select the station with the longest length of the historical record, or the one with the greatest number of variables recorded or measured that interest you in your study and that have quality data. This could be a way to justify it.

greetings and success

I suggest the following

Proximinity to the area of interest

Type and exposure of the instruments

Time since the last Quality Assurance report and its content

Period of record

Date of the last calibration

Owner of site (Federal, state and local gov't best)

I would examine the data to assess if it looks reasonable.

Academic resources on atmospheric dispersion are provided on the project references:

SINGLE PHASE AND MULTIPHASE TURBULENT FLOWS (SMTF) IN NATURE AND ENGINEERING APPLICATIONS | Jamel Chahed | 3 publications | Research Project (researchgate.net)