After Mosaic data loss spectral information

I used sentinel 2A 1C data and used the sen2cor tool to calibrate the atmosphere(20m), then resampled (20m)the data and then mosaic the data and exported it to the BEAM-DIMAP format. Finally, I found that Mosaic data lost spectral information, what caused it?Are there any errors in my data processing methods?(Data processing flow:atmospheric correction—— resampling——mosaic(BEAM-DIMAP))
I hope you can make some Suggestions,thank you!

:joy:Can anyone give me some advice?

This is because the mosaic output can be generated by any arbitrary expression and then the wavelength does often not apply anymore.
However, it is a known issue (SNAP-745). It could be checked if the output is generated by a single band and then the wavelength could be preserved.

You can edit the wavelength manually within SNAP Desktop. That’s fine if you only want to edit a few products but if you want to do the processing on many products it is not feasible anymore.

In this case, you can create a graph XML file which first does the mosaic and then, as the second step, you can use the BandMaths operator to add the wavelength to the bands.

Thank you very much for your advice!
Recently, I installed SNAP latest version (6.0) and used the multi-size Mosaic tool to make data mosaic to solve this problem.
However, processing time is particularly slow. I’ve found that the output data type defaults to float, which cannot be changed?

No this cannot be changed.