search

pytroll / satpy

Python package for earth-observing satellite data processing
http://satpy.readthedocs.org/en/latest/
GNU General Public License v3.0
1.07k stars 297 forks source link

Issues about fog composites #2810

Open augustou18 opened 5 months ago

augustou18 commented 5 months ago

I would like to ask what is the physical significance of the results generated by ahi/hsd's fog composites. The 3D matrix does not seem to confirm the exact distribution of this fog. Is there something inappropriate in our approach？

djhoese commented 5 months ago

@augustou18 Could you please provide the code you are running to generate the data/image you're looking at and include the exact composite name you are loading?

augustou18 commented 5 months ago

@djhoese Hi, djhoese. The composite output fog is a three-dimensional(3 bands) matrix, what is its physical meaning?

djhoese commented 5 months ago

What fog composite are you talking about? Please answer my questions above. My understanding is you are providing Satpy data as AHI HSD files and generating a "fog" composite. Could you please show me the code you are running to do this?

ZitaTan commented 5 months ago

@djhoese Hi, djhoese. I’m using AHI HSD files and generating a "fog" composite too. I get a "fog" image using "scn.load(['fog'])", Could you please tell me what the colors on the picture represent? Thanks. tmpqz87hbe6

mraspaud commented 5 months ago

@ZitaTan I think the fog composite in satpy is the 24h microphysics, which you can read more about here https://resources.eumetrain.org/rgb_quick_guides/quick_guides/24MicroRGB.pdf

djhoese commented 5 months ago

Note that there is always a chance our recipe was implemented poorly so please let us know if something looks off. One thing I noticed is that the AHI configuration specifically has difference channels used for the 10.8um in the two differences. See:

https://github.com/pytroll/satpy/blob/394add5602e3eb052e24febd79839bf32254d8b5/satpy/etc/composites/ahi.yaml#L123-L135

Instead of using B13 for both differences, it uses (B15 - B13) for red and (B14 - B11) for green (instead of B13 - B11). I think this is expected for AHI given the goals of the composite and the properties of those bands. Night microphysics is configured similarly. I'm not a scientist, but this is what I remember from past discussions.