Issues with fpgpy
5 views
Skip to first unread message
maatouk nadia
May 3, 2026, 8:52:31 AM (11 days ago) May 3
to pytroll
Hello,
I run https://github.com/pytroll/fogpy/blob/main/fogpy/algorithms.py. I transformed the channels of msg seviri netcdf to brillaince temperature and reflectance. I run algorithms but I had wrong cloud mask comparing to eumetsat view. I don't know why! help pls
Holl Gerrit
May 8, 2026, 9:25:09 AM (6 days ago) May 8
to pytroll, maato...@gmail.com
Hello Nadia Maatouk,
can you show us the exact code you have tried to run? algorithms.py is a module within fogpy and not the usual "entrance point". What do you mean by wrong cloud mask? Do you mean you get a fog mask, but that it is not the fog mask you are expecting?
Are you using daytime or night time data?
What cloud microphysics (CMICI) are you using? Calculated with NWCSAF GEO or coming from a different source?
Fogpy should work, in particular for SEVIRI, but as a note of warning, it has not been maintained for some years and nobody is using it operationally, so it is possible that some parts are not up to date with the newest satpy.
regards,
Gerrit.
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
From: pyt...@googlegroups.com <pyt...@googlegroups.com> on behalf of maatouk nadia <maato...@gmail.com>
Sent: 02 May 2026 11:17:36
To: pytroll
Subject: [pytroll] Issues with fpgpy
Hello,
I run https://github.com/pytroll/fogpy/blob/main/fogpy/algorithms.py. I transformed the channels of msg seviri netcdf to brillaince temperature and reflectance. I run algorithms but I had wrong cloud mask comparing to eumetsat view. I don't know why! help pls
--
You received this message because you are subscribed to the Google Groups "pytroll" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pytroll+u...@googlegroups.com<mailto:pytroll+u...@googlegroups.com>.
To view this discussion, visit https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com<https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com?utm_medium=email&utm_source=footer>.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
Der Originalabsender dieser E-Mail ist: pytroll+bncBDU3R5M2XA...@googlegroups.com. Wenn Ihnen dieser unbekannt ist, so sollten Sie die enthaltenen Links nicht anklicken.
can you show us the exact code you have tried to run? algorithms.py is a module within fogpy and not the usual "entrance point". What do you mean by wrong cloud mask? Do you mean you get a fog mask, but that it is not the fog mask you are expecting?
Are you using daytime or night time data?
What cloud microphysics (CMICI) are you using? Calculated with NWCSAF GEO or coming from a different source?
Fogpy should work, in particular for SEVIRI, but as a note of warning, it has not been maintained for some years and nobody is using it operationally, so it is possible that some parts are not up to date with the newest satpy.
regards,
Gerrit.
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
From: pyt...@googlegroups.com <pyt...@googlegroups.com> on behalf of maatouk nadia <maato...@gmail.com>
Sent: 02 May 2026 11:17:36
To: pytroll
Subject: [pytroll] Issues with fpgpy
Hello,
I run https://github.com/pytroll/fogpy/blob/main/fogpy/algorithms.py. I transformed the channels of msg seviri netcdf to brillaince temperature and reflectance. I run algorithms but I had wrong cloud mask comparing to eumetsat view. I don't know why! help pls
You received this message because you are subscribed to the Google Groups "pytroll" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pytroll+u...@googlegroups.com<mailto:pytroll+u...@googlegroups.com>.
To view this discussion, visit https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com<https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com?utm_medium=email&utm_source=footer>.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
Der Originalabsender dieser E-Mail ist: pytroll+bncBDU3R5M2XA...@googlegroups.com. Wenn Ihnen dieser unbekannt ist, so sollten Sie die enthaltenen Links nicht anklicken.
Holl Gerrit
May 8, 2026, 10:52:51 AM (6 days ago) May 8
to maatouk nadia, pyt...@googlegroups.com
Hi Nadia,
as a disclaimer, it's been years since I looked at fogpy, so my mind is not very fresh on the topic, sorry. I don't remember what step 8 is. The daytime algorithm needs CMIC data. The night time algorithm uses IR39/IR108 and probably does not work during the day. It's also not very advanced and really only there as a filler for when the daytime algorithm cannot be used.
I would recommend getting some CMIC data. I think there are some in the data store, or otherwise they can be calculated with the NWCSAF GEO software or obtained from CMSAF. Or you can try if the algorithm works at night.
The algorithm implemented by fogpy is by now quite old, and meanwhile superior algorithms (I think) have been developed based on AI/ML, but those are not implemented in fogpy (fog as a product always drops down in the list in priorities at DWD, sadly). One of the reasons why I stopped exploring it further is that CMIC is only available at relatively high SZA (I think <78°), whereas the night time algorithm can only be used if there is no reflectance contamination in IR 3.9 when the Sun is safely down below the horizon. This means the twilight area, including the entire month of January in northern Germany, is missing (I guess one could try to calculate the thermal part of IR 3.9; I haven't tried), but this is the most important time of day for fog. That is a limitation on the algorithms in fogpy that some newer 24/7 fog algorithms manage to avoid.
Regards,
Sent: 08 May 2026 16:40:52
To: Holl Gerrit
Subject: Re: [pytroll] Issues with fpgpy
Thank you a lot for your response. I am running dayfoglowstratus algorithm. As I don't have CMIC data, I retrieve the mask after step 8 and before using cmic. That mask corresponds to fog and low stratus together. I noted that the cloud filter based on ir39 and ir108 displays a message that the threshold is out the expected range for some files. The threshold should be between -5 and 0. Then it gives wrong cloud mask comparing to the mask in eumetsat view for the same date. That day was in 2021 August. Is that because the high values of ir39 for the hot ground ? Should I impose a range for my ir39 and ir108 values? Is my severi msg3 netcdf file correct? I don't know the cause.
Le ven. 8 mai 2026, 14:19, Holl Gerrit <Gerri...@dwd.de<mailto:Gerri...@dwd.de>> a écrit :
Hello Nadia Maatouk,
can you show us the exact code you have tried to run? algorithms.py is a module within fogpy and not the usual "entrance point". What do you mean by wrong cloud mask? Do you mean you get a fog mask, but that it is not the fog mask you are expecting?
Are you using daytime or night time data?
What cloud microphysics (CMICI) are you using? Calculated with NWCSAF GEO or coming from a different source?
Fogpy should work, in particular for SEVIRI, but as a note of warning, it has not been maintained for some years and nobody is using it operationally, so it is possible that some parts are not up to date with the newest satpy.
regards,
Gerrit.
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
From: pyt...@googlegroups.com<mailto:pyt...@googlegroups.com> <pyt...@googlegroups.com<mailto:pyt...@googlegroups.com>> on behalf of maatouk nadia <maato...@gmail.com<mailto:maato...@gmail.com>>
--
You received this message because you are subscribed to the Google Groups "pytroll" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pytroll+u...@googlegroups.com<mailto:pytroll%2Bunsu...@googlegroups.com><mailto:pytroll+u...@googlegroups.com<mailto:pytroll%2Bunsu...@googlegroups.com>>.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
Der Originalabsender dieser E-Mail ist: maato...@gmail.com. Wenn Ihnen dieser unbekannt ist, so sollten Sie die enthaltenen Links nicht anklicken.
as a disclaimer, it's been years since I looked at fogpy, so my mind is not very fresh on the topic, sorry. I don't remember what step 8 is. The daytime algorithm needs CMIC data. The night time algorithm uses IR39/IR108 and probably does not work during the day. It's also not very advanced and really only there as a filler for when the daytime algorithm cannot be used.
I would recommend getting some CMIC data. I think there are some in the data store, or otherwise they can be calculated with the NWCSAF GEO software or obtained from CMSAF. Or you can try if the algorithm works at night.
The algorithm implemented by fogpy is by now quite old, and meanwhile superior algorithms (I think) have been developed based on AI/ML, but those are not implemented in fogpy (fog as a product always drops down in the list in priorities at DWD, sadly). One of the reasons why I stopped exploring it further is that CMIC is only available at relatively high SZA (I think <78°), whereas the night time algorithm can only be used if there is no reflectance contamination in IR 3.9 when the Sun is safely down below the horizon. This means the twilight area, including the entire month of January in northern Germany, is missing (I guess one could try to calculate the thermal part of IR 3.9; I haven't tried), but this is the most important time of day for fog. That is a limitation on the algorithms in fogpy that some newer 24/7 fog algorithms manage to avoid.
Regards,
Gerrit.
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
From: maatouk nadia <maato...@gmail.com>
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
Sent: 08 May 2026 16:40:52
To: Holl Gerrit
Subject: Re: [pytroll] Issues with fpgpy
Thank you a lot for your response. I am running dayfoglowstratus algorithm. As I don't have CMIC data, I retrieve the mask after step 8 and before using cmic. That mask corresponds to fog and low stratus together. I noted that the cloud filter based on ir39 and ir108 displays a message that the threshold is out the expected range for some files. The threshold should be between -5 and 0. Then it gives wrong cloud mask comparing to the mask in eumetsat view for the same date. That day was in 2021 August. Is that because the high values of ir39 for the hot ground ? Should I impose a range for my ir39 and ir108 values? Is my severi msg3 netcdf file correct? I don't know the cause.
Le ven. 8 mai 2026, 14:19, Holl Gerrit <Gerri...@dwd.de<mailto:Gerri...@dwd.de>> a écrit :
Hello Nadia Maatouk,
can you show us the exact code you have tried to run? algorithms.py is a module within fogpy and not the usual "entrance point". What do you mean by wrong cloud mask? Do you mean you get a fog mask, but that it is not the fog mask you are expecting?
Are you using daytime or night time data?
What cloud microphysics (CMICI) are you using? Calculated with NWCSAF GEO or coming from a different source?
Fogpy should work, in particular for SEVIRI, but as a note of warning, it has not been maintained for some years and nobody is using it operationally, so it is possible that some parts are not up to date with the newest satpy.
regards,
Gerrit.
--
Remote Sensing Application Development
Research and development
Deutscher Wetterdienst (DWD)
+49 (69) 8062-3171
________________________________________
Sent: 02 May 2026 11:17:36
To: pytroll
Subject: [pytroll] Issues with fpgpy
Hello,
I run https://github.com/pytroll/fogpy/blob/main/fogpy/algorithms.py<https://ofcsg2dvf1.dwd.de/fmlurlsvc/?fewReq=:B:JVw4MzY6OCR0PzAsMiRrZj8yMzgyMyRxa2VsY3Z3cGc/Ozc3NmA1MDpmZzoyMTBmYzIyYTc3N2c0MGQ2MTo3O2QxOjs7NWQ6NiR2PzM1NTowNzMwOzMkc2tmPzQ2OkdkVGdtMjI7NTUzLzQ2OkdkVGdyMjI7NTUzJHBhcnY/RWdwcGt2LEptbm5CZnVmLGZnJGE/NzAkamZuPzI=&url=https%3a%2f%2fgithub.com%2fpytroll%2ffogpy%2fblob%2fmain%2ffogpy%2falgorithms.py>. I transformed the channels of msg seviri netcdf to brillaince temperature and reflectance. I run algorithms but I had wrong cloud mask comparing to eumetsat view. I don't know why! help pls
To: pytroll
Subject: [pytroll] Issues with fpgpy
Hello,
--
You received this message because you are subscribed to the Google Groups "pytroll" group.
To view this discussion, visit https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com<https://groups.google.com/d/msgid/pytroll/8629dfe0-ef0e-4958-9e73-c8a5060ac528n%40googlegroups.com?utm_medium=email&utm_source=footer>.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
Der Originalabsender dieser E-Mail ist: pytroll+bncBDU3R5M2XA...@googlegroups.com<mailto:pytroll%2BbncBDU3R5M2XALBB...@googlegroups.com>. Wenn Ihnen dieser unbekannt ist, so sollten Sie die enthaltenen Links nicht anklicken.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
----------
Hinweis (DWD): Mindestens eine URL in dieser E-Mail wurde vom Schadsoftware-Erkennungssystem als potentiell gefährlich eingestuft und so umgeschrieben, dass ein Klick darauf diese zunächst auf gefährliche Downloads hin untersucht, bevor Inhalte angezeigt werden. Die Analyse erfolgt ohne menschliches Zutun voll automatisiert. Der Datenschutz ist gewährleistet. Es kann einige Sekunden dauern, bis die Inhalte angezeigt werden.
Reply all
Reply to author
Forward
0 new messages