search

stcorp / harp

Data harmonization toolset for scientific earth observation data
http://stcorp.github.io/harp/doc/html/index.html
BSD 3-Clause "New" or "Revised" License
55 stars 18 forks source link

Add support for `geolocation_flags` of S5P NO2 product #254

Closed zxdawn closed 2 years ago

zxdawn commented 2 years ago

I discussed with Henk Eskes last week and we thought that the geolocation_flags is useful to filter ascending/descending data and determine sun glint. It would be great to add it to harp.

svniemeijer commented 2 years ago

I am not sure that this will add much. Sun glint information is already included in the processing_quality_flags (validity HARP variable). And for ascending/descending you could perhaps use something like solar_zenith_angle<=90 (to only have the ascending part).

In any case, we don't have a good HARP variable to map these additional flags into. Ideally, all flags should have been in a single flag variable in the product. So if certain flags are important for filtering, I would recommend getting them added to processing_quality_flags.

zxdawn commented 2 years ago

Thanks for the reply.

Here's the descending flag (value = 4) from the processing_quality_flags: image

And, here's the result based on SZA>=90. image

It seems the ascending/descending mode can't be easily determined from SZA only.

svniemeijer commented 2 years ago

It is not a straight ascending/descending split, but it does split the orbit in two. Why would you need to know this explicit asc/desc split?

zxdawn commented 2 years ago

We are working on the Arctic NOx emission and there're some significant dividing lines in the monthly average NO2 map. It may be caused by the merge of descending and ascending data.

svniemeijer commented 2 years ago

If you are splitting at the asc/desc split you will just get dividing lines somewhere else, so I doubt that that will fix it for you.

zxdawn commented 2 years ago

Em ... Here's an example of a monthly NO2 map using qa_value>=0.75 filter: image

That dividing line (between dark and bright color) seems to coincide with the SZA==90 border. So, maybe picking ascending ones would fix it?

svniemeijer commented 2 years ago

Have you tried using the solar_zenith_angle<=90 that I mentioned? In the end, it is about not mixing different local times, not about having a clear split at the pole.

svniemeijer commented 2 years ago

But can you move this question to forum.atmospherictoolbox.org? This is no longer a HARP issue, but more a 'how to' question.

zxdawn commented 2 years ago

Thanks, I moved this to the forum post and added the SZA<=90 results.