AAPPv8.7 – MHS flag set to 0 everywhere?

[Adam Dybbroe]

Hi,
We have noticed that the NWCSAF/PPS v2018 fail to generate the MW Scattering index used for the Precipitating Clouds product, when using AAPPv8.7 output, where as it works fine with AAPPv8.3 output.
It concerns all satellites, Metop-B, Metop-C, NOAA-18 and NOAA-19. The flag concerned is supposed to contain information on LAND/OCEAN/FAIL/etc. In 8.7 it is set to zero everywhere as far as we can see.

PS: By the way, just tried to reach the format description being logged in to your website and it says I am not permitted to reach this doc:
https://nwp-saf.eumetsat.int/site/download/NWPSAF-MF-UD-003_Formats_v8.1.pdf

Is that as it is supposed to be?

-Adam

[Stephen Bedford]

Hi Adam,
I don’t work on the software so can’t answer your question on AAPP (I am a site admin). I have changed the access for that file though, you should be able to view it now. Thank you for pointing out the error.

Regards,
Duncan

[Adam Dybbroe]

Thanks, I see the format documentation online now
-Adam

[Nigel Atkinson]

Adam,

I’m not aware of any change in 8.3 and 8.7 that would affect the MW scattering indexes, but will take a look. There are several indexes … could you say which one you are referring to?

Nigel

[Adam Dybbroe]

I was not really precise in my initial report above. We convolve a land-sea mask on the MHS footprints and for that we use the geo-location and the sun-sat viewing angles. And it appears that the latter are wrong from v8.7.

I have some plots showing that, but can’t find a way to attach them here?

[Adam Dybbroe]

Nigel,

Well, it is the scattering index developed by Ralf Bennartz for the NWCSAF/PPS Precipitating Clouds algorithm! Not any AAPP related one! But, when calculating this, the viewing angles are used and they are wrong with 8.7 and okay with 8.3!

At least from what analysis says so far.

We noticed it recently when preparing for an upgrade of AAPP in production. Currently we run PPSv2018 in production using input from AAPPv8.3, and everything runs fine. But changing to 8.7 the Precip product doesn’t get produced, as the MW scattering index calculation (a pre-processing step) fails.

-Adam

[Author]

Posts