search

JCSDA-internal / ioda-converters

Various converters for getting obs data in and out of IODA
8 stars 2 forks source link

[New issue] IODA naming convention for TOA reflectance #1470

Closed weiwilliam closed 4 months ago

weiwilliam commented 4 months ago

Description

The existing GOES ABI converter (goes_converter.py) uses albedo for reflectance factor. And there is no reflectance term in share/ioda/yaml/validation/ObsSpace.yaml.

In the CF convention, there are planetary_albedo and surface_albedo.

planetary_albedo: 
Albedo is the ratio of outgoing to incoming shortwave irradiance, 
where 'shortwave irradiance' means that both the incoming and outgoing radiation are integrated across the solar spectrum.

surface_albedo:
The surface called "surface" means the lower boundary of the atmosphere.

Both of them are integrated across solar spectrum, which is not the case for ABI or VIIRS data. Should we update the ObsSpace.yaml with toa_reflectance or toaReflectance or toa_bidirectional_reflectance? Unsure "bidirectional" is the right term for ABI reflectance factor or VIIRS TOA reflectance. Bidirectional means it depends on angles of incident (zenith) and measured (scan) radiation. Tagging @chengdang and @BenjaminTJohnson for the inputs of terminologies.

Nonetheless, the converted reflectance factor for GOES ABI is the same term stored in VIIRS L1b data. Both of them equal to reflectance * cos(zenith angle), so we can use the same variable name in the converted IODA file.

PatNichols commented 4 months ago

@weiwilliam This might be for @gthompsnJCSDA. If new variables are added we would need to update the data conventions accordingly.

gthompsnJCSDA commented 4 months ago

If you all don't mind, I wish to close this issue. To me, this is not an issue. There is a data convention name of albedo that is intended for use in observed satellite data. I modified the ioda-converter in question from reflectance factor to albedo over a year ago I think and find it sufficient for our use. Instead, what I have not yet done is change my own UFO interface code to CRTM to call the UFO variable the same name. In my own feature branch for visible wavelengths, the variable is referred to as reflectance but I intend to change it to albedo as early as this week.

weiwilliam commented 4 months ago

@gthompsnJCSDA Thanks for the clarification. Please check my follow-up question in JCSDA-internal/discussion.