Help: GFS DSWRF

[peterdudfield]

Hi

Thanks so much for this repo, I think its really good. I'm trying to get some solar irradence data from GFS, but I cant seem to find any.

from herbie import Herbie

H = Herbie("2023-10-12 12:00", model="gfs")
H.inventory().to_csv('gfs.csv')

But then I cant find anything to do with solar irradanece, I thought it would come up as DSWRF and DLWRF.
Can anyone help?

Sorry if this is not the right place to put this

[blaylockbk]

Good question. You're on the right track. I'll have to sit at my computer and dig around, but have you tried looking in of the the different GFS products? Sometimes the different resolutions have different variables.

[blaylockbk]

It looks like the DSWRF variable doesn't exist for fxx=0, but it does for the forecast hours.

from herbie import Herbie
H = Herbie("2023-10-01", model="gfs", fxx=6)
H.inventory("DS")

	grib_message	start_byte	end_byte	range	reference_time	valid_time	variable	level	forecast_time	search_this
652	653	461198994	462155002	461198994-462155002	2023-10-01	2023-10-01 06:00:00	DSWRF	surface	0-6 hour ave fcst	:DSWRF:surface:0-6 hour ave fcst

ds = H.xarray("DSWRF")
ds.dswrf.plot()

[blaylockbk]

Same for DLWRF

ds = H.xarray("DLWRF")
ds.dlwrf.plot()

[blaylockbk]

Thanks for your question, @peterdudfield, and I'm glad you enjoy Herbie 😀

[blaylockbk]

Hint: You can use regex to get both longwave and shortwave flux

ds = H.xarray("D[L,S]WRF")

[peterdudfield]

Thank you very much @blaylockbk , im happy if you want to close this

[williamhobbs]

I have been referring to this NOAA inventory page for double-checking things like this: https://www.nco.ncep.noaa.gov/pmb/products/.

For example, the GFS page (https://www.nco.ncep.noaa.gov/pmb/products/gfs/) let's you browse a typical f000 file (https://www.nco.ncep.noaa.gov/pmb/products/gfs/gfs.t00z.pgrb2.0p50.f000.shtml) and one for f003-384 (https://www.nco.ncep.noaa.gov/pmb/products/gfs/gfs.t00z.pgrb2.0p50.f003.shtml). A Ctrl-F search for "DSWRF" on the first one finds nothing, but it shows up in the second one: