variables, names and units
After discussing with @florian.pinault, the two of us are unsure how to proceed: @awr @fpv07?
What precipitation variable do we want to forecast? @aaron.spring didnt see the difference between EWC and IRI data until today
- Total precipitation as accumulated precipitation since forecast start in units kg m-2?
- Precipitation rate for the given lead_time in units kg m-2 day-1?
This question is really important and needs to be on the website when we announce (and maybe even flyer?).
If we take 1., I still think we should upload 2. to the cloud for 1998-now, so users can calculate accumulated precipitation for the respective timings by themselves. For 2020 we would add accumulated precip observations separately. (Lets check how large this file would be for all thursday forecasts since 1998 with daily steps.)
If we take 1., precipitation terciles would depend on step
. For 2. terciles wouldn't depend on step
(if all inputs are bias corrected).
Currently the data holdings look like
daily output
EWC
variable name | standard_name (CF) | long_name | unit |
---|---|---|---|
t2m | air_temperature | 2 metre temperature | K |
tp | not set | Total Precipitation | kg m**-2 |
IRI
http://iridl.ldeo.columbia.edu/SOURCES/.NOAA/.NCEP/.CPC/.temperature/.daily/
variable name | standard_name (CF) | long_name | unit |
---|---|---|---|
tmin,tmax | air_temperature | Daily Temperature | C |
rain | lwe_precipitation_rate | Precipitation | mm/day |
comparing with the CF conventions: https://cfconventions.org/Data/cf-standard-names/28/build/cf-standard-name-table.html, which we see as the most suitable reference to take as a standard
variable name | standard_name (CF) | Description | unit |
---|---|---|---|
ta | air_temperature | Air temperature is the bulk temperature of the air, not the surface (skin) temperature. | K |
(tp would match here but not given in CF) | precipitation_amount | "Amount" means mass per unit area. | kg m**-2 |
pr | precipitation_flux | In accordance with common usage in geophysical disciplines, "flux" implies per unit area, called "flux density" in physics. | kg m**-2 s**-1 |
lwe_precipitation_rate | "lwe" means liquid water equivalent. | m s**-1 |
@florian.pinault and me agreed that we homogenize the variable names, units and standard_names towards CF standards.
- Temperature: we will stick to
t2m
- Precip: Which precip from the CF table should we take?
tp
as1.
andpr
as2.
?