Opened 7 months ago

Closed 7 months ago

#2920 closed help (answered)

What is the output variable for skin temperature in JULES?

Reported by: pmcguire Owned by: pmcguire
Component: JULES Keywords: skin temperature, surface temperature, air temperature, radiative temperature
Cc: mtodt Platform: JASMIN
UM Version:

Description

What is the output variable for skin temperature in JULES?
Is it tstar or trad or t1p5m? Or some other variable?

Change History (5)

comment:1 Changed 7 months ago by pmcguire

It is not t1p5m (the temperature of the air at 1.5m altitude above the ground). The skin temperature can be defined as the radiometric surface temperature (that's why it's called the skin temperature, like the skin of a person or of a hard object), not the air temperature.

Looking at the JULES 4.9 FORTRAN code:

The variable tstar is the surface temperature for each of the surface tiles.
The variable tstar_gb is the surface temperature averaged over all the surface tiles.

The variable trad is the 4th root of the "surface temperature to the 4th power, averaged over all the surface tiles".
This assumes that the emissivity of each tile is exactly 1.0, and it is effectively calculating the outgoing radiation of a perfect blackbody (BB(T) = T^4) averaged over all surface tiles, and then inverting this to compute trad = BBinv(Ave_Tiles(T^4)).

This is similar to what the authors do in the equation for brightness surface temperature in Equation 5 of:
"Land surface skin temperature climatology: benefitting from the strengths of satellite observations"
Menglin Jin and Robert E Dickinson
Environmental Research Letters (2010)
https://iopscience.iop.org/article/10.1088/1748-9326/5/4/044004/meta#erl361394s2

It's not the same as their Equation 6, where they correct for the effects of surface emissivity not being equal to 1.

So, to first approximation, the skin temperature is the same as trad in JULES, except in JULES the assumption is that the surface emissivity is equal to 1.

There should be differences between trad and tstar_gb in the JULES output, since the 4th root of the mean (over surface tiles) of tstar^4 should be different from the mean (over surface tiles) of tstar.

Jin and Dickinson (2010) state in the above paper that "[f]or a homogeneous surface, the temperature from this [brightness surface temperature] definition is equal to the thermodynamic temperature of the surface (Norman and Becker 1995)".

Also see the discussion in the first paragraph of the introduction of:
"Land surface skin temperatures from a combined analysis of microwave and infrared satellite observations for an all‐weather evaluation of the differences between air and skin temperatures"
Catherine Prigent, Filipe Aires, William B. Rossow
JGR Atmospheres (2003)
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2002JD002301

Does this answer the question?
Patrick

comment:2 Changed 7 months ago by mtodt

Thanks, Patrick! I guess it answers the question — I've added trad to the output list, which also includes tstar_gb. So, we should cover all bases.

Cheers
Markus

comment:3 Changed 7 months ago by pmcguire

  • Status changed from new to accepted

comment:4 Changed 7 months ago by pmcguire

OK, great, Markus!
I will close the ticket now.
Patrick

comment:5 Changed 7 months ago by pmcguire

  • Resolution set to answered
  • Status changed from accepted to closed
Note: See TracTickets for help on using tickets.