search

CliMA / Thermodynamics.jl

A package containing a library of moist thermodynamic relations.
https://clima.github.io/Thermodynamics.jl/dev/
Apache License 2.0
56 stars 2 forks source link

Kinematic air viscosity #217

Open simone-silvestri opened 2 months ago

simone-silvestri commented 2 months ago

I was searching for some way to calculate the kinematic viscosity based on temperature and or humidity. Is there any plans to add a simplified viscosity law?

I guess something simplified that depends on temperature like Sutherland's law might be enough, I am not sure how much humidity might play a role though

trontrytel commented 2 months ago

So far I only used it as constant. Do other models consider the temperature dependence?

simone-silvestri commented 2 months ago

I was looking at the implementation of the COARE3.6 algorithm and it looks like they are using a temperature-dependent viscosity.

However, it looks like in typical atmospheric conditions the viscosity does not change more than 60 - 70%. A constant viscosity is probably totally fine...

trontrytel commented 2 months ago

Should we add it as an option so that you can test if it makes a difference? It's empirical, right? I don't have any intuition on if those differences would be important for your application, or for CloudMicrophysics.jl for that matter. I can test it in CM.jl if you think thats useful.

Assuming t is in Celsius(?) Those are the numbers I was getting. Our default is 1.6e-5 so we are around the warm temperature end there.

Screenshot 2024-04-30 at 10 19 33 AM
glwagner commented 2 months ago

I think it's mostly important for reproducing COARE3.6. More generally I don't think it's super important; it comes into play when diagnosing roughness lengths at very low wind speeds over the ocean --- low enough that the fluxes predicted by the Charnock relation are less than what would be predicted for a laminar boundary layer over a smooth wall.