Why is effect_wspeed (effective wind speed for SPITFIRE) calculated at site level and not patch level?

[adrifoster]

I noticed that the patch%effect_wspeed is calculated using the total site-level average tree_fraction, grass_fraction, and bare_fraction

It seems like this is being done as a comment says "to prevent extreme divergence"

Is this what we want? I suppose it makes sense that the wind would be influenced by all the patches and not just the one we're on. In that case should this not be a site-level variable?

[rosiealice]

Yeah, this is basically to prevent lock in of patches into their existing states. Some simulations I did forever ago had very strong influence of initial conditions wrt the impact of changing climate, and this was something I did to try and ameliorate this somewhat.

[adrifoster]

In that case is it okay if I make it a part of a site-level class?