You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In a recent FATES-Hydro simulation with API 36, FATES_SAPFLOW is uniformly zero, but FATES_SAPFLOW_SZPF is not. Also, FATES_SAPFLOW_SZPF values seem too low and some are negative. I'll see if I can track this down.
The text was updated successfully, but these errors were encountered:
Seems like FATES_SAPFLOW is zero because it's never actually calculated in update_history_hydraulics().
Yes! I looked at this with @ckoven, @JessicaNeedham, and @mpaiao and we realized that's the case. As for the values of FATES_SAPFLOW_SZPF being quite low compared to observed values, @ckoven realized it's because it is being calculated per unit ground area (as opposed to per unit sapwood area). So the plan is to add sapwood area as a history variable.
In a recent FATES-Hydro simulation with API 36, FATES_SAPFLOW is uniformly zero, but FATES_SAPFLOW_SZPF is not. Also, FATES_SAPFLOW_SZPF values seem too low and some are negative. I'll see if I can track this down.
The text was updated successfully, but these errors were encountered: