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
which will not be able to set values > 2000 m to 0 m after clipping all of those values to = 2000 m. I think this line is unnecessary? The (2000 m - LCL) term, not LCL itself, will go to zero with LCL > 2000 m. If functioning, this line would maximize this term with LCL > 2000 instead of minimize.
Looking into this per #545 also reveals that the normalization value employed here is not documented in the referenced sources, and the cited Thompson 2004b has a fifth term (CIN) not included here. Not sure if these are changed documented somewhere to adapt to surface-based parameters instead of the cited mixed-layer parameters. The cited LCL term is given to be (2000 m - ML LCL) / 1500 m, with the same in the Thompson 2003 formulation, instead of our (2000m - SB LCL) / 1000 m. This and the lower clip of 12.5 m/s instead of 10 m/s for shear appear to be the only changes from the citation, so I might look into that a bit more.
Also, here significant_tornado (and calc.supercell_composite, not sure about others) converts scalar values to ndarrays per #622 and so return those ndarrays inside the resulting Quantity instead of scalars. Makes for inconsistencies in strings (and potentially elsewhere?) as brought up here. Not sure if this behavior will be affected by wrapping from #1353 (guess it's as good a time as any to finally get caught up on all that, huh?) @dopplershift suggested checking for this could end up a part of changes introduced with #1388
I can get rid of the 2000/0 line, but I'm not sure if I should otherwise change up the math and/or address this scalar->ndarray behavior, so I'll look into those some more.
The text was updated successfully, but these errors were encountered:
calc.significant_tornado
includes the following lines of code:which will not be able to set values > 2000 m to 0 m after clipping all of those values to = 2000 m. I think this line is unnecessary? The (2000 m - LCL) term, not LCL itself, will go to zero with LCL > 2000 m. If functioning, this line would maximize this term with LCL > 2000 instead of minimize.
Looking into this per #545 also reveals that the normalization value employed here is not documented in the referenced sources, and the cited Thompson 2004b has a fifth term (CIN) not included here. Not sure if these are changed documented somewhere to adapt to surface-based parameters instead of the cited mixed-layer parameters. The cited LCL term is given to be (2000 m - ML LCL) / 1500 m, with the same in the Thompson 2003 formulation, instead of our (2000m - SB LCL) / 1000 m. This and the lower clip of 12.5 m/s instead of 10 m/s for shear appear to be the only changes from the citation, so I might look into that a bit more.
Also, here
significant_tornado
(andcalc.supercell_composite
, not sure about others) converts scalar values to ndarrays per #622 and so return those ndarrays inside the resulting Quantity instead of scalars. Makes for inconsistencies in strings (and potentially elsewhere?) as brought up here. Not sure if this behavior will be affected by wrapping from #1353 (guess it's as good a time as any to finally get caught up on all that, huh?) @dopplershift suggested checking for this could end up a part of changes introduced with #1388I can get rid of the 2000/0 line, but I'm not sure if I should otherwise change up the math and/or address this scalar->ndarray behavior, so I'll look into those some more.
The text was updated successfully, but these errors were encountered: