Canopy lower boundary effects

[arbennett]

In discussion around #223 there was a lot of discussion of an improved algorithm for the case where the snow depth is near or above the canopy bottom. One major concern is that when snow is above the canopy bottom (or canopy bottom is effectively 0), the wind speed may truncate at a relatively high value. This behavior can be seen in this notebook in the Quick Check section (red line).

Mahat2013 provides a description of the current implementation and may provide some insight into different approaches. Figure 1 shows a schematic of the heights. Specifically z_ms is the bottom boundary between the exponential and is set to the height of the bottom canopy boundary in our implementation. Perhaps there is some minimum height for z_ms so that the bottom logarithmic boundary is always included.

It is important to note that the only windspeeds explicitly calculated and used are the ones at the canopy boundaries.

[martynpclark]

@arbennett -- what is the status of this?

[arbennett]

The status of this hasn't changed. When fixing the canopy calculations we never decided if this was something that was important to address. The only thing changing this would affect is a small adjustment to the ground resistance.