Implement basic coarse filtering for ALL variables

[gremau]

Lots of variables have problematic outliers or "spikes" that are very temporally limited ( < 1 day ), despite the qc process we have in place. Often I think this is just because they are relatively rarely used variables, or these outlier/spikes are uncommon and not explainable. Whatever the case, it would be nice to remove many of them with a simple coarse filter. For example:

Wind direction: Remove data < 0 and > 360
RH: Remove outliers > 100 and < 0
Other: Remove outliers greater than 3-4 standard deviations from the 1 or 2 week rolling mean.

This should probably occur as part of the "RemoveBadData" pipeline.