Fix LFC problem and set CIN to 0 if positive

[zbruick]

Description Of Changes

This fixes two problems - one with the LFC checking to make sure it and an EL both don't occur below a LCL - and the other where CIN was returned as a positive value when the parcel is absolutely unstable. CIN will be set to 0 in these cases, per conversation in #1190.

Checklist

• Closes cape_cin ValueError: zero-size array to reduction operation minimum which has no identity #1190
• Tests added
• Fully documented

[jthielen]

Based on the decision in #1190 (comment), would you be able add a citation to the USAF definition of LFC used by MetPy and an brief explanation of why the LFC cannot be below the LCL to the docstring, since that may or may not be what users would initially expect?

[zbruick]

Yup, can do. Thanks for the suggestion. Hopefully this resolves this once and for all.

[dopplershift]

Code changes look good. Let's include a copy of the manual in docs/_static and avoid linking to a horrifically slow site. I have a clean copy of the manual in my archives...

[dopplershift]

Here's the clean copy: aws-tr-79-006.pdf.gz

[zbruick]

Glad we had a better way to do it rather than ignoring the link, as it was failing at a rather historic rate.