The latent heat should depend on temperature

[kmdeck]

Describe the bug

We currently do not include the dependency of the latent heat of freezing and vaporization on temperature. Update all code to do so.

https://github.com/CliMA/Thermodynamics.jl/blob/main/src/relations.jl#L769

To-do

• ID all locations where LH is used: snow parameterizations (T and liquid content from internal energy), soil parameterizations (phase change and T from internal energy), and the bucket model (melt fluxes)
• ID which ones need to depend on temperature: the latent heat appearing in the Clausis-Clapeyron equation should depend on T, I think. I updated the docs so that this is the case. The latent heat appearing in the internal energy should be at the reference point, so this is not updated in snow or in soil.
• TBD: revisit bucket model use of LH