Questions about "skin temperature" definitions

In the current names, the variable `skin_temperature` is mixed between having a `surface_` prefix and not. Our desire is to remove the "`surface_`" part as superfluous, but it does seem like there may be a subtle difference between `surface_skin_temperature` and plain `skin_temperature`, and those differences may even depend on the field. We need to clarify the definition of these terms before resolving #150.

## Examples

[This older paper](https://doi.org/10.1175%2F1520-0442%281997%29010%3C1505%3AACOCBS%3E2.0.CO%3B2) uses "skin temperature" to refer to the satellite-derived temperature of the surface (also referred to as "radiometric temperature"). Likewise [this 2010 paper](https://iopscience.iop.org/article/10.1088/1748-9326/5/4/044004/pdf) goes into some depth on the definitions used for various surface temperatures, including "skin temperature" as being synonymous with "radiometric surface temperature".

However, [this 2016 paper](https://esd.copernicus.org/articles/7/697/2016/esd-7-697-2016.pdf) defines "skin temperature" as synonymous with "TOA temperature", citing two textbooks *Principles of Planetary Climate* (2010) and *Atmospheric Radiation* (1989), the latter of which I was able to verify [via Google Books](https://www.google.com/books/edition/Atmospheric_Radiation/Ji0vfj4MMH0C?hl=en&gbpv=1&dq=%22skin+temperature%22+intitle:atmospheric+intitle:radiation&pg=PA393&printsec=frontcover).

The former is the familiar definition to me, but that does not mean other definitions are invalid.

To make things even more complicated, there seems to be yet another definition [used in ocean observation contexts](https://journals.ametsoc.org/view/journals/clim/15/4/1520-0442_2002_015_0353_tivoss_2.0.co_2.xml): `sea_surface_skin_temperature` is defined as the IR radiatively-derived temperature of the upper ∼500 μm of the *water*, as opposed to other definitions that are a property of the near-surface *air*. It is also unclear to me whether existing variables such as `skin_temperature_at_surface_over_ocean` are referring to the oceanic radiative definition described here, or the atmospheric radiative definition above.

## Proposed definitions

Due to the potential ambiguity, I would propose that `skin_temperature` always contain a clarifying suffix. Since we already have `_at_surface` and `_at_toa` suffixes defined, I propose we use those:

- `skin_temperature_at_surface`: the "radiometric surface temperature" definition
- `skin_temperature_at_TOA`: the "TOA temperature" definition
- (*if needed*) `sea_surface_skin_temperature`: The SST skin temperature defined above.

But a follow-up question is, is `skin_temperature_at_TOA` meaningfully different from `temperature_at_TOA`? I am not sure if there is a potential difference between those definitions in an observational vs modeling context.