soil data - variable mapping and conversion

[ValentineHerr]

@teixeirak , So far I've only been dealing with non-soil data file. The soil data file is in a very close format to the non-soil data and it should be easy to import. If we do want to import it, I think there is a need for a few extra variable name conversions, instructions for ref column and maybe how to deal with min and max depth if it is different than for non-soil data.

[teixeirak]

Thanks.

We'll only use soil data with min.depth=0, so max.depth will be recorded in depth covariate field. I don't think we'll do anything with the ref column. I'll provide variable name conversions / do carefufl review later.

[teixeirak]

@ValentineHerr, Here is the variable mapping:

GROA variable	ForC variable	ForC units	ForC description
SOC	soil_C	Mg C ha-1	Soil organic carbon to specified depth in mineral soil
soil_perC	secondary	soil_pC	%
SOM_per	soil_OM	Mg ha-1	Soil organic matter to specified depth in mineral soil

@CookPatton, please confirm.

[CookPatton]

@ValentineHerr

That's correct for SOC and soil_perC, but SOM_per is % as well, just oddly percent organic matter.

[teixeirak]

In that case, SOM_per won't map to anything. I updated variable_name_conversion.

[CookPatton]

@ValentineHerr and @teixeirak - I realized this morning that my "sum to surface" approach works for SOC (t/ha/depth), but not for percent SOC. Instead of providing the most raw form of the data, I could give you the data where I've converted to SOC . But in many cases, this involves using modeled bulk density data from SoilGrids. How would you like to proceed?

[teixeirak]

We’ll just exclude it from ForC or use the originally reported depths (low priority for import). It’s only a supplementary variable, not a C stock.

[CookPatton]

Okay great - @ValentineHerr please do exclude the soil-perC numbers in the soil summed to surface.csv, since those are not correct.