Feature: Multi-source Initial Conditions (IC) assembly in `ic_process`

[dlebauer]

Description

Current behavior: ic_process supports one source per run.

Problem: Many data sources provide soil but not veg c, and vice versa - in these cases it is not possible to generate a valid IC from a single source.

The motivating use case is to enable ic_process() to combine soil carbon initial conditions from one source with vegetation carbon pools from a second source into a single poolinitcond file - originally discussed in #3508 (comment).

Proposed Solution

Update ic_process() to support multiple sources.

Inputs:

• Soil C:
  • SoilGrids
  • gSSURGO
  • custom CSV
  • function arguments (e.g. soil_c = ...)
• Vegetation C pools: (choose one):
  • CSV
  • optional function args e.g. agb_c, root_c, leaf_c
  • Bare ground (0),

Optional for MVP (could be separated into a separate feature request):

• MODIS (LAI) + LandTrendr (biomass) as implemented in
  2a_grass/02_ic_build.R.

Outputs:

• IC files, including provenance of values
• set settings$run$inputs[["poolinitcond"]]$path <- newfile

[dlebauer]

To follow up conversation from #3508 (comment)

we should just extract the vegetation and other non-soil pools from BADM and merge those with the SoilGrids SOC data.

For BADM, when it provides SOC, we should use the BADM IC. BADM isn't the priority in this case.

if SoilGrids is the passed as source in settings, another option would be to call Landtrendr_AGB_prep() to get aboveground biomass medians and standard deviations, and use extract_SM_CDS() for soil moisture

I like the general idea, but I don't think there should be a case where SoilGrids is passed in and then the other IC sources are chosen automatically. It should be an option (perhaps required?) that a user specify each source of IC. I think both the settings and the IC file should identify the IC sources.