Evapotranspiration is never removed from the soil water in LandModel

[olivierbonte]

When working on my implementation plan for #201, claude pointed me to an issue with the coupling between evapotranspiration and the soil scheme.

Terrarium.jl/src/processes/soil/soil_coupled.jl

Line 97 in 9f0d786

compute_tendencies!(state, grid, soil.hydrology, soil, constants)

does not have the evapotranspiration argument although it is an option in

Terrarium.jl/src/processes/soil/hydrology/soil_hydrology_rre.jl

Line 89 in 9f0d786

function compute_tendencies!(

MWE for the problem

using Terrarium

# Bare-ground column with Richards soil hydrology; no rainfall
function run_column(specific_humidity; nsteps = 60, Δt = 60.0)
    grid = ColumnGrid(CPU(), Float64, UniformSpacing(Δz = 0.1, N = 10))
    hydrology = SoilHydrology(Float64, RichardsEq())
    soil = SoilEnergyWaterCarbon(Float64; hydrology)
    model = LandModel(grid; soil, vegetation = nothing)
    initializers = (temperature = 15.0, saturation_water_ice = 0.5)
    integrator = initialize(model; initializers)
    set!(integrator.state.specific_humidity, specific_humidity)
    set!(integrator.state.rainfall, 0.0)
    for _ in 1:nsteps
        timestep!(integrator, Δt)
    end
    state = integrator.state
    return (
        evaporation_ground = state.evaporation_ground[1, 1, 1],
        saturation_top = state.saturation_water_ice[1, 1, grid.Nz],
    )
end

dry = run_column(1.0e-4)   # strong evaporative demand
humid = run_column(1.0e-2) # weak (or negative) evaporative demand

@show dry.evaporation_ground humid.evaporation_ground
@show dry.saturation_top humid.saturation_top
# Expected: different top-layer saturation. Actual: identical, since ET is never applied to the soil.
@show dry.saturation_top == humid.saturation_top

[bgroenks96]

Could be a dispatch bug introduced by some recent changes? Can you check whether or not the forcing method is being called?

[bgroenks96]

Removed my previous comment because it was just me making the usual indexing mistake of end without interior... I'll just blame it on being in a train for 10+ hours.

Anyway, see #204.