Improper values of PAW (plant available water) for OM-rich clay soils

[yuki-nmi]

[Problem]
It is usually expected that water retention is better in OM-rich and clay-rich soils.
However, with the current PTF for water retention curve (Wosten 1999), plant available water (PAW = theta_fc = theta_wp) decreases with increasing SOM, when clay content is high. This gives unwanted results for the I_P_WRI score for clay-rich regions, such as Flevoland.

[Analysis]
This is because the difference between water content at field capacity and that of wilting point slightly decreases for the intermediate SOM range for high-clay soils. Water content at field capacity and wilting point do increase with SOM, which is fine. So, the negative influence of SOM on PAW is merely due to smalll error/uncertainty of pedotransfer function, which is probably based on the dataset with not much high-clay soils.

[Possible solution]
Use of another available PTF, Wosten 2001, is not a solution, because it gives similar outputs:

Alternatively, we can use Water Holding Capacity to evaluate water retention. This gives more robust results.

The evaluation function for WHC is alreadyincluded in OBI, as follows:

dt[, D_WRI_WHC := calc_waterretention(A_CLAY_MI,A_SAND_MI,A_SILT_MI,A_SOM_LOI,
                                    type = 'water holding capacity', ptf = 'Wosten1999')]
dt[, I_P_WRI := ind_waterretention(D_WRI_WHC, type = 'water holding capacity')]