Using a Visnir Penetrometer to Map Field-Scale Clay Content, Bulk Density and Restrictive Layers for the Palouse Region, Washington, USA.

High resolution maps of important hydrological properties of soil such as clay content, bulk density and presence of low hydraulic conductivity layer (i.e. restrictive layer) improve the spatial modelling of soil moisture and hydrologic processes. In this study, we used a newly design VisNIR penetrometer foreoptic, capable to collect spectra and insertion force simultaneously in situ to estimate clay content and bulk density and determine the presence of restrictive layer. Three fields in the Palouse region (Washington and Idaho, USA) were interrogated: we collected soil spectra and insertion force data along 35m x 35m grid points (2 field) and 50m x 50m grid points (1 fields) to ≈80cm depth. Based on terrain indexes, we stratified the fields and randomly selected 36 locations per field outside the grid points (12 for calibration and 24 for validation purposes). Next to the 36 points, we extracted two set of calibration cores. In laboratory, we cut the first set of cores in 3 cm slices, centered on depth of in situ interrogation and analyze for clay content (pipette method). The second set was cut at 10cm depth interval and we determined the bulk density for each interval. Finally, we measured soil apparent electroconducitvity (EC_a) across the three fields. We combined soil spectra and insertion force and used partial least square regression (PLSR) to build calibration models of soil clay content and bulk density. We applied spline functions to describe clay and bulk density profiles at each points (grid and 12 locations). We then used regression kriging with terrain indexes and EC_a data as covariates to generate 3D soil map as in Hengl et al. (2014). Preliminary results showed the VisNIR penetrometer ability to adequate estimate clay content and detect variation of bulk density consistently across the field. Work is ongoing to evaluate the prediction accuracy of the 3D soil map in estimate clay and restrictive layer presence.