Development of Pedotransfer Functions to Predict Soil Physical Properties in Southern Quebec (Canada)

Pedotransfer functions (PTFs) are empirical fits to soil property data and have been used as an alternative tool in situ measurements for estimating hydraulic properties the last few decades. PTFs of Saxton Rawls, 2006 (PTFs’S&R.2006) some most widely because their global aspect. However, yield more accurate results when trained locally. This study proposes a set agricultural developed southern Quebec, Canada three horizons (A, B, C). Four response variables (bulk density (ρb), saturated conductivity (Ksat), volumetric water content at field capacity (θ33), permanent wilting point (θ1500)) four predictors (clay, silt, organic carbon, coarse fragment percentages) were this modeling process. The new using stepwise forward regression (SFR) canonical correlation analysis (CCA) algorithms. CCA- SFR-PTFs cases accurate. Θ1500 θ33 estimates improved with SFR. ρb A horizon was moderately estimated by PTFs’S&R.2006, while performed equally well B C horizons, yet qualified weak. all Ksat unacceptable. Estimation could be considering other morphological (soil structure, drainage information, etc.).