Predicting Soil Bulk Density for Incomplete Databases.

Soil bulk density (ρ_b) measurements are commonly missing from databases for reasons that include omission due to sampling constraints and laboratory mishandling. The objective of this study was investigate the performance of novel pedotransfer functions (PTFs) in predicting ρ_b as a function of textural class and basic pedon description information extracted from the horizon of interest (the horizon for which ρ_b is being predicted), and ρ_b, textural class, and basic pedon description information extracted from horizons above or below and directly adjacent or not to the horizon of interest. Twelve ρ_b PTFs were developed by combining PTF types, database configurations, and horizon limiting depths. Different PTF types were created considering the direction of prediction in the soil profile: upward and downward prediction models. Multiple database configurations were used to mimic different scenarios of horizons missing ρ_b values: random missing (e.g., ρ_b sample lost in transit) and patterned or systematic missing (e.g., no ρ_b samples collected for horizons > 30 cm depth). For each database configuration scenario, upward and downward models were developed separately. Three limiting depths were tested to identify any threshold depth between upward and downward models. For both PTF types, validation results indicated that models derived from the database configuration mimicking random horizons missing ρ_b performed better than those derived from the configuration mimicking clear patterns of missing ρ_b measurements. All 12 PTFs performed well (RMSPE: 0.10−0.15 g cm^-3). The threshold depth of 50 cm most successfully split the database between upward and downward models. For all PTFs, the ρ_b of other horizons in the soil profile was the most important variable in predicting ρ_b. The proposed PTFs provide accurate ρ_b predictions, and have the potential to help researchers and surveyors fill gaps in their database without complicated data acquisition.