Soil Consistence and Structure as Predictors of Water Retention

摘要: It is impractical to measure water retention for large-scale hydrologic, agronomic, and ecological applications or at the design stages of many projects; therefore, estimates are often used. Field soil descriptions routinely include structure consistence characterization. The objective this work was use National Resource Conservation Service (NRCS) database evaluate potential structural properties serve as predictors hydraulics properties. Total =2140 samples were found that had (i) values contents -33 kPa -1500 kPa, (ii) characterized with grade, size, shape, (iii) dry moist consistency, stickiness, plasticity, (iv) textural class determined in field from lab analysis. Because parameters represented by categories rather than numbers, regression trees used recursive partitioning data sets into groups decrease overall variability measured sum squared errors within groups. Plasticity class, grade consistency leading both matric potentials. accuracy lower classes. Using along classes provided a small, although significant improvement compared estimation texture alone. Soil can because those reflect basic affect hydraulic