Analysis of the suitability of Polish soil texture classification for estimating soil water retention and hydraulic properties

Abstract The objective of this study was to examine whether the Polish soil textural classification is useful for evaluation of soil water retention and hydraulic properties and, furthermore, for determining which textural classes are characterized by the highest diversity of soil water retention and hydraulic properties. The texture triangle was divided into a 1% grid of particle-size classes resulting in 5151 different data points. For each data point, soil water retention parameters and saturated hydraulic conductivity were obtained using the ROSETTA program. The silt classes showed the highest uncertainty in the estimation of the saturated water content based on the soil texture. These classes are characterized by high variations of saturated water content within the class. Estimations of field capacity and permanent wilting point on the basis of textural classes are encumbered with highest errors for gp, pg, pl and pyg soils, which are characterized by the highest values of coefficient of variation. Saturated soil hydraulic conductivity is better classified into homogeneous classes by the Polish texture classes than by the clusters obtained by the k-means cluster analysis based on the soil hydraulic and retention properties. Soil water retention parameters are better classified into homogeneous groups by the k-means cluster analysis than by the traditional textural classes. Cluster analysis using the k-means can be helpful for grouping similar soils from the point of view of their retention properties.