Abstract
The available water capacity (AWC) of especially coarse-textured forest soils of the federal state of Brandenburg - estimated by various pedotransfer functions (PTF) – induce huge systematic differences. Thus further validations of the PTF seem to be necessary. The PTF were evaluated, using a dataset of overall 489 forest soil horizons of the north-eastern lowlands. The dataset includes measured AWC-values and further soil parameters required for the PTF. The comparison of measured and estimated AWC-values indicates that PTF, which were developed using large, geographically widespread datasets and which consider especially soils similar to those of the validation sample respectively, tend to smaller systematic prediction errors. The PTF, which consider a higher number of predicting variables generally induce higher coefficients of determination. The lowest overall error (square root of the mean square prediction error) was determined for the PTF Renger et al. (2009). The highest values describing the strength and accuracy of the linear relationship between measurements and predictions (coefficient of determination and slope) were found out for the approach of Riek et al. (1992, 1995).