Predicting root size and frequency from one-dimensional consolidation data — A mathematical model