Predicting The Soil Water Characteristic Curves From The Particle Size Distribution

Predicting the soil water characteristic curve(SWCC)based on the shape similarity between the SWCC and the particle size distribution(PSD)curve appeared rapid and economical,and which can be applied potentially.But these traditional models underestimated the water content in the dry range of the SWCC.The factors resulted in this errors were discussed and three different improved methods predicting the SWCC were proposed.Besides,abundant measured data were used to validate proposed methods,and then the performance of traditional method and the three improved methods were compared.It was proven that the error of the dry range of SWCC originate from the poor translation of pore size distribution(PoSD)from PSD.In improved methods,the suction head calculation followed the approach of the traditional model,which used Yong-Laplace equation and the liner relation between the pore size and the particle size.Primarily,the water content calculation approach was improved.Improved method?calculated the water content as the sum of film water and capillary water content,which increased the water content obviously in the dry range of SWCC.Improved method ? calculated the PoSD based on the pore model containing the slit shaped space,and the pore volume fraction at the smaller pore diameter increased obviously.In improved method ?,a regression equation described the residual pore fraction and the clay fraction facilitated to predict PoSD from PSD based on the characteristic relation between measured particle fraction curve and pore fraction curve,then to predict the SWCC.The calculation procedure of improved method?were constructed based on the physical situation in the drainage process,which could gain the best prediction of the dry range of SWCC and could be potentially promoted to predict the SWCC from saturation to oven-dryness.In which,the estimated specific surface area fairly influenced the predicting accuracy.The calculation procedure in improved method ? was slightly complex but it could gain stable predicting performance.The selection of pore parameters and the estimated specific surface area would influence the predicting accuracy.Improved method ? was easily to use,but the error resulted from index ? obtained from statistical analysis possibly gained unfaithful predictions of SWCC for few samples.As a whole,the SWCC predicted from the three improved methods fitted the measured data well and were more accurate than those obtained from the traditional method,especially the model performance in the dry range of the SWCC was improved obviously.