@article{7198603,
  abstract     = {{Applicability of time domain reflectometry (TDR) under naturally distributed stone fragments in soils has seldom been investigated. A multilayer profile of a 30-m-deep well was sampled and the natural distribution of stone fragments in the soils was replicated in the laboratory. Gravimetric soil water content (SWC) was measured simultaneously with TDR dielectric permittivity (Ka) readings and bulk densities in three subsamples as replications. Two connector and buriable probes and three reflection-time capture windows (10, 20 and 40 ns) were used for the measurements. These were repeated for sieved soil samples <2mm with fixed, pre-measured bulk densities. Measurements of Ka and observed SWC were repeated for extension-cable lengths of 3–30 m. All measurements were taken in samples saturated from the bottom. A semi-empirical mixture model was applied for different fractions of stony samples in order to convert bulk Ka to bulk volumetric SWC (qv) by the mixture model (qvmx), to be compared with qv by the conventional Topp equation (qvTp). An improvement in model performance was observed with lower root-mean-square error (RMSE, 0.02–0.04 v. 0.07–0.1) and ratio of RMSE to observation standard deviation (0.32–0.87 v. 1.07–3.05) for qvmx compared with qvTp. This approach for converting the in-situ measured dielectric permittivity to the qv of the bulk soil can be applied based on the determined stoniness. The 15-cm, 2-rod (connector) probe type with capture windows 20 ns resulted in a better performance than the 20-cm, 3-rod (buriable) probe type with capture windows 10 and 40 ns. Development of regression equations for the stonefree samples resulted in calibrated equations for converting the measured Ka to qv with better results (RMSE ~0.002m3 m–3) than those obtained using the Topp equation. In contrast to the traditional equation, new sets of coefficients for the Topp equation were also capable of estimating extremely low qv values of <=0.02m3 m–3 where the minimum calculated qv values were adequately similar to the observed ones. Noticeable effects of cable length on measured Ka were found for lengths exceeding 10 m. Accurate Ka values might be obtained in similar soil conditions if the suggested regression equations are employed, provided a correction is made for the extension cables.}},
  author       = {{Pakparvar, Mojtaba and Cornelis, Wim and Gabriëls, Donald and Mansouri, Z and Kowsar, SA}},
  issn         = {{1838-675X}},
  journal      = {{SOIL RESEARCH}},
  keywords     = {{Gareh Bygone Plain,stony soils,soil-water,Iran,TIME-DOMAIN REFLECTOMETRY,ELECTROMAGNETIC DETERMINATION,ELECTRICAL-CONDUCTIVITY,BULK-DENSITY,TDR,CALIBRATION,MOISTURE,PERFORMANCE,TEMPERATURE,VALIDATION}},
  language     = {{eng}},
  number       = {{3}},
  pages        = {{360--370}},
  title        = {{Enhancing modelled water content by dielectric permittivity in stony soils}},
  url          = {{http://doi.org/10.1071/SR15154}},
  volume       = {{54}},
  year         = {{2016}},
}

Altmetric

View in Altmetric

Web of Science

Times cited: