Performance evaluation of laser diffraction and sieve-pipette techniques for grid-scale hydraulic characterization of lateritic vadose zones

The vadose zone plays a vital role in sustaining society and ecosystems on the Earth. The complexity of vadose zones, coupled with expensive and laborious field investigations pose a formidable challenge in their characterization, and hence they are neglected in hydrogeologic studies. This study demonstrates a novel scientific framework for the grid-scale hydraulic characterization of vadose zones using the laser diffraction (LD) technique and grain-size analysis (GSA) models. To demonstrate the proposed methodology, a total of 126 vadose zone samples and 95 core samples were collected from four layers across 33 grids in a lateritic terrain of Eastern India. Thereafter, the efficacy of the advanced LD technique and the conventional sieve-pipette analysis (SP) was compared by generating grain-size distribution curves (GSDCs) and estimating vadose-zone hydraulic conductivity (K vz ) by using selected GSA models at four different depths (0–50 cm, 50–100 cm, 100–150 cm, and 150–200 cm), followed by an uncertainty analysis. Finally, the estimates of K vz based on the SP-GSA and LD-GSA models were verified using field-measured K vz . The analysis of GSDCs for different depths revealed that the LD technique yields smoother S-shaped GSDCs than the SP technique. Consistently, low uncertainty was found in the K vz estimated by the LD-GSA models (0.0002–1.203 m/day) compared to the K vz estimated by the SP-GSA models (0.0008–2.171 m/day). The verification results based on statistical and graphical indicators revealed that the K vz yielded by most of the LD-GSA models for all layers of the vadose zone are in close agreement with the measured K vz values [MAE: 0.38 to 0.89 m/day and RMSE: 0.6 to 0.98 m/day]. On the other hand, the performance of SP-GSA models was also found to be reasonable, yielding K vz values with MAE and RMSE ranging from 0.36 to 1.45 m/day and 0.64–1.74 m/day, respectively, across all the vadose-zone layers. It is concluded that the LD technique provides more reliable GSDCs, whereas both SP and LD techniques yield reliable K vz (using suitably selected GSA models), thereby offering a useful scientific framework for estimating K vz at a grid scale. The proposed framework provides a basis for transferability and future evaluation at a larger spatial scale under different hydrogeologic and hydro-climatic conditions.