Tomijiro Kubota

Three-dimensional inversion of in-line resistivity data for monitoring a groundwater recharge experiment in a pyroclastic plateau

An artificial groundwater recharge experiment was conducted in a pyroclastic plateau in Kagoshima Prefecture in Japan, and time-lapse electrical resistivity data were collected to monitor the recharge process. In the experiment, time-efficient in-line resistivity surveys were performed along four intersecting lines, because a large amount of water was released from two recharge areas and a relatively fast migration of water into the vadose zone was expected. The migration of recharged water may be estimated from changes in electrical resistivity because resistivity in the vadose zone is largely controlled by water saturation variations there. The geological setting at the experiment site was interpreted from the resistivity distribution inverted from the in-line survey data, which were obtained before the recharge experiment. The resistivity distribution showed an approximately layered structure, which could be correlated with four borehole logs in the area. Three-dimensional (3D) distributions of the resistivity change ratio were derived through constrained nonlinear ratio inversion. Three-dimensional inversion of the in-line resistivity data was more suitable than two-dimensional inversion to describe the 3D phenomena associated with groundwater recharge. During the recharge experiment, the zones of decreased resistivity shifted with time, indicating non-uniform penetration of water from the recharge areas into the ground and a horizontal flow of the recharged water, especially in the secondary Shirasu layer, which comprises lacustrine or marine sediments of pyroclastic origin. These interpretations agree with the variation in water content observed in a borehole. An artificial groundwater recharge experiment was conducted in a pyroclastic plateau in Kagoshima Prefecture in Japan, and time-efficient in-line resistivity surveys were performed along four intersecting lines. The zones of decreased resistivity shifted with time, indicating non-uniform penetration of water from the recharge areas and a horizontal flow of the recharged water.

Scenario analysis for reduction of pollutant load discharged from a watershed by recycling of treated water for irrigation

A model in which a river model was layered on a distributed model (double-layered model) was developed to analyse the transport of water and pollutants (nitrogen, phosphorus, and BOD as organic matter) in watersheds and rivers. The model was applied to the watershed of Abragafuchi Lake, Japan, where serious water pollution has occurred over three decades, and the applicability of the model was demonstrated. Scenarios of recycling of sewage treated-water into agriculture to reduce pollutant load discharged into the lake were analysed. The results showed that irrigating paddy fields with the sewage-treated water could contribute to conserving water and reducing pollutant load, with reduction rate in BOD, nitrogen, and phosphorus ranging from 6%-36%, 16%-46%, and 18%-51%, respectively. Particularly, the results indicated that, irrigating paddy fields with the treated water during non-cropping periods and the accompanying reduction in withdrawn water from the river were more effective in reducing pollutant loads discharged into the lake. Further study is required on the effect of recycled water on crop cultivation and soil conditions for safe implementation.

Estimating high hydraulic conductivity locations through a 3D simulation of water flow in soil and a resistivity survey

In this study, we propose a method to estimate high hydraulic conductivity locations that uses 3D simulation of soil water flow and in-line resistivity survey data acquired during a groundwater recharge experiment, and we apply this method to numerical and field experiments. The high hydraulic conductivity locations are estimated from a combination of field-observed and simulated apparent resistivities using the following simple steps. (1) Assuming that high hydraulic conductivity zones exist in the first layer, simulations of saturated-unsaturated seepage are conducted for several possible water-flow models that have high hydraulic conductivity zones in different locations. (2) The simulated volumetric water contents are converted into bulk resistivities, which are used to produce apparent resistivity data through simulation of a resistivity survey. (3) The differences between the simulated apparent resistivities and the field-observed data are examined, and the best-fit hydraulic conductivity model is identified by minimising the above differences. In the numerical experiment, 3D inversion of the simulated resistivity survey provides an image of the preferential flow, although the infiltration locations are unclear. Comparing the field model with the possible models, the high hydraulic conductivity location in the field model corresponds to the high hydraulic conductivity location in the possible model with the minimum errors. In the field, an in-line resistivity survey was conducted during a groundwater recharge experiment on a pyroclastic plateau. The 3D inversion of the in-line resistivity survey data provides an image of the preferential flow. Comparing the field apparent resistivity data with the simulated apparent resistivity data, the high hydraulic conductivity location of the possible model that provides the minimum error corresponds to the recharge water range, whereas the hydraulic conductivity location of the possible model that gives the maximum errors corresponds to ranges with no recharge water. These results indicate that it is possible to estimate high hydraulic conductivity locations using 3D simulations of the soil water flow and a resistivity survey. We propose a simple method for estimating high hydraulic conductivity locations. The proposed method uses the 3D simulations of soil water flow and resistivity survey during a groundwater recharge experiment. Results of numerical and field experiments indicate that the proposed method estimates the high hydraulic conductivity locations more precisely compared with 3D inversion of in-line data.