Using basic metrics to analyze high-resolution temperature data in the subsurface

Abstract

Time-series temperature data can be summarized to provide valuable information on spatial variation in subsurface flow, using simple metrics. Such computationally light analysis is often discounted in favor of more complex models. However, this study demonstrates the merits of summarizing high-resolution temperature data, obtained from a fiber optic cable installation at several depths within a water delivery channel, into daily amplitudes and mean temperatures. These results are compared to fluid flux estimates from a one-dimensional (1D) advection-conduction model and to the results of a previous study that used a full three-dimensional (3D) model. At a depth of 0.1 m below the channel, plots of amplitude suggested areas of advective water movement (as confirmed by the 1D and 3D models). Due to lack of diurnal signal at depths below 0.1 m, mean temperature was better able to identify probable areas of water movement at depths of 0.25–0.5 m below the channel. The high density of measurements provided a 3D picture of temperature change over time within the study reach, and would be suitable for long-term monitoring in man-made environments such as constructed wetlands, recharge basins, and water-delivery channels, where a firm understanding of spatial and temporal variation in infiltration is imperative for optimal functioning.