Dielectric constant of ice in natural rivers

Abstract

Ground-penetrating radar (GPR) is a non-contact measurement method widely used to determine ice thickness, which is an important hydrological parameter of rivers in winter. The accuracy of GPR measurements of ice thicknesses is highly dependent on the dielectric constant value of ice. Research on the dielectric constant of ice has mainly focused on pure ice; however, pure ice is not commonly found in natural rivers. In fact, hydrodynamic and meteorological factors, as well as boundary conditions, affect ice in natural rivers, which can be treated as a mixture of pure ice, air bubbles, sediment, and pure water. Therefore, the dielectric constant of ice in natural rivers often differs from the theoretical value of pure ice, a fact that has often been neglected in prototype measurements. During eight years of prototype monitoring in the Heilong River and Yellow River (China), it was found that variations in the dielectric constants of ice in the two rivers produced errors in the GPR ice thickness measurements. In this study, we measure the dielectric constant of ice at 158 measurement points across more than 20 river sections, and we present a recommended range of values for the dielectric constant. Regarding media, such as air bubbles, sediment, and water in natural river ice, we analyze sand (both dry and wet) in ice-bound sediment and the effects of air bubbles, sediment, and water upon the ice's dielectric constant. Using temperature and voltage measuring rods installed in the ice, we demonstrate that the water content in the ice differs significantly according to the freeze-up and breakup periods. The present study recommends guidelines including suitable timing, ice-condition, and dielectric-constant reference values for GPR-based ice thickness measurements to improve the measurement accuracy. Thus, this work can serve as a reference for future applications in similar rivers.