Multiple tracer study in Horonobe, northern Hokkaido, Japan: 1. Residence time estimation based on multiple environmental tracers and lumped parameter models

Abstract

A multiple tracer study was conducted in the coastal plain of Horonobe, northern Hokkaido, Japan. To accomplish reliable estimates of groundwater residence time, multiple tracer approach, employing both gaseous and nongaseous environmental tracers, and lumped parameter models (LPMs) were applied in combination. Environmental tracers including 3H, 3H/3He, 36Cl, and SF6 were analyzed and compared with values estimated from LPMs to select the most appropriate tracer and models to describe groundwater flow regime in the study area. Time series of 36Cl/Cl input in the study area were reconstructed and used in the LPMs. From comparison of the analyzed and modeled results, an index of 3H/3He was selected as the most appropriate for investigation of the study area. The relationship between the 3H/3He index and 36Cl/Cl indicated reliability of the 3H/3He index, but 36Cl/Cl values of the samples were lower than the models. This discrepancy was caused by underestimation of Cl− concentration of recharging water in the calculation of 36Cl/Cl input. SF6 mixing ratios of the samples were slightly higher than the models and it suggested involvement of terrigenic SF6. Three well groups, northern Shimonuma wells (NSW), southern Shimonuma wells (SSW), and Hamasato wells (HW), were respectively consistent with exponential piston flow model (EPM), binary mixing model (BMM), and exponential mixing model (EMM). Relational expressions between the 3H/3He index and mean residence time or mixing fraction of young groundwater (assumed to contain no tritiogenic 3He) with old groundwater (all 3H decayed to tritiogenic 3He) were obtained based on the models. The mean residence time determined from the 3H/3He index and the expressions were 19–78 years for NSW and 1–648 years for HW. The mixing fraction of young groundwater was less than 10% at all SSW. Nongaseous tracers, including 36Cl are not influenced by factors specific for gaseous tracers, such as excess air and degassing. Analyses of multiple tracers, including both gaseous and nongaseous tracers, and comparison using LPMs were effective to show plausibility of the estimated residence time, and they contributed to understanding of groundwater flow regime in the study area.