[A-HW22] 流域生態系における物質輸送と循環：源流から沿岸まで

[AHW22-P19] Physically Based Groundwater Flow Simulation using Tracer-aided model in Kumamoto Region, Japan

*A T M Sakiur Rahman¹、Takahiro Hosono^2,3、Yasuhiro Tawara⁴、Youichi Fukuoka⁴、Aurelien Hazart、Jun Shimada³ (1.Postdoctotal researcher, Kumamoto University、2.Faculty of Advanced Science and Technology, Kumamoto University、3.International Research Organization for Advanced Science and Technology, Kumamoto University、4.Geosphere Environmental Technology Corporation, NCO Kanda, Awajicho Building 3F, 2-1 Kanda, Awajicho, Chiyoda-ku, Tokyo 101-0063, Japan)

キーワード：Water Modeling, Kumamoto, Tracer aided model

Integrated watershed modeling techniques are being applied to examine the surface and subsurface water interactions in recent years. Generally, modeling approaches heavily rely on the best fit of the hydrograph which cannot alone describe entire catchment water dynamics. For holistic understanding of hydrological processes, it is necessary to incorporate and simultaneously simulate tracers which can provide important inferences about water ages, flow paths and origin. Hence, this study incorporated multiple tracers like tritium (³H), Krypton-85 (⁸⁵Kr), and groundwater temperature in model and seamlessly simulated tracers coupled with surface and subsurface flows using a fully distributed physically based modeling approach (GETFLOWS) for detailed characterizations of Kumamoto water regimes in southern Japan. The first model developed using the regular hydrometric parameters could not characterize isotopic compositions and groundwater temperature, though it showed acceptable model performance for simulating surface water and groundwater hydrographs for several observation stations located in the area. The second model calibrated by integrating tracers with regular parameters well characterized hydrographs (surface water, groundwater), isotopic compositions and groundwater temperature. Hence, the tracer-aided model was used for simulating groundwater storage, flow paths and groundwater age which showed a close agreement with estimation of water ages using isotopic approaches. The findings of this study clearly indicated that multiple tracers aided can better visualize the subsurface water dynamics than hydrometric model. It is expected that the learning framework presented in this study will be beneficial for field hydrologists and modelers to make joint efforts to build robust models for hydrological processes.