This project is a comprehensive interdisciplinary study of the March 11 initial M=9 and M=7.9 earthquakes, the resulting tsunami wave generation, propagation and coastal inundation along northern Honshu Island, and the initial pathways and changes in Cs-237 concentrations as it enters the coastal waters at the Fukushima Daiichi nuclear facility and spreads across the shelf to deeper water. The approach is to use a combination of advanced seismic and nested coupled atmospheric/3-D ocean circulation numerical models plus available field measurements to simulate these processes starting with the initial March 11 M=9 earthquake bottom movement through April 12. During this 33-day simulation, the Cs-137 source concentration levels peaked and decreased towards the increasing coastal and off-shelf concentration levels, indicative of cross-shelf transport and shelf-ocean exchange processes, with a potential sedimentation loss and biological accumulation in the near-shore region. Detailed descriptions of the different model simulations, the resulting ocean circulation and water property output fields and initial analysis will be uploaded to a project website on a frequent basis for use by others interested in coastal physical and bio/chemical processes in the study area and as initial conditions for studies of the long-term spread of Cs-137 and other radionuclides in the Pacific Ocean.
Intellectual Merit: The March 11 earthquakes, tsunami waves, coastal inundation, and initial release of Cs-137 into the ocean cover a wide range of time (from seconds to a month) and space (meters to 100's of km) scales. The multi-scale modeling approach with advanced models should produce a comprehensive and integrated description and understanding of the key physical processes involved and an independent assessment of the initial fate and spread of Cs-137 and its impact on the coastal ecosystem within the RAPID grant period.
Broader Impacts: This study will foster U.S-Japanese collaboration in several areas of ocean sciences (marine geophysics, physical oceanography, and bio-chemistry related to Cs-137 and other released radionuclides). The team includes one Japanese PhD student, and it is anticipated that the models and model results posted on a website will be used by researchers, students, and others as the study progresses. One outcome of this study will be a tested new combined earthquake/3-D ocean model system that can be used by researchers and coastal planners for assessing potential tsunami flooding from future earthquakes in the megathrust zones east of Japan. This system can be applied to other earthquake and or tsunami-prone areas.
A team of US and Japanese investigators applied a high-resolution nested global-Japan coastal coupled atmosphere-ocean numerical model system to simulate the March 11, 2011 T?hoku M9 and M7.9 earthquake-induced tsunami wave generation and coastal inundation along the northeastern coast of Honshu Island in the western Pacific Ocean. The ocean model used was the Finite-Volume Community Ocean Model (FVCOM) that uses an unstructured triangular grid in the horizontal and a generalized terrain-following vertical coordinate. This allowed the earthquake-tsunami-inundation model domain to be configured with a horizontal resolution up to 5 m in regions of complex inner-shelf-land topography and around the Fukushima Dai-ichi Nuclear Power Plant (FDNNP) and lower resolution off shore. Under real oceanic conditions with inclusion of the Kuroshio, tides and wind forcing, we conducted model experiments using the seafloor change predicted inversely by five seismic rupture models as initial conditions, and the results were evaluated via a) observed sea level changes at tidal gauges and b) field survey estimates of the locations and spatial extents of the observed flooding. The good agreement of the model simulations with observations demonstrates that with sufficient resolution of coastal geometry for a given accurate local bathymetry, an unstructured-grid nested-domain ocean model like FVCOM is capable of reproducing the Japan March 11, 2011 tsunami wave and resulting coastal inundation. This study also provides an alternative way to evaluate the seismic rupture models. Significant differences in sea level and inundation areas were found in the seismic model-data comparisons, which suggest that it is critical to capture the earthquake moment and spatial distribution of the seafloor change to accurately predict the tsunami and flooding using an ocean model. With the appropriate seismic rupture model and sufficiently high spatial resolution around the FDNNP, our model system accurately predicted the flooding observed at the facility and identified the key role of the breakwaters in the flooding process. While the initial tsunami waves broke over the entrance breakwaters, much of the wave energy and transport was diverted by the submerged part of the breakwaters around the entrance towards the coast north and south of the facility and up onto land, helping to cause the flooding from the land side. Based on the model’s success in reproducing the March 11 observed tsunami and coastal inundation, tracer experiments were conducted with differing model grid resolution to assess the initial spread of Cs-137 over the east shelf of Japan. The Cs-137 was tracked as a conservative tracer in the three-dimensional model flow field over the period March 26 -August 31, 2011. The results clearly show that for the same Cs-137 discharge, the model-predicted spreading of Cs-137 was sensitive not only to model resolution but also the FDNPP seawall structure. A coarse-resolution (about 2 km) model simulation led to an overestimation of lateral diffusion and thus faster dispersion of Cs-137 from the coast to the deep ocean, while advective processes played a more significant role when the model resolution at and around the FDNPP was refined to about 5 m. By resolving the pathways from the leaking source to the southern and northern discharge canals, the high-resolution model better predicted the Cs-137 spreading in the inner shelf where in situ measurements were made 30 km off the coast. The overestimation of Cs-137 concentration near the coast is thought to be due to the omission of sedimentation processes in the model, which was evident in sediment measurements taken after the accident.
