Abstract In this study, earthquake rupture models for future mega-thrust earthquakes in the Nankai–Tonankai subduction zone are developed by incorporating the main characteristics of inverted source models of the 2011 Tohoku earthquake. These scenario ruptures also account for key features of the national tsunami source model for the Nankai–Tonankai earthquake by the Central Disaster Management Council of the Japanese Government. The source models capture a wide range of realistic slip distributions and kinematic rupture processes, reflecting the current best understanding of what may happen due to a future mega-earthquake in the Nankai–Tonankai Trough, and therefore are useful for conducting probabilistic tsunami hazard and risk analysis. A large suite of scenario rupture models is then used to investigate the variability of tsunami effects in coastal areas, such as offshore tsunami wave heights and onshore inundation depths, due to realistic variations in source characteristics. Such investigations are particularly valuable for tsunami hazard mapping and evacuation planning in municipalities along the Nankai–Tonankai coast.

Geothermal energy depends on high subsurface temperature, adequate permeability and fracture volume, and accessible groundwater supply to support heat exchange with surrounding rock. Some regions may have adequate thermal resources but lack the necessary permeability or deep circulating water. Exploitation of such areas for geothermal energy could occur if permeability can be enhanced enough to provide the necessary heat exchange. These improvements to the geothermal reservoir would produce what is termed an “enhanced geothermal system” (EGS). The Snake River Plain (SRP) in southern Idaho is a geological region with high heat flux (~110 mW/m 2 ) that has been recommended as an EGS target. In this study, we consider how the geologic and thermal history of the SRP might influence its EGS potential. We describe the fracture distribution (mean = 28.63 fractures/10 m) in a welded tuff core recovered from one of the few deep boreholes located on the SRP and provide a preliminary discussion of the likely geomechanical behavior under in situ stress. Spatial autocorrelation of fracture features is defined with geostatistical techniques and used in a stochastic simulation of possible structures in other welded tuff reservoirs. Autocorrelation scales for the continuous date are on the order of 70 meters with high subsample scale variability (56 m). Results should aid in designing criteria for a hydraulic fracturing plan that would augment the permeability and connectivity of an SRP reservoir's preexisting fracture network.

Quantifying erosion rates and processes remains a central focus of studying the Earth's surface. Measurement of in situ–produced cosmogenic radionuclides (CRNs) enables a level of quantification that would otherwise be impossible or fraught with uncertainty and expense. Remarkable success stories punctuate the field over the last decade as CRN-based methodologies are pushed to new limits. Inherent to all is an assumption of steady-state rates and processes. This paper focuses on the use of cosmogenic 10 Be and 26 Al, extracted from quartz in bedrock, saprolite, and detrital material to quantify sediment production or erosion rates and processes. Previous results from two very different field areas are reviewed to highlight the potential for non-steady-state processes in shaping soil-mantled landscapes. With this potential in mind, a numerical model is presented, following a review of the CRN conceptual framework, to test the effects of non-steady-state erosion rates and processes on CRN concentrations. Results from this model focus on 10 Be concentrations accumulated under modeled variations in erosion rates with different ranges, frequencies, and styles of variability. In general, the higher the maximum erosion rate, the higher the impact on the CRN concentration and, therefore, the more likely that point measurements will capture the variable signal. Conversely, the higher the frequency of erosional variation, the less likely point measures are to accurately determine rates, but the closer the inferred rate is to the mean of the long-term erosion rate. Modeling results are applicable for point-specific erosion rates, but endorse the catchment-averaged approach for determining average rates. Potentially large uncertainties emphasize the need for careful sample selection, with adequate numbers of samples collected for quantifying the processes eroding the land. The two field examples show how analyzing enough samples can define a clear soil production function despite the potential for non-steady-state processes. The model presented here is ready for application to catchment-averaged processes, as well as modeling the role of muons in variable erosion rate scenarios.