Search

Unlike conventional crowd simulations for what-if analysis, agent-based crowd simulations for real-time applications are an emerging research topic and an important tool for better crowd managements in smart cities. Recent studies have attempted to incorporate the real-time crowd observations into crowd simulations for real-time crowd forecasting and management; however, crowd flow forecasting considering individual-level microscopic interactions, especially for large crowds, is still challenging. Here, we present a method that incorporates crowd observation data to forecast a large crowd flow, including thousands of individuals, using a microscopic agent-based model. By sequentially estimating both the crowd state and the latent parameter behind the crowd flows from the aggregate crowd density observation with the particle filter algorithm, the present method estimates and forecasts the large crowd flow using agent-based simulations that incorporate observation data. Numerical experiments, including a realistic evacuation scenario with 5000 individuals, demonstrated that the present method could successfully provide reasonable crowd flow forecasting for different crowd scenarios, even with limited information on crowd movements. These results support the feasibility of real-time crowd flow forecasting and subsequent crowd management, even for large but microscopic crowd problems.

Next-generation high-performance computing (HPC) platform requires to run at higher speed with higher power consumption levels. Integrated voltage regulators (IVR) with a conversion ratio of 12 V to 1 V can help to reduce the power distribution network (PDN) impedance, increase the voltage conversion efficiency, and increase the regulation bandwidth. In this work, we present a new package embedded inductor array for multi-phase IVRs with a DC resistance as low as $\mathrm{22}.\mathrm{8}\ \mathrm{m}\Omega$ . Inductors with three magnetic materials are demonstrated. With one material a small signal inductance as high as 475 nH is obtained, suitable for 12 V to 1 V conversion at 2 MHz. Another material gives an inductance of 192 nH suitable for 12 V to 1 V IVRs at 5 MHz. Each inductor occupies less than 6.25 mm2 and are built with $\mathrm{400}\ \mu \mathrm{m}$ thick metal polymer composite magnetic sheets. The metric effective AC resistance per unit inductance or $R_{acx}$ is used to predict the inductor performance.

Rapid and accurate hazard forecasting is important for prompt evacuations and reducing casualties during natural disasters. In the decade since the 2011 Tohoku tsunami, various tsunami forecasting methods using real-time data have been proposed. However, rapid and accurate tsunami inundation forecasting in coastal areas remains challenging. Here, we propose a tsunami forecasting approach using convolutional neural networks (CNNs) for early warning. Numerical tsunami forecasting experiments for Tohoku demonstrated excellent performance with average maximum tsunami amplitude and tsunami arrival time forecasting errors of ~0.4 m and ~48 s, respectively, for 1,000 unknown synthetic tsunami scenarios. Our forecasting approach required only 0.004 s on average using a single CPU node. Moreover, the CNN trained on only synthetic tsunami scenarios provided reasonable inundation forecasts using actual observation data from the 2011 event, even with noisy inputs. These results verify the feasibility of AI-enabled tsunami forecasting for providing rapid and accurate early warnings., Rapid and accurate hazard prediction is important for prompt evacuation and casualty reduction during natural disasters. Here, the authors present an AI-enabled tsunami forecasting approach, which provided rapid and accurate early warnings.

We investigated tree development phenomena leading to a breakdown in silicone gel under the non-uniform electric field with a high-speed video camera. As a result, the tree developed stepwise, and its developing speed was similar with that in the silicone oil as subsonic speed. This phenomenon can be explained by that the tree (cavity) formed by discharges was a conductive condition which was supported by the result of electromagnetic waves during the tree development.

In Japan, seismic observation using modern instruments was started at the end of the 19th century, and the nationwide seismic observation network was developed in the early 20th century. Therefore, to study earthquakes that occurred before the start of modern seismic observation, it is necessary to conduct research and analysis based on historical documents.Since the end of the 6th century, there have been various types of historical documents describing earthquakes and their damage in Japan. Especially since the 10th century, historical source journals have described not only large earthquakes that caused great damage, but also small earthquake quakes. In Japanese history, pre-modern diaries were considered as semi-official archives intended for future reference and are used in various studies as reliable primary archives.The history diaries have the following features: (1) their descriptions are highly reliable because they were written immediately after the event or on the same day. (2) You can specify exactly where the history diaries were recorded. (3) The same author kept a diary for ten to several decades, providing continuous and stable information. (4) In some areas in Japan, historical diaries have existed almost continuously since the 10th century.In this study, many such history diaries are accumulated over a wide area in Japan, and the descriptions of felt earthquakes are extracted and digitized, and a database has been constructed so that the digitized information can be easily used for seismology research. In the prototype of the "Historical Diary-based Sensible Earthquake Database" introduced at the previous meeting, the diary historical data at 28 locations and 2,767 felt earthquakes were registered for four years from 1853 to 1856 (Nishiyama et. al., 2019).Since then, the database was released in a website, and the data has been expanded to 47 locations and 3,376 felt earthquake data for the same period. The database includes the date and possible occurrence time of the earthquake, the recording location, and the magnitude of shaking. In order to efficiently utilize this database for seismology research, we also have developed a graphic user interface (GUI) with the following functions.(a) A function to display the locations where shaking was recorded during the same time period on the same map, and plays the specified period as a moving image.(b) A function to examine whether a felt earthquake recorded at different locations on the same day is caused by the same earthquake. The possible time periods of each shaking are displayed side by side.This database and the newly developed GUI eliminate the need to read historical diaries, which requires the specialty of history. Utilizing digitized data, it is expected to promote seismology research, such as elucidation of long-term seismic activity including not only large-scale but also small- and medium-scale earthquakes.

Compared to evacuation behaviours, behaviours preceding evacuation including milling during tsunamis remain unclear. This study developed an analysis approach that evaluates the similarities in behaviour sequences throughout the evacuation process and investigated both the preceding behaviours and evacuation behaviours in Ishinomaki during the 2011 tsunami. The developed method was first applied to 101 detailed evacuation interview data and identified 16 major evacuation trip patterns in the study area during the 2011 tsunami. Then, we characterised the major patterns based on narratives and found key behavioural characteristics during tsunamis that both expanded and mitigated human loss. The analysis only considered the different destinations visited by people during evacuation processes; however, the proposed method successfully distilled evacuation processes with not only similar behavioural patterns but also similar thoughts or intentions. The collected behaviours revealed the complex evacuation processes during the tsunami, i.e., nearly half of the interviewees had two or more destinations in their evacuation trips. Evacuating alone was rarely observed, and many local residents first attempted to perform safety or damage confirmations, which caused evacuation delays. On the other hand, the results also indicated that this tendency of unification along with immediate school evacuation promoted evacuations in the study area by triggering household movements towards a hill, which unintentionally worked as a tsunami evacuation trip. We inferred that such secondary household movements potentially triggered other evacuation movements as tertiary effects by being sighted by other people. The findings of this study provide evidence suggesting that evacuation processes are largely affected by social structures or ties and highlight the importance of evacuation preparedness considering the revealed behaviour characteristics.

To contribute to tsunami early warning systems, we investigated the currently achievable speed of tsunami inundation simulations on a parallel computer as well as the benefits of high-resolution and faster than real-time inundation predictions. We found that a 5 m resolution inundation simulation can be 75 times faster than real time, requiring only 1.5 min to overview the inundation situation in Sendai City for the 2011 Tohoku tsunami. We developed a novel parallel tsunami model based on the well-known TUNAMI-N2 model and achieved 9.17 tera Floating-point Operations Per Second (FLOPS) on 9469 CPU cores. The present model can accurately hindcast the observed inundated regions of the 2011 Tohoku tsunami using tsunami source estimations of the tFISH and tFISH/RAPiD algorithms, which can be instantly derived from real-time observation data. The present high-resolution predictions can provide clear images of imminent hazards/disasters and can provide guidance for appropriate evacuation actions.

[1] Large-scale tsunami propagation simulations from the fault region to the coast are conducted using a three-dimensional (3-D) parallel unstructured mesh finite element code (Fluidity-ICOM). Unlike conventional 2-D approximation models, our tsunami model solves the full 3-D incompressible Navier-Stokes (NS) equations. The model is tested against analytical solutions to simple dispersive wave propagation problems. Comparisons of our 3-D NS model results with those from linear shallow water and linear dispersive wave models demonstrate that the 3-D NS model simulates the dispersion of very short wavelength components more accurately than the 2-D models. This improved accuracy is achieved using only a small number (three to five) of vertical layers in the mesh. The numerical error in the wave velocity compared with the linear wave theory is less than 3% up to kH = 40, where k is the wave number and H is the sea depth. The same 2-D and 3-D models are also used to simulate two earthquake-generated tsunamis off the coast of Japan: the 2004 off Kii peninsula and the 2011 off Tohoku tsunamis. The linear dispersive and NS models showed good agreement in the leading waves but differed especially in their near-source, short wavelength dispersive wave components. This is consistent with the results from earlier tests, suggesting that the 3-D NS simulations are more accurate. The computational performance on a parallel computer showed good scalability up to 512 cores. By using a combination of unstructured meshes and high-performance computers, highly accurate 3-D tsunami simulations can be conducted in a practical timescale.

To evaluate tsunami hazards with strong locality in urban areas, this study developed a novel tsunami inundation model based on nonlinear shallow water wave equations and a porous body model (PBM). By applying a kinematic boundary condition that includes both porosity and surface permeability of the porous medium, the proposed model could accurately incorporate geometric effects such as the flow anisotropy caused by the distributions of buildings. The proposed PBM demonstrated as good accuracy for the inundation heights around buildings near the coastline as with a conventional three-dimensional simulation with high resolution. In addition, the model showed its capability to reproduce a tsunami’s essential behaviors in urban areas. In particular, the amplification effect of flow velocity along straight roads surrounded by buildings was reasonably reproduced. It can be expected that the present model can become a useful tool to accurately evaluate the tsunami risks in urban areas.

In three-dimensional geodynamo simulations, the most widely used method is the spectral transform method in which variables are expanded by spherical harmonics. This method is suitable because it has good accuracy and enables appropriate treatment of boundary conditions of the magnetic field. However, it causes a heavy drag especially on the high-resolution calculation because of the large amount of computation required for the spectral transform in the latitudinal coordinate (the Legendre transform). Therefore, we introduce a method for high-resolution geodynamo simulations to decrease the computation time by avoiding the Legendre transform and performing only the Fourier transform in longitude. In this Fourier transform method, the resulting spectral equations are solved in a meridional plane using a finite difference technique. We verify the applicability of the Fourier transform method for geodynamo simulations by solving the dynamo benchmark problem [Christensen, U.R., Aubert, J., Cardin, P., Dormy, E., Gibbons, S., Glatzmaier, G.A., Grote, E., Honkura, Y., Jones, C., Kono, M., Matsushima, M., Sakuraba, A., Takahashi, F., Tilgner, A., Wicht, J., Zhang, K., 2001. A numerical dynamo benchmark. Phys. Earth Planet. Inter. 128, 25–34] and some cases of higher Rayleigh (up to 4 times critical) and lower Ekman ( 1 0 − 4 at minimum) numbers than the benchmark problem. The results confirm the validity of this method, but suggest that its superiority over spherical harmonic models is limited to such a very high-resolution calculation as is realizable in near future.

In order to help combat climate change, Japan plans to introduce a large-scale photovoltaic (PV) power generation capacity. In this paper, we focus on reactive power in large-scale PV systems, first proposing a two-layered shunt capacitor (SC) scheduling method using multi-objective particle swarm optimization (MOPSO) and dynamic programming (DP). Using MOPSO, cost efficiency and reliability in power system planning and operation can be more accurately analyzed, while DP allows switching times to be assessed. To control reactive power flows caused by changing PV output, we then propose an effective static var compensator (SVC) control method, consider SC and transformer load tap charger (LTC) efficiencies, for use in situations in which PV outputs fluctuate sharply over short time intervals. Based on simulations run on a modified Ward-Hale 6-bus system to verify the validity of the proposed method, we confirm its capability to control power systems with a large-scale penetration over a normal voltage range.

The double heating technique of the Shaw method with low-temperature demagnetisation (LTD-DHT Shaw method) for determination of geomagnetic palaeointensity is applied to samples exhibiting high-temperature oxidation states from the Kilauea 1970 lava, Hawaii Island. Results are obtained for 11 of the 12 specimens prepared from five block samples, yielding an average palaeointensity of 38.2 ± 2.8 T( N = 11, ±1σ). This value is consistent with the expected value determined from DGRF 1970 (35.8T), and does not appear to be significantly dependent on the high-temperature oxidation state. Coe’s version of the Thellier method was also applied to nine specimens prepared from the same block samples, and successful results were obtained for seven specimens, giving an average palaeointensity of 43.2 ± 8.4 T( N = 7). Although this average is statistically consistent with the expected value, the results include erroneously high palaeointensities (52.1 and 55.4 T) for specimens from one block sample of intermediate high-temperature oxidation state. The present results therefore reinforce the broader applicability of the LTD-DHT Shaw method for samples with high-temperature oxidation states compared with Coe’s version of the Thellier method. It is also shown that the samples yielding overestimated Thellier palaeointensities tend to fall close to the mixing line between single-domain (and/or pseudo-single-domain) and multidomain components on the Day plot. This relationship may be useful as a pre-selection technique for application of the Thellier method. © 2004 Elsevier B.V. All rights reserved.

We have performed a palaeointensity study of the Oshima 1986 lava in order to examine the reliability of the Thellier and LTD-DHT Shaw methods with special reference to high-temperature oxidation states of magnetic grains. Coe’s version of the Thellier method was applied to fifteen specimens from five block samples. The palaeointensities determined by this Thellier method ranged from 44.8 to 59.0 μT with an average of 51.0 ± 4.1 μT ( N = 15). They are systematically higher than the expected intensity (45.5 μT from the DGRF 1985). Since the in situ magnetic survey at the sampling site indicates no strong local magnetic anomaly, these high palaeointensities are overestimation, probably due to some rock magnetic property and/or some thermal alteration during the experiments. Microscopic observations show that the samples, where some titanomagnetites surrounding vesicles were classified with intermediate oxidation indices of III–V, gave 10–30% higher palaeointensity values. Therefore, the erroneously high Thellier palaeointensities have a relation to the high-temperature oxidation state. For eight of the fifteen specimens used in the Thellier experiments, the applied DC field was perpendicular to the primary NRM component for quantitative evaluation of laboratory-induced chemical remanent magnetisation (CRM). Directional changes of NRM thermally demagnetised during the Thellier experiments indicate that most of the samples more or less acquired CRM due to laboratory heating suggesting that some magnetic phase with high blocking temperatures was produced. For the specimens of intermediate oxidation level, the CRM appears to have been gradually acquired at relatively low temperatures (≤400 °C). It would appear to be difficult to detect this gradual thermal alteration with the standard pTRM check, weak-field susceptibility measurements and hysteresis parameter changes. We also applied the double heating technique of the Shaw method combined with low temperature demagnetisation (LTD-DHT Shaw method) to seven specimens from the same block samples. The palaeointensities determined by the LTD-DHT Shaw method yielded an average of 46.4 ± 4.7 μT ( N = 6), which agrees with the expected intensity indicating no dependence on the high-temperature oxidation states.

The regiochemistry of the Michael addition of thiols to unsymmetrical fumaric derivatives was investigated. Conjugate addition of thiols to unsymmetrical fumaric diester was well controlled by the presence of lithium cation and one of the two possible regioisomers was prepared in a highly selective manner. Fumaric ester amides underwent the regioselective Michael addition that was controlled by the presence or absence of the base; either of the regioisomers was prepared as an almost diastereomerically pure form. The present control of the regiochemistry can be explained by the factors of change of active site for the addition by the coordination or non-coordination of proton or lithium cation to the carbonyls. To clarify the origin of the regioselectivity, the relative rates of the conjugate addition of thiol to acrylate derivatives were measured under competitive conditions. Ethyl acrylate reacted with thiol faster than tert-butyl acrylate and the rate difference was enhanced by the presence of lithium cation. In the presence of base, ethyl acrylate gave the adducts much faster than acrylamide, while under non-basic conditions acrylamide showed higher reactivity than the ester. This regioselectivity was also observed in the Michael/aldol reaction and multi-substituted γ-butyrolactones were prepared in a stereoselective manner. The thio groups introduced here served as a leaving group and a convenient stereoselective synthesis of β-, γ- and δ-lactams was developed.

Macrophage cells differentiate into killer macrophages, named M1 macrophages, that effectively eliminate cancer cells by generating cytokines. We examined the feasibility of a lung cancer therapy using lipopolysaccharide (LPS), which triggers this differentiation. It is expected that the delivery via inhalation of LPS directly into the lungs, where alveolar macrophages reside abundantly, would be effective at minimizing the possible toxic effect of LPS.We determined the effects of intratracheal insufflation of LPS on (i) the exudation of lactate dehydrogenase, (ii) generation of tumor necrosis factor-α and interleukin-12, and (iii) tumor metastases in Lewis lung carcinoma-bearing C57BL/6 mice.Pulmonary insufflation of LPS resulted in a consistent accumulation of tumor necrosis factor-α, and transient increase in interleukin-12 without significant release of lactate dehydrogenase from the lung cells. In addition, a significant antitumor effect of LPS was observed; and this antitumor effect was potentiated by combination of LPS with the antitumor agent cyclophosphamide.Pulmonary inhalation of LPS combined with a chemotherapeutic agent is a promising approach to lung cancer therapy.

Waste fiber-reinforced plastics (FRPs), which are formidable composite plastics for chemical treatment, can be efficiently depolymerized by treatment with supercritical methanol in the presence of catalytic amounts of 4-(dimethylamino)pyridine (DMAP), which is a well-known catalyst for the formation of esters and amides. This novel depolymerization reaction also provides ready separation of the depolymerized products into three components: methanol-soluble oil, CHCl3-soluble solid, and an inorganic residue. Thus, almost complete decomposition of FRP as well as useful separation of the decomposed products was achieved with the present method, and the latter were ready for chemical recycling. Investigation of the reaction profile revealed that the reaction rate increased as the amount of DMAP increased. Unfortunately, DMAP employed in the reaction could not be recovered because of its decomposition catalyzed by dimethyl phthalate, a depolymerized product from waste FRP, under the reaction conditions employed.

We examined the decomposition of polyamides such as nylon-6 and nylon-12 by using supercritical MeOH as the reaction media. The treatment of waste nylon-6 with supercritical MeOH resulted in smooth depolymerization, forming caprolactam as the first product which was then converted to a mixture of methyl 6-hydroxycapronate and methyl 5-hexenoate in a ratio of approximately 1 : 1. The reaction progress was traced using gas chromatography (GC) analyses, and precise product distribution was estimated. During the decomposition of nylon-6, N-methylcaprolactam and methyl 6-(N,N-dimethylamino)capronate were detected as intermediates. The sum of the all detectable products and intermediates exceeded 80%. In addition, we examined the decomposition reaction initiating from caprolactam, N-methylcaprolactam, and methylN,N-dimethylcapronate under similar reaction conditions, and observed that the final two products were formed in similar yields and ratios. Kinetic analyses by using a simulation study based on the experimental data were performed, and kinetic parameters for each step were estimated. Nylon-12 underwent similar conversion to produce methyl 12-hydroxydodecanoate in good yield. Because methyl ω-hydroxyalkanoate is known to be an important intermediate in the chemical industry, the present method has the potential for producing valuable compounds from waste material. Thus, the first upgrade in the chemical recycling of plastics was accomplished.