Tropical Storm Pabuk in 2019, the most severe storm event in the past two decades, provided a unique opportunity to assess storm surge impact in the Gulf of Thailand. Employing the COMCOT-SURGE model with a two-layer nested-grid configuration, this study examined storm surges from offshore to nearshore driven by the ERA5 reanalysis winds and the 1980 Holland wind model. The spatial and temporal variations of storm surges generated by Tropical Storm Pabuk were both investigated. The ERA5-driven case indicated severe storm surges in the coastal regions of Thailand and matched well with the tide-gauge observations in the timing and amplitude of peak surges. However, the Holland wind model case showed the maximum storm surges shifted southward and poorly agreed with the measured surges. This discrepancy was attributed to the absence of environmental winds in the parametric wind model. Moreover, the effect of the radius of maximum winds on storm surges was tested. To further explore potential storm surges in the Gulf of Thailand, this study focused on the variations in storm tracks along the latitudinal axis and associated storm surges. Among the proposed storm track scenarios, the one landing more northward in Thailand than Pabuk represented the most severe case to the Thailand coast, followed by Pabuk’s track. In the proposed scenarios, cities such as Chumphon, Surat Thani, and Nakhon Si Thammarat were found to face higher storm surge risks than other regions.

Objectives: Total Parenteral Nutrition (TPN) is a lifesaving therapy providing intravenous nutrition when enteral use is contraindicated. Prolonged TPN can induce hepatic complications (i.e., PN associated liver dysfunction [PNALD]) affecting 40% of patients. TPN is often infused at a steady rate which contributes to PNALD. Restricting TPN infusion to the evening can delay the condition, raising the possibility that specific infusion patterns may prevent PNALD. We hypothesized that, compared to steady-rate infusions, a sinusoidal infusion schedule matched to diurnal consumption patterns in mice, would attenuate PNALD. Methods: 17 male C57B/6 mice aged 9-11 weeks were randomly assigned to 3 groups: chow (n=5), steady rate (n=6), and sinusoidal (n=6). The jugular vein was cannulated with 0.12 mm diameter silicon catheter attached to TPN syringe on a programmable pump. Mice were housed in a metabolic chamber (Promethion Core, Sable Systems International, Las Vegas NV) at 30°C and infused with 5.5 mL/d of TPN formula (21.5% dextrose, 5.6% amino acids, 2% Intralipid-20) at a steady or sinusoidal rate for 5 days. Mice were euthanized and tissues were weighed, processed, and snap frozen. Average daily respiratory exchange rate (RER) was extracted for analysis and qPCR was conducted on hepatic metabolic targets. Results: Two-way ANOVA revealed significant group and time effects on RER (F = 1.19, r2 = 0.14, p = 0.004) with a mean difference of 0.063 (95% CI = 0.058 – 0.068, p < 0.001) between steady (mean = 0.96, min = 0.88, max = 1.11) and sinusoidal groups (mean = 0.90, min = 0.78, max = 0.96). PCR revealed trending differential expression of Srebp1 (F = 2.053, p = 0.18) and Acaca (F = 2.041, p = 0.18) driven by relatively higher expression under steady rate infusion. There were no significant differences in metabolic tissue weights or intestine length between groups. Conclusions: Sinusoidal infusions, matched to ad-libitum consumption patterns, are tolerated and do not induce complications in murine TPN. This novel approach may improve lipid utilization and match with circadian rhythm of metabolism. Further investigations are underway to establish this mouse model of human PNALD. Funding Sources: Metabolism and Nutrition Training Program T32 (DK 007665)Wisconsin Dairy Innovation Hub Start-up Funding.

The Pencil Beam Scanning (PBS) technique in modern particle therapy offers a highly conformal dose distribution but poses challenges due to the interplay effect, an interaction between respiration-induced organ movement and PBS. This study evaluates the effectiveness of different volumetric rescanning strategies in mitigating this effect in liver cancer proton therapy. We used a Geant4-based Monte Carlo simulation toolkit, 'TOPAS,' and an image registration toolbox, 'Elastix,' to calculate 4D dose distributions from 5 patients' four-dimensional computed tomography (4DCT). We analyzed the homogeneity index (HI) value of the Clinical Tumor Volume (CTV) at different rescan numbers and treatment times. Our results indicate that dose homogeneity stabilizes at a low point after a week of treatment, implying that both rescanning and fractionation treatments help mitigate the interplay effect. Notably, an increase in the number of rescans doesn't significantly reduce the mean dose to normal tissue but effectively prevents high localized doses to tissue adjacent to the CTV. Rescanning techniques, based on statistical averaging, require no extra equipment or patient cooperation, making them widely accessible. However, the number of rescans, tumor location, diaphragm movement, and treatment fractionation significantly influence their effectiveness. Therefore, deciding the number of rescans should involve considering the number of beams, treatment fraction size, and total delivery time to avoid unnecessary treatment extension without significant clinical benefits. The results showed that 2-3 rescans are more clinically suitable for liver cancer patients undergoing proton therapy.

This study presents a numerical tool for calculating storm surges from offshore, nearshore, and coastal regions using the finite-difference method, two-way grid-nesting function in time and space, and a moving boundary scheme without any numerical filter adopted. The validation of the solitary wave runup on a circular island showed the perfect matches between the model results and measurements for the free surface elevations and runup heights. After the benchmark problem validation, the 2013 Super Typhoon Haiyan event was selected to showcase the storm surge calculations with coastal inundation and flood depths in Tacloban. The catastrophic storm surges of about 8 m and wider, storm-induced inundation due to the Super Typhoon Haiyan were found in the Tacloban Airport, corresponding to the findings from the field survey. In addition, the anti-clockwise, storm-induced currents were explored inside of Cancabato Bay. Moreover, the effect of the nonlinear advection terms with the fixed and moving shoreline and the parallel efficiency were investigated. By presenting a storm surge model for calculating storm surges, inundation areas, and flood depths with the model validation and case study, this study hopes to provide a convenient and efficient numerical tool for forecasting and disaster assessment under a potential severe tropical storm with climate change.

Introduction: The gut microbiota has been implicated in various heart diseases by producing metabolites that modulate host immunity and metabolism. In our previous study, we found that mice with dysbiosis suffered more from cardiac rupture, resulting in higher mortality than mice possessing commensal gut flora after myocardial infarction. This suggests that there are unexplored routes to cardiovascular health through the gut microbiota. Hypothesis: We hypothesize that a community of symbiotic gut microbiota is required for heart mechanical modifications during adaptive cardiac remodeling under stress. Methods and Results: Pressure-overload stress was induced by a transverse aortic constriction (TAC) surgery and dysbiosis was induced by antibiotic treatment (ABX) in mice. Under echocardiography, ABX-TAC mice showed worse cardiac outcomes versus controls (ejection fraction = 59% ± 2% vs 67% ± 2%, P < 0.001, N ≥ 12/group). Under microscopic examination of the extracellular matrix and tensile tests, ABX-TAC mice had larger fibrotic areas (9.4% ± 0.4% vs 5.7% ± 0.4%, P < 0.0001) and collagen disarray, accompanied by more severe ventricular stiffening (Young’s moduli = 360 ± 10 kPa vs 280 ± 43 kPa, P < 0.01, N ≥ 6/group). When establishing normal gut flora before surgery, germ-free mice had heart malfunctioning rescued (change of myocardial performance index = 15% ± 13% vs 120% ± 32% untreated, P < 0.05, N ≥ 5/group). The normal gut flora was profiled by third-generation 16S sequencing to acquire bacterial information with high accuracy, followed by PICRUSt analysis revealing that the microbes favored the production of short-chain fatty acids (SCFAs) (N ≥ 8/group). When supplemented with SCFA before surgery, ABX mice gained better cardiac outcomes (ejection fraction = 65% ± 2% vs 53% ± 3% untreated, P < 0.01, N ≥ 5/group). Cardiac fibroblasts treated with SCFA were less susceptible to TGF-β1-triggered fibrogenesis (COL1A1/GAPDH = 0.8 ± 0.3 vs 2.3 ± 0.3 untreated, P < 0.01, n ≥ 4/group). Conclusions: In conclusion, our study demonstrates firstly that gut microbiota-derived SCFAs manipulate heart mechanical functioning under stress potentially acting via cardiac fibroblasts. This provides new insights into the management of heart diseases through the gut microbiota.

Even though the impact from large tsunamis are limited to coastal areas, these events are still devastating. Knowledge is crucial in minimizing the losses from natural disasters, as it can aid in creating better and more proactive preparation. Focusing on natural hazards’ mitigation in Asia, a collaboration between 10 Asian and two European countries, based on a deeper understanding approach, has been conducted since 2015. Deeper understanding aims to discover the physical mechanisms and drivers behind a hazard event. Innovative models and simulation facilities are developed correspondingly to achieve more accurate numerical simulations of the whole lifespan of the target event. An application framework composed from the knowledge, data, simulation facility, software tools, and case studies is designed to provide an advanced estimation of hazard risk and would be evolved progressively with more case studies and observation data. For tsunamis, based on the COMCOT (COrnell Multi-grid Coupled Tsunami Model), the simulation portal (iCOMCOT) implementing parallelized tsunami wave propagation calculation over distributed clouds had been established. The iCOMCOT system finished the simulation of the whole lifecycle of the 2011 Great East Japan Earthquake Tsunami in 1 min. In this regional collaboration, case studies on historical events and tsunami impact analysis were conducted. The goal is to capture the physical characteristics of the tsunami as much as possible, such as tsunami wave propagation, tsunami refraction, and tsunami run-up on land, as well as their drivers and root causes. The whole processes of the tsunami, from its initiation to its impacts in selected locations, then could be simulated accurately by iCOMCOT based on the scientific explorations and the quantitatively revised models. The Sulawesi Tsunami (2018) case is presented to demonstrate the processes of the deeper understanding approach and how to achieve the capacity building. At the same time, ways to take advantage of citizen science are also explored. The citizen science model is valuable in supporting data collection, such as data of run-up height, inundation range, flow depth, disruption information, impact area, from publication, news reports, and interviews from local people. According to experiences on case studies, suggestions to simplify and optimize the integration of the citizen science model with the deeper understanding approach to result in a lower operation cost are provided.

This work aims to study the variation, robustness, and feature redundancy of PET/MR radiomic features in the primary tumor of nasopharyngeal carcinoma (NPC). PET/MR scans of 21 NPC patients were used in this study. The primary tumor volumes were defined using PET, T2-weighted-MR (T2-MR), and diffusion-weighted MR (DW-MR) images. A random-dilation-erosion method was used to simulate 10 sets of tumor volumes for identifying features invariant with manual segmentation uncertainties. Feature robustness was evaluated against imaging modalities, pixel sizes, slice thickness, and grey-level bin sizes using intraclass correlation coefficient (ICC) and spearman correlation coefficient. Feature redundancy was analyzed using the hierarchical cluster analysis. Voxel size of 0.5 × 0.5 × 1.0 mm3 was found optimal for robust feature extraction from PET and MR. Normalized grey level of 64 and 128 was suggested for PET and MR, respectively. The features from wavelet-transformed images were less stable than those from the original images. The robustness analysis and volume correlation analysis identified 335 (62.04 %) PET features, 240 (44.44 %) T2-MR features, and 366 (67.78 %) DW-MR features. The cluster analysis grouped PET, T2-MR, and DW-MR features into 106, 83, and 133 representative features, respectively. The present study analyzed and identified robust features extracted from tumor volumes on PET/MR, which can provide guidance and promote standardization for PET/MR radiomic studies in NPC.

Driven by the need to carefully plan and optimise the resources for the next data taking periods of Big Science projects, such as CERN's Large Hadron Collider and others, sites started a common activity, the HEPiX Technology Watch Working Group, tasked with tracking the evolution of technologies and markets of concern to the data centres. The talk will give an overview of general and semiconductor markets, server markets, CPUs and accelerators, memories, storage and networks; it will highlight important areas of uncertainties and risks.

This study explores the discrepancies of storm surge predictions driven by the parametric wind model and the numerical weather prediction model. Serving as a leading-order storm wind predictive tool, the parametric Holland wind model provides the frictional-free, steady-state, and geostrophic-balancing solutions. On the other hand, WRF-ARW (Weather Research and Forecasting-Advanced Research WRF) provides the results solving the 3D time-integrated, compressible, and non-hydrostatic Euler equations, but time-consuming. To shed light on their discrepancies for storm surge predictions, the storm surges of 2013 Typhoon Haiyan in the Leyte Gulf and the San Pedro Bay are selected. The Holland wind model predicts strong southeastern winds in the San Pedro Bay after Haiyan makes landfall at the Leyte Island than WRF-ARW 3 km and WRF-ARW 1 km. The storm surge simulation driven by the Holland wind model finds that the water piles up in the San Pedro Bay and its maximum computed storm surges are almost twice than those driven by WRF-ARW. This study also finds that the storm surge prediction in the San Pedro Bay is sensitive to winds, which can be affected by the landfall location, the storm intensity, and the storm forward speed. The numerical experiment points out that the maximum storm surges can be amplified by more 5–6% inside the San Pedro Bay if Haiyan’s forward speed is increased by 10%.

This study investigates the effects of horizontal resolution and surface flux formulas on typhoon intensity and structure simulations through the case study of the Super Typhoon Haiyan (2013). Three sets of surface flux formulas in the Weather Research and Forecasting Model were tested using grid spacings of 1, 3, and 6 km. Increased resolution and more reasonable surface flux formulas can both improve typhoon intensity simulation, but their effects on storm structures differ. A combination of a decrease in momentum transfer coefficient and an increase in enthalpy transfer coefficients has greater potential to yield a stronger storm. This positive effect of more reasonable surface flux formulas can be efficiently enhanced when the grid spacing is appropriately reduced to yield an intense and contracted eyewall structure. As the resolution increases, the eyewall becomes more upright and contracts inward. The size of updraft cores in the eyewall shrinks, and the region of downdraft increases; both updraft and downdraft become more intense. As a result, the enhanced convective cores within the eyewall are driven by more intense updrafts within a rather small fraction of the spatial area. This contraction of the eyewall is associated with an upper-level warming process, which may be partly attributed to air detrained from the intense convective cores. This resolution dependence of spatial scale of updrafts is related to the model effective resolution as determined by grid spacing.

It is now important to prepare spectrum auction for the 5th Generation Mobile Networks that will start operation in 2020. This paper proposes a novel approach for the optimization for the 5G spectrum. Our ultimate target is to determine the various variables of 5G to optimize the revenue of the spectrum auction by the optimization algorithms. For the optimization, we develop advanced Simulated Annealing Algorithm and Genetic Algorithm. We use the costs and benefits of telecommunication companies as a constraint to achieve the goal of revenue maximization. Finally, this paper shows the optimal results by chart. This study is the first of its kind thus far.

Hazard risk assessment is essential to coexist with the natural disasters. An open application framework for knowledge-oriented hazard risk assessment based on deeper understanding of the root cause and drivers of a disaster is developed and verified in this study. Hazard risks estimation by numerical simulation could be effective and reliable only if we have enough knowledge and observation data to the hazards. Insufficient knowledge to the natural hazards is the most critical barrier for efficient disaster risk assessment. To transform the knowledge into valuable simulations on hazard risks is still a challenge. To overcome the challenges, we proposed and implemented the application framework for hazard risk evaluation based on deeper understanding approach. For each case study, scientists achieved the accurate simulation processes with best knowledge to the root cause and physical mechanism of the target hazard. Simulation portal was then developed for easier access to the most updated analysis facility. All the data, services, methods and knowledge were compiled into a knowledge base. Those shared data, analysis and simulation facility, knowledge base and value-added services, and the distributed could infrastructure constitute the open application framework. This is a practical framework verified by case studies and will be evolved progressively with more cases of various types of disasters in different places. In this paper, design details of the application framework based on deeper understanding approach and the verification by a case study are explained. The application framework would be a facilitator for capacity building on hazard risk assessment and will be a solid ground for open science development of this domain in the future.

A knowledge-oriented hazard risk assessment approach based on deeper understanding of the root cause and drivers of a hazard is developed and verified in this study. The open collaboration framework consists of case study, simulation facility and knowledge base to carry out the hazard assessment by this approach has been initiated. Several case studies of different types in different countries were implemented. Simulation facility is built from the requirements of the target case study. Design of the knowledge base for disaster assessment by compiling all the materials and resources from case studies in an organized way is also proposed. A positive feedback loop is formed by the case study, simulation facility and knowledge base. Both the knowledge to the hazard physical processes and the simulation facility will progressively reinforced by the growing of case studies. Through integration and share of data, simulation facility and innovative applications, workflow, and details of computational environment that generate published findings in open trusted repositories from the open collaboration platform, an open science platform for disaster mitigation would be realized.

It is now important to prepare spectrum auction for the 5 th Generation Mobile Networks that will start operation in 2020. This paper proposes a novel approach for the optimization for the 5G spectrum. Our ultimate target is to determine the various variables of 5G to optimize the revenue of the spectrum auction by the optimization algorithms. For the optimization, we develop advanced Simulated Annealing Algorithm and Genetic Algorithm. We use the costs and benefits of telecommunication companies as a constraint to achieve the goal of revenue maximization. Finally, this paper shows the optimal results by chart. This study is the first of its kind thus far.

Distributed Cloud Operating System (DiCOS) is designed to support large and long-tail data analysis efficiently in all disciplines by federating distributed Cloud resources. The distributed infrastructure established by DiCOS federates small data centers and provides low latency access and high performance applications to local users. Job is submitted locally and run globally while reducing the number of data transmitted across resource centers. DiCOS could provide more than 400K CPU-days computing power to support several application domains with 6 resource sites in Taiwan, US and CERN. DiCOS is becoming the new generation research infrastructure of Academia Sinica. In addition to high throughput, application efficiency is improved continuously by the optimization of workflow and system efficiency. The objectives and status of DiCOS, as well as the development strategy and plan for the second stage from 2016 are described in this study.

In this paper, we report the development of fanless single rack data centre (SRDC) which is noise-free and of efficiency in power, energy and operation. SRDC could be deployed as a building block according to different needs and close to users to reduce the latency. Distributed cloud infrastructure constituted by these SRDCs could provide federated computing capability as a large cloud centre. Private cloud could be also implemented according to the application needs. The first prototype of fanless SRDC with a dedicated cooling system is accomplished and verified. It consists of fanless computing nodes with dual CPUs, fanless 12 bays disk servers and a fanless 10G switch with fibre interfaces. Test results show that heat generated by components in each server can bring out effectively by the conductional cooling. The temperature of CPUs can be kept within 60oC under the full load condition. The cooling system modified from the existing air conditioner can achieved PUE < 1.25 in the summer condition.

It is hoped that through the cultivation of a crew of volunteer citizen seismologists, public involvement could be encouraged and the discovery and inquiry into earthquake knowledge could be promoted. These volunteers can contribute to data collection, analysis, and reporting, and have the potential to greatly improve the emergency response to earthquakes. The Citizen Seismologists in Taiwan Project (CSTaiwan) is designed to elevate the quality of earthquake science education by incorporating earthquake and tsunami stories and educational earthquake games into traditional school curricula. The project aims to build a cloud-based computing service incorporating an earthquake school (i.e., a website for online learning) where teachers can easily teach their students about earthquakes and children can learn about earthquakes in a fun environment. Here we demonstrate how students perform P-and S-wave picking and measure seismic intensity through an interactive learning platform, how scientists and school teachers work together, and how we create a near-real-time earthquake games competition to facilitate continuous learning while making earthquake science fun. We also develop 49 questions associated with participants' preknowledge, attitude, and skills in earthquake sciences, called Citizen Seismological Literacy (CSL). The CSL model may serve as an example to quantify citizen's background in earthquake sciences and could be applied as a framework for seismologists around the world who wish to approach the public for educational purposes, while considering promoting the public's seismologic literacy.

The Manila subduction zone is identified as one of the most hazardous tsunami source regions, and the countries around the South China Sea are under threat from tsunami hazard. However, the number of early-warning tsunami buoys being deployed in this area is far fewer than that in the Pacific Ocean. This study investigates a feasible approach for establishing a tsunami early warning system in the South China Sea without deploying buoys. The idea is to integrate existing earthquake early warning systems with a fast computing system for estimating tsunami threats. This study presents an efficient and low-cost tsunami fast computing system for early warning. The widely validated tsunami model, COMCOT is chosen as the kernel. The COMCOT source code has been optimized and parallelized in order to meet the requirements of real-time simulation. The optimized model, iCOMCOT performs at least 10 times faster than the original COMCOT. In addition, a flexible and user-friendly grid/cloud-based portal service has been built, which is also made available for mobile devices. As for the automatic generation of the tsunami sources, a new Source-Scaling relationship, which has been validated by recent mega-earthquakes, is implemented. The 2011 Tohoku tsunami is adopted as a case for validation and demonstration. The modelling results of Manila trench, chosen to demonstrate the application of iCOMCOT, show that the western coast of the Philippines is prone to get tsunami attacks. Furthermore, tsunami from the north segment of Manila trench tend to strike southern Taiwan, Hong Kong and Macau area; tsunami from the middle segment, Vietnam.