Browsing by Author "Wu, Hao"
Now showing 1 - 6 of 6
Results Per Page
Sort Options
- Computational methodology to analyze the effect of mass transfer rate on attenuation of leaked carbon dioxide in shallow aquifersFucik, Radek; Solovsky, Jakub; Plampin, Michelle R.; Wu, Hao; Mikyska, Jiri; Illangasekare, Tissa H. (2021-01)Exsolution and re-dissolution of CO2 gas within heterogeneous porous media are investigated using experimental data and mathematical modeling. In a set of bench-scale experiments, water saturated with CO2 under a given pressure is injected into a 2-D water-saturated porous media system, causing CO2 gas to exsolve and migrate upwards. A layer of fine sand mimicking a heterogeneity within a shallow aquifer is present in the tank to study accumulation and trapping of exsolved CO2. Then, clean water is injected into the system and the accumulated CO2 dissolves back into the flowing water. Simulated exsolution and dissolution mass transfer processes are studied using both near-equilibrium and kinetic approaches and compared to experimental data under conditions that do and do not include lateral background water flow. The mathematical model is based on the mixed hybrid finite element method that allows for accurate simulation of both advection- and diffusion-dominated processes.
- Forecasting the Flu: Designing Social Network Sensors for EpidemicsShao, Huijuan; Hossain, K.S.M. Tozammel; Wu, Hao; Khan, Maleq; Vullikanti, Anil Kumar S.; Prakash, B. Aditya; Marathe, Madhav V.; Ramakrishnan, Naren (Virginia Tech, 2016-03-08)Early detection and modeling of a contagious epidemic can provide important guidance about quelling the contagion, controlling its spread, or the effective design of countermeasures. A topic of recent interest has been to design social network sensors, i.e., identifying a small set of people who can be monitored to provide insight into the emergence of an epidemic in a larger population. We formally pose the problem of designing social network sensors for flu epidemics and identify two different objectives that could be targeted in such sensor design problems. Using the graph theoretic notion of dominators we develop an efficient and effective heuristic for forecasting epidemics at lead time. Using six city-scale datasets generated by extensive microscopic epidemiological simulations involving millions of individuals, we illustrate the practical applicability of our methods and show significant benefits (up to twenty-two days more lead time) compared to other competitors. Most importantly, we demonstrate the use of surrogates or proxies for policy makers for designing social network sensors that require from nonintrusive knowledge of people to more information on the relationship among people. The results show that the more intrusive information we obtain, the longer lead time to predict the flu outbreak up to nine days.
- High density oilfield wastewater disposal causes deeper, stronger, and more persistent earthquakesPollyea, Ryan M.; Chapman, Martin C.; Jayne, Richard S.; Wu, Hao (Nature, 2019)Oilfield wastewater disposal causes fluid pressure transients that induce earthquakes. Here we show that, in addition to pressure transients related to pumping, there are pressure transients caused by density differences between the wastewater and host rock fluids. In northern Oklahoma, this effect caused earthquakes to migrate downward at ~0.5 km per year during a period of high-rate injections. Following substantial injection rate reductions, the downward earthquake migration rate slowed to ~0.1 km per year. Our model of this scenario shows that the density-driven pressure front migrates downward at comparable rates. This effect may locally increase fluid pressure below injection wells for 10+ years after substantial injection rate reductions. We also show that in north-central Oklahoma the relative proportion of high-magnitude earthquakes increases at 8+ km depth. Thus, our study implies that, following injection rate reductions, the frequency of high-magnitude earthquakes may decay more slowly than the overall earthquake rate.
- Moments in a pavilionWu, Hao (Virginia Tech, 2012-04-30)This thesis is my answer to the following question: What is a good building? I believe that a good building must have the moments that can touch people. It should have some spiritual qualities. These moments can be achieved by form, meterial, light, details, and color.
- Numerical Investigations of Geologic CO2 Sequestration Using Physics-Based and Machine Learning Modeling StrategiesWu, Hao (Virginia Tech, 2020-08-06)Carbon capture and sequestration (CCS) is an engineering-based approach for mitigating excess anthropogenic CO2 emissions. Deep brine aquifers and basalt reservoirs have shown outstanding performance in CO2 storage based on their global widespread distribution and large storage capacity. Capillary trapping and mineral trapping are the two dominant mechanisms controlling the distribution, migration, and transportation of CO2 in deep brine aquifers and basalt reservoirs. Understanding the behavior of CO2 in a storage reservoir under realistic conditions is important for risk management and storage efficiency improvement. As a result, numerical simulations have been implemented to understand the relationship between fluid properties and multi-phase fluid dynamics. However, the physics-based simulations that focus on the uncertainties of fluid flow dynamics are complicated and computationally expensive. Machine learning method provides immense potential for improving computational efficiency for subsurface simulations, particularly in the context of parametric sensitivity. This work focuses on parametric uncertainty associated with multi-phase fluid dynamics that govern geologic CO2 storage. The effects of this uncertainty are interrogated through ensemble simulation methods that implement both physics-based and machine learning modeling strategies. This dissertation is a culmination of three projects: (1) a parametric analysis of capillary pressure variability effects on CO2 migration, (2) a reactive transport simulation in a basalt fracture system investigating the effects of carbon mineralization on CO2 migration, and (3) a parametric analysis based on machine learning methods of simultaneous effects of capillary pressure and relative permeability on CO2 migration.
- Probabilistic Modeling of Multi-relational and Multivariate Discrete DataWu, Hao (Virginia Tech, 2017-02-07)Modeling and discovering knowledge from multi-relational and multivariate discrete data is a crucial task that arises in many research and application domains, e.g. text mining, intelligence analysis, epidemiology, social science, etc. In this dissertation, we study and address three problems involving the modeling of multi-relational discrete data and multivariate multi-response count data, viz. (1) discovering surprising patterns from multi-relational data, (2) constructing a generative model for multivariate categorical data, and (3) simultaneously modeling multivariate multi-response count data and estimating covariance structures between multiple responses. To discover surprising multi-relational patterns, we first study the ``where do I start?'' problem originating from intelligence analysis. By studying nine methods with origins in association analysis, graph metrics, and probabilistic modeling, we identify several classes of algorithmic strategies that can supply starting points to analysts, and thus help to discover interesting multi-relational patterns from datasets. To actually mine for interesting multi-relational patterns, we represent the multi-relational patterns as dense and well-connected chains of biclusters over multiple relations, and model the discrete data by the maximum entropy principle, such that in a statistically well-founded way we can gauge the surprisingness of a discovered bicluster chain with respect to what we already know. We design an algorithm for approximating the most informative multi-relational patterns, and provide strategies to incrementally organize discovered patterns into the background model. We illustrate how our method is adept at discovering the hidden plot in multiple synthetic and real-world intelligence analysis datasets. Our approach naturally generalizes traditional attribute-based maximum entropy models for single relations, and further supports iterative, human-in-the-loop, knowledge discovery. To build a generative model for multivariate categorical data, we apply the maximum entropy principle to propose a categorical maximum entropy model such that in a statistically well-founded way we can optimally use given prior information about the data, and are unbiased otherwise. Generally, inferring the maximum entropy model could be infeasible in practice. Here, we leverage the structure of the categorical data space to design an efficient model inference algorithm to estimate the categorical maximum entropy model, and we demonstrate how the proposed model is adept at estimating underlying data distributions. We evaluate this approach against both simulated data and US census datasets, and demonstrate its feasibility using an epidemic simulation application. Modeling data with multivariate count responses is a challenging problem due to the discrete nature of the responses. Existing methods for univariate count responses cannot be easily extended to the multivariate case since the dependency among multiple responses needs to be properly accounted for. To model multivariate data with multiple count responses, we propose a novel multivariate Poisson log-normal model (MVPLN). By simultaneously estimating the regression coefficients and inverse covariance matrix over the latent variables with an efficient Monte Carlo EM algorithm, the proposed model takes advantages of association among multiple count responses to improve the model prediction accuracy. Simulation studies and applications to real world data are conducted to systematically evaluate the performance of the proposed method in comparison with conventional methods.