Update about.md

_pages/about.md

@@ -10,11 +10,14 @@ redirect_from:
 
 __Biography__
 
-I am Shuolin Xiao. I received my Ph.D. in Mechanical Engineering from the University of Houston in 2020 and a B.A. in Electrical Engineering from Beijing Institute of Technology, China, in 2011. During my graduate studies, I developed numerical simulations for oil spills and wind turbine wake flows, with a primary objective of enhancing offshore environmental sustainability. After my Ph.D., I joined the School of Civil and Environmental Engineering at Cornell University as a Postdoctoral Associate. There, my research focused on the transport and fate of microplastics across urban, atmospheric, and oceanic environments. I am currently a Postdoctoral Fellow at the Ralph S. O’Connor Sustainable Energy Institute at Johns Hopkins University, where I specialize in numerical simulations of wind farms using large-eddy simulation and machine learning technologies for analysis.
+I am Shuolin Xiao. I received my Ph.D. in Mechanical Engineering from the University of Houston in 2020 and a B.A. in Electrical Engineering from Beijing Institute of Technology, China, in 2011. During my graduate studies, I developed numerical simulations for oil spills and wind turbine wake flows, with a primary objective of enhancing offshore environmental sustainability. After my Ph.D., I joined the School of Civil and Environmental Engineering at Cornell University as a Postdoctoral Associate. There, my research focused on the transport and fate of microplastics across urban, atmospheric, and oceanic environments. I am currently a Postdoctoral Fellow at the Ralph S. O’Connor Sustainable Energy Institute at Johns Hopkins University, where I specialize in coarse-grained large-eddy simulations of wind farms and the buidlup of the web-service accessible public database. Also, very recently, I started with another project on very coarse large-eddy simulation with a closure model discovered by a reinforcement learning approach, aiming to capture the long-term behavior of a chaotic system.
 
 __Research Interests__
 
-My research focuses on micrometeorological applications that leverage AI methodologies and digital twins to explore how micrometeorology interacts with man-made structures, such as agricultural facilities and renewable energy systems, facilitates the long-range transport of emerging contaminants, including microplastics and per- and polyfluoroalkyl substances (PFAS), and subsequently impacts local ecology and biodiversity under large-scale forcing, particularly during extreme weather events. Due to limited localized data availability for AI/ML training, a comprehensive repository must be developed, integrating data collected from high-fidelity simulations, sensors, and localized imaging. This repository is critical for developing low-cost, data-driven surrogate models. High-fidelity simulations combined with localized data are essential for understanding the underlying physics and informing data-driven models that incorporate sufficient physical and local details. Once these models are built, digital twins—virtual replicas of physical systems—can provide dynamic, interactive systems for monitoring, prediction, and decision-making by continuously updating the models with real-time data. Collectively, these advancements enhance our ability to understand and predict dynamic processes in micrometeorology, while also reducing operational and maintenance costs for man-made structures and agricultural facilities.
+My research focuses on Long-term behaviors in chaotic systems, such as bifurcations or transitions between meta-stable states, are particularly relevant in single or hybrid energy systems, rotating denotation engines, and carbon capture and storage when accounting for temporal effects that evolve over extended periods. For wind farm, if we consider only daytime or nighttime conditions, the first row of wind turbines is generally expected to generate more power compared to the last row, primarily due to the presence of wake flow that reduces wind speeds downstream. However, when examining wind farm performance over an extended period, such as a full diurnal cycle (24 hours), a different pattern can emerge. During certain morning transitions, the first row of wind turbines may generate less power compared to the last row. This occurs because, during the morning transition, the atmospheric boundary layer height is reduced relative to the size of the wind turbines, causing the incoming flow to be significantly impacted by the blockage effect of the wind farm. This wind farm blockage effect reduces wind speeds at hub height upstream of the farm, diminishing the power output of the front-row turbines. At the same time, the formation of a low-level jet (a narrow, high-velocity wind band) and enhanced turbulent kinetic energy, driven by mixing effects triggered by the downstream turbines, increase wind speeds experienced by the turbines in those rows. As a result, for a few morning hours, the last row of wind turbines benefits from these conditions and outperforms the first row in power generation. This dynamic highlights the importance of considering extended periods and long-term behaviors when optimizing wind farm design and improving energy forecasting. 
 
+For classical-fluid turbulent flow, including Von Kármán swirling flow, which serves as a foundation for understanding \uline{rotating detonation engines, the transition between meta-stable states plays a critical role. This transition, however, can only be captured by observing the long-term behavior of the system, as short-term analysis often misses the intricate dynamics involved. Understanding and characterizing this long-term transition is crucial for many applications, such as improving the efficiency of energy systems, optimizing mixing processes in industrial flows, and enhancing the predictive accuracy of climate models that rely on turbulent dynamics. One of the key challenges for carbon capture and storage, particularly in the context of upscaling, is the ability to reliably evaluate long-term effects. Upscaling involves transitioning from pilot projects to large-scale deployment, where factors such as storage capacity, reservoir stability, and leakage risks become critical. Monitoring and predicting the behavior of stored carbon over extended periods require advanced modeling and long-term observational data. This is essential to ensure that the injected $CO_2$ remains securely trapped and does not negatively impact surrounding ecosystems or infrastructure. Addressing these challenges is crucial for validating carbon capture and storage as a sustainable and effective solution for mitigating climate change.
+
+My methodology leverages AI- or theory-aided, very-coarse high-fidelity simulations to model turbulent systems over long-term runs. This approach is complemented by the development of web-service accessible public databases to support the creation of data-driven surrogate models. These models incorporate sufficient long-term physical and other relevant details while reducing computational costs. Once constructed, these surrogate models integrate process simulations, optimization techniques, and uncertainty quantification to predict system behavior over extended periods. In particular, digital twins—virtual replicas of physical systems—will be created to enable dynamic, interactive monitoring, prediction, and decision-making by continuously updating the models with real-time data. Collectively, these advancements enhance our ability to understand, optimize, and predict dynamic processes across various applications over the long term.
 ![test](./images/Application_graphical_abstract.jpg)