Computer modeling makes possible the simulation of shoreline hazards, from tsunamis and tropical storms. To make predictions of these hazard events requires many simulations, to explore the high dimensional space of input parameters, and massive computational budgets. Data science methods are needed to detect nascent storms and tsunami waves and feed information to simulation models, to monitor the evolving hazard or make long-term predictions. Statistical emulators can estimate the output of simulations and greatly reduce the computational burden. However the necessary outputs are often spatio-temporal fields, and conventional methods for constructing emulators cannot be applied. This mini-symposium, which emerged from research activity during the 2018-19 SAMSI program on Uncertainty Quantification, will bring together scientists working on computational and statistical methodology to better predict and track storms and tsunamis.
08:30
A Combined Physical-Statistical Approach for Estimating Storm Surge Risk
Whitney Huang | Clemson University | United States
Show details
Author:
Whitney Huang | Clemson University | United States
Storm surge is an abnormal rise of seawater caused by a storm. According to the National Hurricane Center, storm surge is often the most damaging part of a hurricane and poses the most severe threat to property and life in a coastal region. Thus, it is crucially important to assess the storm surge risk, typically summarized by r-year surge return level with return period r ranging from 10, 50, 100, or even much longer along a coastline. It is however very difficult to reliably estimate this quantity due to the limited storm surge observations in space and time. This talk presents an approach to integrate physical and statistical models to estimate extreme storm surge. Specifically, A physically-based hydrodynamics model is used to provide the needed interpolation in space and extrapolation in both time and atmospheric conditions. Statistical modeling is needed to 1) estimate the input distribution for running the computer model, 2) develop a statistical emulator in place of the computer simulator, and 3) quantify estimate uncertainty due to input distribution, statistical emulator, missing/unresolved physics.
This is joint work with SAMSI storm surge working group under the year-long program: Model Uncertainty: Mathematical and Statistical (MUMS)
09:00
Uncertainty Quantification in Assessing Storm Surge Hazards
Pulong Ma | Statistical and Applied Mathematical Sciences Institute | United States
Show details
Author:
Pulong Ma | Statistical and Applied Mathematical Sciences Institute | United States
Storm surge is one of the most severe natural hazards that can lead to significant flooding in coastal areas and severe damages to the life and property from a hurricane. Risk assessment of storm surges is addressed through a synthesis of computer modeling, statistical modeling, and extreme-event probability computation. A computer model of the natural hazard developed based on mathematical representations of real-world processes provides the capability of predicting storm surges over unseen parts of the hazard space. Statistical modeling of the available data is needed to determine the initializing distribution for exercising the computer model. In dealing with rare events, large-scale simulations involving the computer model are required. Physics-based computer models of storm surge can be implemented with a wide range of fidelity due to the nature of the system, though the danger posed by surge makes greater fidelity highly desirable. However, such models and their high-dimensional outputs tend to come at great computational cost, which can make highly detailed studies prohibitive. A crucial need is the development of an emulator - a fast approximation to the computer model. We develop an emulator that is capable of handling very large spatial domains and enables the computation of hazard probabilities. This methodology is used to facilitate risk assessment of storm surges over coastal areas of the United States.
09:30
How high could be the coastal waves generated by a storm surge? A computationally efficient surrogate-based optimization approach
Serge Guillas | University College London | United Kingdom
Show details
Author:
Serge Guillas | University College London | United Kingdom
Storm surges cause coastal inundations due to the setup of the water surface resulting from atmospheric pressure, surface winds and breaking waves. The latter is particularly difficult to be accounted for. For instance, it was observed that during Typhoon Haiyan (2013, Philippines), a stretch of coral reef near the coast, which was expected to protect the coastal communities, actually amplified the waves. The generation of such large nearshore waves and their breaking process can be successfully captured by a Boussinesq-type phase-resolving wave model. Building defences requires estimating the maximum storm surge height, but also maxima of breaking wave height and run-up on land, with respect to a variety of storm characteristics. However, the computational complexity of the wave model makes optimization tasks impractical. To find the maximum breaking wave (bore) height and the maximum run-up, we employ optim-MICE, a new surrogate-based optimization scheme based on Gaussian Process emulation and information theory, which, we show, outperforms other methods on a large range of optimization problems. In two idealized settings we efficiently identify the conditions that create the largest storm waves at the coast since optim-MICE uses a minimal number of simulations. This is the first maximization of storm wave heights and run-ups using surrogate-based optimization. It opens the door to previously inconceivable coastal risk assessments.
10:00
Uncertainty Quantification of Tsunami Currents and Heights
Devaraj Gopinathan | University College London | United Kingdom
Show details
Author:
Devaraj Gopinathan | University College London | United Kingdom
Hazard due to tsunami currents in ports is under-studied and under-rated. The catastrophic tsunamis of 2004 and 2011 necessitate a hazard assessment due to tsunami waves and accompanying currents in ports. This study is the first to employ statistical emulation to investigate the port hazard from the Makran subduction zone at the port of Karachi. First, a simple three-parameter characterization of the earthquake source is carried out. The computer code simulates two physical models – the deformation of sea floor due to the source and the consequent propagation of the tsunami to the port. Multiple emulators (~10k) are constructed to approximate the functional dependence of the wave heights and velocities on the earthquake source. Then, we propagate one million scenarios representing the source uncertainties through the emulators. The resulting distributions of wave heights and velocities reveal remarkable peak structures that may point to resonance in ports. We also observe high velocities in the range of 10 – 12 m/s for certain ranges of parameters. As such, statistical emulation in conjunction with high resolution (~10m) simulations and realistic bathymetry data provides an effective tool by which port hazard and resonant frequencies may be characterised.