<p>Landfalling tropical cyclones bring tremendous coastal and inland hazard, which depends strongly on the evolution of the wind field after the landfall. This work investigates the inland evolution of the tropical cyclone wind field via idealized numerical simulation experiments and existing theories explaining the physics of storms over the ocean. The complicated landfall process is idealized as a transient response of a mature axisymmetric tropical cyclone to instantaneous surface forcings associated with landfall.</p><p><br></p><p>First, idealized landfall experiments are performed in the f-plane Bryan Cloud Model (CM1), where surface drag coefficient and evaporative fraction are individually or simultaneously modified systematically beneath an axisymmetric mature storm. Surface drying stabilizes the eyewall and consequently weakens the overturning circulation, thereby reducing inward angular momentum transport that slowly decays the low-level wind field only within the inner-core. In contrast, surface roughening first weakens the entire low-level wind field rapidly and enhances the overturning circulation dynamically despite the concurrent thermodynamic stabilization of the eyewall; thereafter, the storm gradually decays in a manner similar to drying. As a result, total precipitation temporarily increases with roughening but uniformly decreases with drying. Storm inner size and outer size decrease monotonically and rapidly with surface roughening, while the radius of maximum wind can increase with moderate surface drying.</p><p><br></p><p>Second, the extent to which existing intensity theory formed for tropical cyclones over the ocean can explain the intensity response to idealized landfall is explored in this work. Existing theoretical predictions for the equilibrium response and transient response of storm intensity are compared against the simulated response found in previous idealized simulations. The equilibrium and transient response of storm intensity to combined surface forcings can be reproduced by the product of their individual responses, in line with traditional potential intensity theory. The intensification theory of Emanuel (2012) is generalized for predicting the weakening process and found capable of reproducing the transient intensity decay. Specifically, the rapid initial decay of near-surface wind can be captured by how kinetic energy is instantaneously reduced by surface friction, where the decay is a function of surface roughness.</p><p><br></p><p>Third, existing structural theory and TC radial length scale formed or identified for storms over the ocean are tested against the idealized landfall experiment where surface is individually dried or roughened. The equilibrium storm radial length scale can predict the transient response of storm size to surface roughening throughout the decay evolution. For surface drying experiments, TC size scales with the intensity after around 12h. The E04 wind field model can generally capture the transient response of TC low-level wind field to individual surface drying and roughening, from radius of maximum wind speed to the outer region. The E04 prediction for these two types of experiments exhibits limited dependence on the subsidence cooling rate applied in the model.</p><p><br></p><p>Overall, though results are insufficient to explain the complicated wind field evolution of every real-world landfalling storm, it provides a fundamental understanding of how storm low-level wind fields respond to inland surface properties. This work also indicates the potential for existing theory to predict how tropical cyclone intensity evolves after landfall in the real world, which is essential for improving the forecasts on any timescale and the risk assessments.</p>

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第18931号 / 工博第3973号 / 新制||工||1612(附属図書館) / 31882 / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 間瀬 肇, 教授 平石 哲也, 准教授 森 信人 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

ropical cyclone-induced tornadoes (TCIT) exacerbate the devastation that landfalling tropical cyclones have on the United States. This research applied machine learning techniques in conjunction with midlatitude severe weather parameters to create an artificial intelligence (AI) capable of predicting TCIT occurrence. Severe weather diagnostic variables were collected at thousands of gridpoints from the North American Regional Reanalysis (NARR) to characterize the environments within tropical cyclones between 1991 and 2011. A support vector machine (SVM) was generated in various configurations to obtain the most effective AI. This approach revealed many parameters that were ineffective at predicting TCITs (primarily those utilizing the effective inflow layer). In addition, the most highly configured AI were capable of predicting TCIT occurrence with a Heidke Skill Score around 0.48.

Decades of research have fostered a greater understanding of the environmental controls that drive tropical cyclone (TC) intensity change, yet the community has achieved only small improvements in intensity forecasting. Numerous environmental factors impact TC intensity, such as vertical wind shear and sea surface temperatures (SSTs), but little research has focused on establishing if SST change under the TC, or SST gradients, influence these intensity changes. This study investigated three methods to compute SST gradients. The first method calculated the SST change within fixed distances along the track. In the second and third methods, the SST was calculated over the distance traversed by the TC in two separate six-hour periods. By examining 455 24-hour weakening episodes in the eastern North Pacific, this study revealed that the first SST gradient method explained the highest 24-hour weakening variance for TCs located within SSTs at or lower than 26.5 degrees C.

京都大学 / 新制・課程博士 / 博士(地球環境学) / 甲第24766号 / 地環博第238号 / 新制||地環||46(附属図書館) / 京都大学大学院地球環境学舎地球環境学専攻 / (主査)教授 小林 広英, 准教授 落合 知帆, 准教授 深町 加津枝 / 学位規則第4条第1項該当 / Doctor of Global Environmental Studies / Kyoto University / DFAM

While the standard wind product (L2B) available operationally in near-real time from SeaWinds on QuikSCAT is only 25 km in resolution, QuikSCAT data can be enhanced to yield a 2.5 km ultra-high resolution (UHR) product. The latter can be used to help estimate Tropical Cyclone (TC) characteristics such as TC eye center and wind radii. Two studies are conducted in this thesis, in which QuikSCAT UHR wind product's effectiveness in estimating these TC characteristics is evaluated. First, a comparison is made between the analyst's choice of eye location based on UHR images and interpolated best-track position. In this analysis, the UHR images are divided into two categories, based on the analyst's confidence level of finding the eye center location. In each category, statistical error quantities are computed. UHR images within the high confidence category can provide, for a given year and basin, mean error distance as small as 19 km with a 10 km standard deviation. Second, a visual comparison of QuikSCAT's performance in estimating wind radii is made. QuikSCAT's performance is gauged against H*wind dataset and the Extended Best-Track (EBT) dataset. Results show that QuikSCAT UHR data yields a correct 34-kt wind radius most of the time regardless of the TC category when compared to both H*wind and EBT, whereas the 50- and 64-kt wind radii visual estimates do not always agree with H*wind and EBT. A more sophisticated method is also implemented to automatically estimate wind radii based on a model fit to QuikSCAT data. Results from this method are compared with EBT wind radii. Wind radii obtained from QuikSCAT model fit are generally highly correlated with EBT estimated wind radii. These two studies show that QuikSCAT UHR wind products are helpful in estimating TC eye location and wind radii, thus improving TC forecasting and analysis.

No description available.

Tsunami and storm surge are two of the main destructive and costly natural hazards faced by coastal communities around the world. To enhance coastal resilience and to develop effective risk management strategies, accurate and efficient tsunami and storm surge prediction models are needed. However, existing physics-based numerical models have the disadvantage of being difficult to satisfy both accuracy and efficiency at the same time. In this dissertation, several surrogate models are developed using statistical and machine learning techniques that can rapidly predict a tsunami and storm surge without substantial loss of accuracy, with respect to high-fidelity physics-based models. First, a tsunami run-up response function (TRRF) model is developed that can rapidly predict a tsunami run-up distribution from earthquake fault parameters. This new surrogate modeling approach reduces the number of simulations required to build a surrogate model by separately modeling the leading order contribution and the residual part of the tsunami run-up distribution. Secondly, a TRRF-based inversion (TRRF-INV) model is developed that can infer a tsunami source and its impact from tsunami run-up records. Since this new tsunami inversion model is based on the TRRF model, it can perform a large number of tsunami forward simulations in tsunami inversion modeling, which is impossible with physics-based models. And lastly, a one-dimensional convolutional neural network combined with principal component analysis and k-means clustering (C1PKNet) model is developed that can rapidly predict the peak storm surge from tropical cyclone track time series. Because the C1PKNet model uses the tropical cyclone track time series, it has the advantage of being able to predict more diverse tropical cyclone scenarios than the existing surrogate models that rely on a tropical cyclone condition at one moment (usually at or near landfall). The surrogate models developed in this dissertation have the potential to save lives, mitigate coastal hazard damage, and promote resilient coastal communities. / Doctor of Philosophy / Tsunami and storm surge can cause extensive damage to coastal communities; to reduce this damage, accurate and fast computer models are needed that can predict the water level change caused by these coastal hazards. The problem is that existing physics-based computer models are either accurate but slow or less accurate but fast. In this dissertation, three new computer models are developed using statistical and machine learning techniques that can rapidly predict a tsunami and storm surge without substantial loss of accuracy compared to the accurate physics-based computer models. Three computer models are as follows: (1) A computer model that can rapidly predict the maximum ground elevation wetted by the tsunami along the coastline from earthquake information, (2) A computer model that can reversely predict a tsunami source and its impact from the observations of the maximum ground elevation wetted by the tsunami, (3) A computer model that can rapidly predict peak storm surges across a wide range of coastal areas from the tropical cyclone's track position over time. These new computer models have the potential to improve forecasting capabilities, advance understanding of historical tsunami and storm surge events, and lead to better preparedness plans for possible future tsunamis and storm surges.

Introduction: Tropical Cyclones (TCs) inflict considerable damage to life and property every year. A major problem is that residents often hesitate to follow evacuation orders when the early warning messages are perceived as inaccurate or uninformative. The root problem is that providing accurate early forecasts can be difficult, especially in countries with less economic and technical means. Aim: The aim of the thesis is to investigate how cyclone early warning systems can be technically improved. This means, first, identifying problems associated with the current cyclone early warning systems, and second, investigating if biologically based Artificial Neural Networks (ANNs) are feasible to solve some of the identified problems. Method: First, for evaluating the efficiency of cyclone early warning systems, Bangladesh was selected as study area, where a questionnaire survey and an in-depth interview were administered. Second, a review of currently operational TC track forecasting techniques was conducted to gain a better understanding of various techniques’ prediction performance, data requirements, and computational resource requirements. Third, a technique using biologically based ANNs was developed to produce TC track and intensity forecasts. Systematic testing was used to find optimal values for simulation parameters, such as feature-detector receptive field size, the mixture of unsupervised and supervised learning, and learning rate schedule. Five types of 2D data were used for training. The networks were tested on two types of novel data, to assess their generalization performance. Results: A major problem that is identified in the thesis is that the meteorologists at the Bangladesh Meteorological Department are currently not capable of providing accurate TC forecasts. This is an important contributing factor to residents’ reluctance to evacuate. To address this issue, an ANN-based TC track and intensity forecasting technique was developed that can produce early and accurate forecasts, uses freely available satellite images, and does not require extensive computational resources to run. Bidirectional connections, combined supervised and unsupervised learning, and a deep hierarchical structure assists the parallel extraction of useful features from five types of 2D data. The trained networks were tested on two types of novel data: First, tests were performed with novel data covering the end of the lifecycle of trained cyclones; for these test data, the forecasts produced by the networks were correct in 91-100% of the cases. Second, the networks were tested with data of a novel TC; in this case, the networks performed with between 30% and 45% accuracy (for intensity forecasts). Conclusions: The ANN technique developed in this thesis could, with further extensions and up-scaling, using additional types of input images of a greater number of TCs, improve the efficiency of cyclone early warning systems in countries with less economic and technical means. The thesis work also creates opportunities for further research, where biologically based ANNs can be employed for general-purpose weather forecasting, as well as for forecasting other severe weather phenomena, such as thunderstorms.

Recent studies have found that equatorial Atlantic sea surface temperature (SST) variability may be influencing tropical Indian Ocean climate (Kucharski 2009, Wang 2009). Due to the economic and social impact of tropical cyclones, it is important to investigate how an Atlantic-Indian Ocean connection may be affecting tropical cyclone behavior in the southern Indian Ocean. In this study, the International Best Track Archive for Climate Stewardship (IBTrAC) tropical cyclone database is used to derive metrics of tropical cyclone behavior, which are then compared with indices of tropical Atlantic SST variability representing Atlantic Niño, and Benguela Niño events. Changes in tropical Atlantic SSTs are found to coincide with significant differences in tropical cyclone activity for portions of the southern Indian Ocean. In addition, for these same regions, tropical Atlantic SST variability is associated with changes in large-scale atmospheric conditions, including steering flow, low level vorticity, and humidity, typically associated with tropical cyclogenesis, and tropical cyclone track. The changes in steering flow related to both indices of Atlantic SST anomaly are reproduced by an atmospheric model. The changes in steering flow are also found to be linked to changes in TC translational velocity, and TC tracking. These findings indicate a possible link between tropical Atlantic conditions and cyclone activity in the Indian Ocean mediated through a teleconnection between tropical Atlantic SSTs and large scale atmospheric conditions over the southern Indian Ocean. The teleconnection related to the Benguela Niño region of SST variability was found to consist of a Rossby wave initiated off the coast of South America, and propagating into the Indian Ocean, thereby influencing several atmospheric variables, including steering flow. The teleconnection related to the Atlantic Niño region of SST variability was not well defined, although there was some evidence of a Walker circulation anomaly extending from the equatorial Atlantic over the continent of Africa, and influencing SIO steering flow. / Graduation date: 2012