Skip navigation
Please use this identifier to cite or link to this item: http://arks.princeton.edu/ark:/88435/dsp01s4655k66h
Title: Statistical Modeling of Tropical Cyclone Climatology
Authors: Jing, Renzhi
Advisors: Lin, Ning
Contributors: Civil and Environmental Engineering Department
Keywords: Risk Analysis
Statistical Modeling
Tropical Cyclones
Subjects: Atmospheric sciences
Environmental science
Civil engineering
Issue Date: 2020
Publisher: Princeton, NJ : Princeton University
Abstract: Tropical cyclones (TCs) are among the most destructive natural hazards on the earth. The changing character of TC-related risks under climate change is of significant concern to both coastal and inland areas. In a risk assessment framework, a large number of synthetic TCs are often needed to evaluate the risk posed to a specific region. However, current TC statistical modeling methods that are used to generate storms often rely very little on the environment but heavily on storm characteristics. This dissertation aims to improve existing tropical cyclone statistical models by developing a new TC probabilistic model that is dependent on climate predictors so that it is suitable for climate change studies. Starting from TC intensity modeling, the dependence of TC intensification on the environment is firstly explored through various statistical modeling approaches. Mixture modeling is performed to capture the heterogeneity in TC intensification and potential physically-based environmental predictors are carefully examined. Based on these analyses, a hidden Markov model, the MeHiM (short for Markov environment-dependent hurricane intensity model), is developed to simulate tropical cyclone intensity evolution dependent on six essential environmental and storm predictors. The MeHiM is then coupled with a clustering-based genesis model and a data-driven track model to form a complete TC probabilistic model, PepC (Princeton environment-dependent probabilistic tropical cyclone model). All components of PepC are dependent on local environmental predictors. PepC is capable to generate large samples of synthetic TCs in an efficient way. The synthetic storms match well with observational records under the current climate condition. The PepC is applied to investigate climate change effects on TCs by generating large numbers of TCs under current and future climates, with environmental conditions taken from the Geophysical Fluid Dynamics Laboratory (GFDL) High-Resolution Forecast-Oriented Low Ocean Resolution (HiFLOR) model. Storms are shown to become more intense under future climate, however no significant change in TC frequency is detected by the end of 21st century. The intensity component of PepC, the MeHiM, is applied and examined for real-time forecasting. The performance is compared with the state-of-the-art TC statistical forecasting skill. The MeHiM shows great potential when coupled with an accurate rapid intensification indicator. Inspired by this finding, TC real-time satellite imagery data is used to predict the onset of rapid intensification using deep learning. Following these two example applications, PepC is potentially a powerful tool for TC risk assessment to inform strategic risk management and policy making.
URI: http://arks.princeton.edu/ark:/88435/dsp01s4655k66h
Alternate format: The Mudd Manuscript Library retains one bound copy of each dissertation. Search for these copies in the library's main catalog: catalog.princeton.edu
Type of Material: Academic dissertations (Ph.D.)
Language: en
Appears in Collections:Civil and Environmental Engineering

Files in This Item:
File Description SizeFormat 
Jing_princeton_0181D_13485.pdf30.66 MBAdobe PDFView/Download


Items in Dataspace are protected by copyright, with all rights reserved, unless otherwise indicated.