Publication:
Nuevas estrategias para pronosticar la trayectoria e intensidad de los huracanes en el océano atlántico
Nuevas estrategias para pronosticar la trayectoria e intensidad de los huracanes en el océano atlántico
Authors
Castro-Sánchez, Joan M.
Embargoed Until
Advisor
Ramírez-Beltrán, Nazario D.
College
College of Engineering
Department
Department of Electrical and Computer Engineering
Degree Level
M.S.
Publisher
Date
2007
Abstract
Durante muchos años, los científicos, matemáticos e ingenieros han estudiado sobre la formación de los huracanes. Dos aspectos muy importantes que los investigadores han tomado en consideración dentro del área de huracanes son los relacionados con la predicción del desplazamiento y la intensidad. Los expertos dentro del campo de la meteorología han desarrollado diferentes métodos numéricos, dinámicos y estadísticos para pronosticar la trayectoria y la intensidad de los huracanes. Algunos modelos usados actualmente para predecir los parámetros relacionados con el movimiento de translación y los vientos máximos sostenidos son el Geophysical Fluid Dynamic Laboratory Model (GFDL) y el Statistical Hurricane Intensity Prediction Scheme (SHIPS). Nuestro trabajo de investigación desarrolló un método mixto (estadístico y dinámico) que pronosticó la trayectoria e intensidad de huracanes a 12 y 24 horas. Para ello, usamos los modelos numéricos que simulan las condiciones atmosféricas, oceánicas y de superficie. Seleccionamos los índices meteorológicos que aportaron al pronóstico de trayectoria e intensidad de huracanes. Mediante la aplicación de métodos estadísticos para la selección de variables y técnicas de optimización no lineales, desarrollamos funciones de transferencia que estimaron los cambios en el desplazamiento zonal y meridional de la tormenta, así como los cambios en la intensidad. Nuestro objetivo fue desarrollar un modelo de predicción rápida y eficiente que logre reducir los errores de predicción para trayectoria e intensidad en comparación con los modelos utilizados y validados por el “National Hurricane Center” (NHC).
Scientists, mathematicians and engineers, have studied the evolution of the hurricane for a long time. Two of the most interesting aspects that the hurricane researchers take in consideration are hurricane displacement and hurricane intensification. People who work in the meteorology field have developed different tools and methods to analyze, predict and forecast hurricane intensity and hurricane displacement. Hurricane experts are using dynamic models, for example Geophysical Fluid Dynamics Laboratory Model (GFDL) to predict hurricane tracks and statistical models, such as, Statistical Hurricane Intensity Prediction Scheme (SHIPS) to forecast hurricane intensity. Our proposed statistical and dynamic prediction method was developed to predict hurricane tracks and intensity during twelve and twenty-four hours in advance. We used numerical models that simulated synoptic variables for the atmosphere, surface and ocean. Source of information mainly come from radiosondes, satellites, aircrafts, ships, and radars. In this work, we showed some strategies to select the best predictors to forecast hurricane tracks and hurricane intensity. To do this, we used statistics tools; for example, a variable selection procedure is used to identify the best predictors to forecast displacement and change wind intensity. We also applied no-lineal optimization techniques to identify the structure and parameters of the transfer function model. Our main objective was to develop a simple, quick and efficient hurricane tracking and intensity forecast scheme that reduce predictions errors generated by National Hurricane Center models.
Scientists, mathematicians and engineers, have studied the evolution of the hurricane for a long time. Two of the most interesting aspects that the hurricane researchers take in consideration are hurricane displacement and hurricane intensification. People who work in the meteorology field have developed different tools and methods to analyze, predict and forecast hurricane intensity and hurricane displacement. Hurricane experts are using dynamic models, for example Geophysical Fluid Dynamics Laboratory Model (GFDL) to predict hurricane tracks and statistical models, such as, Statistical Hurricane Intensity Prediction Scheme (SHIPS) to forecast hurricane intensity. Our proposed statistical and dynamic prediction method was developed to predict hurricane tracks and intensity during twelve and twenty-four hours in advance. We used numerical models that simulated synoptic variables for the atmosphere, surface and ocean. Source of information mainly come from radiosondes, satellites, aircrafts, ships, and radars. In this work, we showed some strategies to select the best predictors to forecast hurricane tracks and hurricane intensity. To do this, we used statistics tools; for example, a variable selection procedure is used to identify the best predictors to forecast displacement and change wind intensity. We also applied no-lineal optimization techniques to identify the structure and parameters of the transfer function model. Our main objective was to develop a simple, quick and efficient hurricane tracking and intensity forecast scheme that reduce predictions errors generated by National Hurricane Center models.
Keywords
Usage Rights
Persistent URL
Cite
Castro-Sánchez, J. M. (2007). Nuevas estrategias para pronosticar la trayectoria e intensidad de los huracanes en el océano atlántico [Thesis]. Retrieved from https://hdl.handle.net/20.500.11801/2362