Improving Time Structure Patterns of Orthogonal Markov Chains and its Consequences in Hydraulic Simulations

• Autor: Jaimes Correa, Juan Carlos
• Formato: tesis de maestría
• Fecha de Publicación: 2013
• Institución: Superintendencia Nacional de Educación Superior Universitaria
• Repositorio: Registro Nacional de Trabajos conducentes a Grados y Títulos - RENATI
• Lenguaje: inglés
• OAI Identifier: oai:renati.sunedu.gob.pe:renati/6909
• Enlace del recurso: https://renati.sunedu.gob.pe/handle/sunedu/3448668
• Nivel de acceso: acceso abierto
• Materia: Lluvia; Simulación estocástica; Generadores de clima; Procesos de Markov; Cuencas hidrográficas; https://purl.org/pe-repo/ocde/ford#2.01.01

The frequency of rainfall is relevant for agriculture because its distribution affects crop production. Few stochastic models successfully generate daily rainfall events and preserve spatiotemporal dependency between multiple sites. This work evaluated an extension of the traditional Orthogonal Markov Chain (TOMC) model to reproduce the temporal structure of rainfall events in Florida, Nebraska, and California. A simulation of a hydrographic basin was also carried out from rainfall reproduced by a meteorological generator. The results show that (i) an extended temporal structure of the TOMC (EOMC) preserved the spatial correlation between observed and synthetic rainfall events; (ii) the EOMC used a smaller number of simulations to reproduce the observed frequencies of wet periods than those used by the TOMC to achieve similar precision; (iii) the use of EOMC-generated rainfall in the SWMM model produced similar runoff values using observed rainfall data; and (iv) the EOMC decreased 50% of the computational time needed to generate synthetic rainfall. The use of the EOMC would benefit the simulation of future climate scenarios due to the reduction in hardware needs.