Uso de herramientas estadísticas para la predicción estacional del campo de precipitación en América Central como apoyo a los Foros Climáticos Regionales. 1: Análisis de tablas de contingencia.
Files
Date
2012-11-21
Authors
Fallas López, Berny
Alfaro Martínez, Eric J.
Journal Title
Journal ISSN
Volume Title
Publisher
Revista de Climatología, Vol 12 (2012): 61-79
Abstract
Se utilizó la técnica estadística del análisis de tablas de contingencia para elaborar esquemas predictivos de los campos de precipitación en América Central. Como primer paso, se produjeron índices de estos campos utilizando el análisis de componentes principales a partir de los registros de 146 estaciones con datos diarios. Se obtuvieron dos componentes principales para la precipitación, asociados con las vertientes Pacífico y Caribe de América Central. Debido a que uno de los objetivos de este estudio era el de apoyar al proceso de los Foros Regionales de Predicción Climática, los esquemas predictivos utilizaron los trimestres de Mayo-Junio-Julio, Agosto-Setiembre-Octubre y el cuatrimestre de Diciembre-Enero-Febrero-Marzo como periodos a predecir de la precipitación. Como predictores se utilizaron diferentes índices asociados con fuentes de variabilidad climática que influencian los patrones climáticos de América Central, usando uno o dos bimestres anteriores a la estación predicha. Se encontraron esquemas predictivos útiles para prácticamente todas las relaciones señaladas anteriormente y se observó que gran parte de la variabilidad de América Central se puede explicar con los índices asociados a El Niño (La Niña) (variabilidad interanual) y del Atlántico (AMO, principalmente, variabilidad multidecadal).
The statistical technique of contingency table analysis was used to produce predictive schemes associated with rainfall in Central America. As a first step, principal component analysis was used to produce indices using 146 daily station records. Two rainfall components were obtained associated with Central America Pacific and Caribbean slopes. Keeping in mind that one of the work objectives is to support the Regional Climate Outlook Forums process, the predictive schemes used the trimesters of May-June-July, August- September-October and the four month period of December-January-February-March as targets for predictions in rainfall. Different climate indices were used as predictors, associated with several climate variability sources that influence the climate patterns in Central America, using one or two bimester previous to the predicted season. Useful predictive schemes were found for practically all the relationships mentioned previously, noticing that most of the Central America climate variability could be explained by the El Niño (La Niña) (e.g. interanual variability) and the Atlantic (AMO, mainly, e.g. multidecadal variability) indices.
The statistical technique of contingency table analysis was used to produce predictive schemes associated with rainfall in Central America. As a first step, principal component analysis was used to produce indices using 146 daily station records. Two rainfall components were obtained associated with Central America Pacific and Caribbean slopes. Keeping in mind that one of the work objectives is to support the Regional Climate Outlook Forums process, the predictive schemes used the trimesters of May-June-July, August- September-October and the four month period of December-January-February-March as targets for predictions in rainfall. Different climate indices were used as predictors, associated with several climate variability sources that influence the climate patterns in Central America, using one or two bimester previous to the predicted season. Useful predictive schemes were found for practically all the relationships mentioned previously, noticing that most of the Central America climate variability could be explained by the El Niño (La Niña) (e.g. interanual variability) and the Atlantic (AMO, mainly, e.g. multidecadal variability) indices.
Description
Artículo científico -- Universidad de Costa Rica. Centro de Investigaciones Geofísicas, 2012
Keywords
América Central, índices climáticos, variabilidad climática, Central America, climate indices, Climatología, Climate variability
Citation
http://webs.ono.com/reclim7/reclim12e.pdf