Pronóstico de El Niño en la zona 1+2 usando lenguaje Python
No Thumbnail Available
Date
2024
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Nacional de Trujillo
Abstract
En este informe se presentan los hallazgos de una investigación llevada a cabo sobre la influencia que tienen las TSM (temperatura superficial de mar) de 23 diferentes zonas oceánicas a nivel mundial con las anomalías de TSM de la zona 1+2 (90ºW - 80ºW, 0º - 10ºS). En la primera etapa de esta investigación se desarrolló un método para la obtención de un valor representativo de MMSST (Monthly Mean Sea Surface - Promedio mensual de la temperatura superficial del mar en español) en zonas o áreas delimitadas por coordenadas geográficas, el cuál consistió en 2 bloques de código desarrollados enteramente en lenguaje de programación Python, los cuales permitieron decodificar datos descargados en formato NC (netCDF) sobre el promedio mensual de la temperatura superficial de mar desde enero de 1854 hasta el mes de Febrero del 2023 a nivel global y obtener información útil a partir de ellos, es decir, obtener el Promedio por Área Georreferenciada de MMSST de distintas zonas oceánicas, además de observar el cambio de este valor representativo a lo largo del intervalo temporal mencionado. Con la información preliminar obtenida durante el desarrollo de la in-festinación, se creó un modelo de pronostico diseñado en base a redes neuronales artificiales donde se realizaron ensayos pero los resultados no fueron los esperados. Sin embargo, en la etapa consecuente, se encontró que 23 zonas de las mas de 40 zonas estudiadas son las que influyen de manera determinante en la TSM de la zona 1+2, y por ende, en sus anomalías que categorizan el Índice Costero El Nino (ICEN). Los resultados indicados obtuvieron valores ˜P<0.05, lo que muestra que son significativos estadísticamente, es decir, que no ocurrieron por casualidad y en consecuencia con esto se rechaza la hipótesis nula
Abstract This report presents the findings of an investigation carried out on the influence that the SST (sea surface temperature) of 23 different oceanic zones worldwide has on the SST ano malies of zone 1+2 (90ºW - 80ºW, 0º - 10ºS). In the first stage of this research, a method was developed to obtain a representative value of MMSST (Monthly Mean Sea Surface - Monthly average of sea surface temperature in Spanish) in zones or areas delimited by geographical coordinates, which consisted of 2 blocks of code developed entirely in Python programming language, which allowed decoding data downloaded in NC format (netCDF) on the monthly average sea surface temperature from January 1854 to the month of February 2023 at a global level and obtaining useful information from them, that is, obtaining the Average per Geore ferenced Area of MMSST of different oceanic zones, in addition to observing the change of this representative value throughout the mentioned time interval. With the preliminary infor mation obtained during the development of the research, a prognostic model designed based on artificial neural networks was created where tests were carried out but the results were not as expected. However, in the subsequent stage, it was found that 23 zones of the more than 40 zones studied are those that decisively influence the SST of zone 1+2, and therefore, in its anomalies that categorize the Coastal Index. Child (ICEN). The indicated results obtained values P<0.05, which shows that they are statistically significant, that is, that they did not occur by chance and consequently, the null hypothesis is rejected
Abstract This report presents the findings of an investigation carried out on the influence that the SST (sea surface temperature) of 23 different oceanic zones worldwide has on the SST ano malies of zone 1+2 (90ºW - 80ºW, 0º - 10ºS). In the first stage of this research, a method was developed to obtain a representative value of MMSST (Monthly Mean Sea Surface - Monthly average of sea surface temperature in Spanish) in zones or areas delimited by geographical coordinates, which consisted of 2 blocks of code developed entirely in Python programming language, which allowed decoding data downloaded in NC format (netCDF) on the monthly average sea surface temperature from January 1854 to the month of February 2023 at a global level and obtaining useful information from them, that is, obtaining the Average per Geore ferenced Area of MMSST of different oceanic zones, in addition to observing the change of this representative value throughout the mentioned time interval. With the preliminary infor mation obtained during the development of the research, a prognostic model designed based on artificial neural networks was created where tests were carried out but the results were not as expected. However, in the subsequent stage, it was found that 23 zones of the more than 40 zones studied are those that decisively influence the SST of zone 1+2, and therefore, in its anomalies that categorize the Coastal Index. Child (ICEN). The indicated results obtained values P<0.05, which shows that they are statistically significant, that is, that they did not occur by chance and consequently, the null hypothesis is rejected
Description
Keywords
Lenguaje de programación Python, área georreferenciada, temperatura superficial de mar, índice costero El Nino