Stochastic analysis of temperature records in Santa Ana - Manabí - Ecuador

Abstract

In the present work, time series are modeled applying the Box - Jenkins methodology on the data collected from a weather station for the variables vapor pressure, dry thermometer, wet thermometer, dew point and evaporation.  Obtaining as result several adjustments of autoregressive models and mobile averages in the different variables were found. For example, in the first station, the steam tension was adjusted to an Autoregressive model with a delay AR (1), that of the dry thermometer was adjusted to an Autoregressive model with a delay of grade 3 AR (3) and moving averages with a delay of grade 5 MA (5) and the evaporation was adjusted to an Autoregresive model with a delay AR (1) and moving averages with a challenge of grade 5 MA (5).