Modelos estadísticos para la detección de patrones en medio ambiente y economía aplicación de gráficos de control funcional para monitorizar series temporales meteorológicas y climáticas para la detección de anomalías.

Abstract

One of the tools of analysis and problem detection in statistical control is the control chart, used to collect and analyze data in order to monitor, control and identify variations in the process that are due to common and special causes. Phase I of the control chart is proposed using depths that will be ex tended to the analysis of functional data to monitor smooth curves, ca lling this method as functional control charts, where the modal depth and bootstrap procedures are considered to determine the quantile and perform the functional control chart of ranges to detect outlier functions. The method was applied to meteorological data of ambient temperatu re from 7 stations in the Chimborazo province monitoring and comparing the climate over time. Detecting that the atypical data are part of one or several stations produced by natural phenomena presented in the results section