Determinación de umbrales horarios de temperatura en la provincia de Chimborazo, mediante metodología estadística con datos funcionales

Abstract

The province of Chimborazo faces the challenge of not having specific hourly thresholds for environmental temperature, which makes it difficult to clean, validate and impute missing data in records of this variable. In response to this need, this research focused on defining such thresholds using functional statistical techniques. Using data collected from 11 weather stations monitored by the GEAA, an exploratory functional analysis was performed to understand the temporal variability of temperatures, identifying seasonal patterns and groupings between stations with similar thermal behavior. Subsequently, the functional Bootstrap method was applied to obtain null distributions of temperatures, which provided accurate estimates of climate variability and helped to evaluate the uncertainty of the measurements. Confidence bands were established for temperatures at each station, facilitating the assessment of data reliability. Hourly temperature thresholds were determined at each station, successfully validated using the vertical temperature gradient, confirming their accuracy and relevance to thermal variations at different altitudes. This methodology identified seasonal groups and made it possible to distinguish two seasons with different temperatures, obtaining validated hourly thresholds that significantly improve the understanding and management of climate data in the province of Chimborazo.