<p>Climate change has caused an accelerated increase in temperature worldwide, which has brought negative consequences for the environment and our society. This study aims to determine monthly temperature patterns in six Chilean cities between 1980-2022 (Calama, Antofagasta, Santiago, Valparaíso, Chillán, and Concepción) using Functional Data Analysis (FDA). The cities are located in the north, center, and south of the country, as well as on the coast and valleys, representing diverse climates in Chile. The methodology combines a statistical description of the data, functional principal component analysis (FPCA), and cluster analysis. FPCA explains the covariance structure of the data, identifying and modeling the main modes of variation as continuous functions. Cluster analysis groups similar patterns in the temperature series, identifying similar climatic behaviors across cities. Results show increasing anomalies in the 21st century, especially for Calama and Santiago. The average minimum temperature in Calama and Santiago has increased steadily over the last 40 years. The trend of the anomaly is +0.36°C/decade for Calama and +0.38°C/decade for Santiago. Significant change points are detected in the maximum temperature (September 2001) and minimum temperature (October 2011) of Calama. Findings may also indicate that minimum temperatures have been increasing from the north to the south and have already reached the central part of the country. Average minimum temperature has increased first in the northern cities, then in the central regions of the country, and the increase may advance towards the south of Chile at some point. The sea tempers the temperatures in the coastal cities.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Functional PCA and cluster analysis for determining temperature patterns in Chile

  • Matilda Tapia-Díaz,
  • Alba Martínez-Ruiz,
  • Pablo Lemus-Henríquez

摘要

Climate change has caused an accelerated increase in temperature worldwide, which has brought negative consequences for the environment and our society. This study aims to determine monthly temperature patterns in six Chilean cities between 1980-2022 (Calama, Antofagasta, Santiago, Valparaíso, Chillán, and Concepción) using Functional Data Analysis (FDA). The cities are located in the north, center, and south of the country, as well as on the coast and valleys, representing diverse climates in Chile. The methodology combines a statistical description of the data, functional principal component analysis (FPCA), and cluster analysis. FPCA explains the covariance structure of the data, identifying and modeling the main modes of variation as continuous functions. Cluster analysis groups similar patterns in the temperature series, identifying similar climatic behaviors across cities. Results show increasing anomalies in the 21st century, especially for Calama and Santiago. The average minimum temperature in Calama and Santiago has increased steadily over the last 40 years. The trend of the anomaly is +0.36°C/decade for Calama and +0.38°C/decade for Santiago. Significant change points are detected in the maximum temperature (September 2001) and minimum temperature (October 2011) of Calama. Findings may also indicate that minimum temperatures have been increasing from the north to the south and have already reached the central part of the country. Average minimum temperature has increased first in the northern cities, then in the central regions of the country, and the increase may advance towards the south of Chile at some point. The sea tempers the temperatures in the coastal cities.