Reconstruction of data and detection of heat and cold waves in Porto and neighbouring municipalities – Portugal

Abstract

Reconstruction of the historical series was performed using simple linear regression analysis (RLS), and then the existing data from the classic (EMC) and automatic (EMA) weather stations were transformed in single and continuous series. Heat waves (OdC) and cold waves (OdF) were detected from daily temperature data obtained between 1970 and 2017. The RLS for the maximum temperature and for the minimum temperature between the four stations were significant (p < 0.0001). The correlations (rs) between the series reconstructed for the stations of the same sites were very strong (p < 0.01), ranging from 0.93 to 0.99. Based on the Mann-Kendall test (α = 5%), the OdC in Pedras Rubras tend to be more frequent and more durable. However, for the OdC in the Serra do Pilar station, all parameters showed a significant upward trend. As for the OdF in Pedras Rubras, there is a trend towards less frequent, shorter and more intense events. The OdF in Serra do Pilar were less frequent, shorter and less severe.