Extreme temperatures changes from 1950 to 2018 in 5 Mediterranean cities: trend analysis from E-OBS time-series data

11:15 Tuesday 28 May


Room S2


Ana Oliveira (Portugal) 1; António Saraiva Lopes (Portugal) 2; Samuel Niza (Portugal) 3

1 - MIT Portugal, Instituto Superior Técnico, Universidade de Lisboa; 2 - IGOT - Instituto de Geografia e Ordenamento do Território/Centro de Estudos Geográ ficos, Universidade de Lisboa; 3 - IN+, Instituto Superior Técnico, Universidade de Lisboa

Climate change prospects for the Mediterranean bioregion indicate higher average temperatures and more frequent, intense and longer heatwaves. In this paper, recent trends in terms of extreme temperatures were examined in five Mediterranean cities: Athens (Greece), Barcelona (Spain), Lisbon (Portugal), Marseille (France) and Naples (Italy).

Maximum Temperature (TX) and Minimum Temperature (TN) time-series data (daily values from 1950 to 2018) were extracted from the E-OBS gridded dataset (ensemble version 17.e) at the location of each city. Using the ClimPACT2 software, extreme heat temperatures were identified on a yearly basis, using indices approved by the ET-SCI, among others: summer days (SU), tropical nights (TR), warm spell duration indicator (WSDI), warmest TX (TXx), amount of hot days (TX90p), amount of warm nights (TN90p), Excess Heat Factor (EHF, frequency, duration and intensity indices) and Cooling Degree-days (CDD). Extreme cold indices, mirror the previous one’s: cold spell duration indicator (CSDI), coldest TN (TNn), amount of cool days (TX10p), amount cold nights (TN10p), Excess Cold Factor (ECF, frequency, duration and intensity indices) and Heating Degree-days (HDD).

A Mann Kendall’s tau-b correlation was run to determine the trends of each index, from 69 observations (yearly values). Significant (p < 0.01) positive correlations were found regarding the number of SU, TR and percentage of days with temperatures above the TX90p or TN90p threshold. Also, the frequency of heatwaves, as defined by the EHF, were shown to be increasing with significant correlations. Accordingly, CDD were also found to be increasing. On the other hand, remaining extreme heat indices related to the intensity or duration of heatwaves showed positive correlations without significance. Regarding cold temperatures, the opposed situation occurred, with significant negative correlations found amongst indices that express number of days with extreme cold, namely the amount of days with temperatures below TX10p or TN10p, the frequency of coldwaves expressed by the ECF, as well as HDD.

Furthermore, coldest yearly temperatures expressed by the TNn index presented significant increasing trends in most cities. Hence, main results indicate that the warming process has been translated in increasing frequency of heatwaves and number of days involved, paralleling significant decrease in the number of coldwaves and number of extreme cold days. Conversely, intensity indices regarding extreme cold temperatures showed more significant positive trends than their extreme heat counterparts, thus implying that TN has been increasing more significantly than TX. Future work will consider climate change projections to address future extreme temperatures trends.