3.2.1. The Heat Wave of the Year 1987

Listed as one of the major natural disasters by Berz [40], the heat wave of the summer of 1987 in Greece resulted in over a thousand deaths all over the country due to heat strokes, heat exhaustion, and heart-related conditions [16,30]. As a result, the heat wave of 1987 has been the subject of several studies, all agreeing that even though higher absolute Tmax values had been recorded before, the duration of these very intense and severe hot conditions was the main characteristic of this specific heat wave [29]. Using two physiological discomfort indices, Giles et al. [41] confirmed that this heat wave was continuous. For nine consecutive days, from 19 to 27 July [29], Tmax values exceeded 40 ◦C for most areas of Greece [41]; minimum temperatures were also relatively high and the days of the heat waves were characterized by a rather small diurnal range [29]. Matzarakis and Mayer [30] also mention that the thermal indices used in their study showed a very high thermal stress on people. Especially in the case of Athens, each afternoon was considered as "extremely hot" [31] and the heat wave was found to be more intense in northern Greece, especially in Thessaloniki [31,41].

Because of the above characteristics of the heat wave of 1987, this study now attempted to investigate if the EHF index was efficient enough to detect the heat wave, to identify it, and to provide a thorough analysis of this extremely hot summer in Greece. For this purpose, the acclimatization and the significance indices (EHIaccl and EHIsig) as well as the EHF index were computed during the month of July 1987 for the 14 stations under study in comparison with the daily Tmean of that month and the Tmean95 of the reference period 1971–2000 (Figure 4). For all the stations, a gradual increase of the Tmean daily values was observed, rising above the Tmean95 value from 17 to 18 July. The peak of these daily values was found during days 25 to 27 of the month, and they dropped again below the 95th percentile at the end of the month (from 29 to 30 July 1987). If the study of the heat wave was based only on the daily mean temperatures that exceeded the Tmean95, then the duration of the heat wave would be defined from days 18 to 30 of the month. However, using the EHF index, additional and more detailed information was provided.

More specifically, EHIaccl began to have positive values much sooner, on 12–13 July, meaning that the averaged three-day temperature was higher than the recent past, according to the index definition. This indicated that there was now a lack of human acclimatization to the upcoming warmer conditions, which could result in an adverse impact on their everyday life and health [12]. Thus, even though the "actual" heat wave had not started, this rising index (EHIaccl) indicated that humans were not able to physically adapt to this warming, which could be used as a useful alert for heat wave policy management measures. The second computed index (EHIsig) turned positive for several days (in general on 20 July). This was the starting point of the heat wave, since by definition, the EHIsig values should be positive in order to consider the temperature conditions abnormally hot, higher than the 95th percentile Tmean. Both the indices dropped below zero (yellow line in Figure 4) at the end of July, either on day 29 or 30 of the month. In most of the stations, these were the same days that the daily Tmean also fell below the 95th percentile value. Overall, the examination of the combined EHF index showed that the heat wave of July 1987 started on day 20 of the month (in most of the stations under study) ending on day 30 when the EHF values were again negative. An interesting finding of the application of this index was that the peak of the heat wave was not placed on the same day that the actual daily mean temperature values reached their maximum, but one or two days later. In addition, it is worth mentioning that during this specific heat wave, not only were the temperatures exceptionally high (indicated by the EHIsig) but, due to the rapid temperature increase, people did not have the chance to acclimatize themselves to these new "much hotter" conditions (positive EHIaccl), resulting in harmful effects on their health. Moreover, the previous studies' finding that the heat wave was more intense in northern Greece was also detected from the index application that showed higher values for the stations in the northern parts of the country. It is obvious that the definition and the description of this heat wave using the EHF match the previous findings, making it a suitable index for the study of heat waves.
