Technical ReportPDF Available

Climate variability and extreme weather events in the Mediterranean region

Climate variability and extreme weather events
in the Mediterranean region
A large increase in greenhouse gases is a driver of climate change, leading to many other changes in
the atmosphere, on land, and in the oceans. Climate change refers to a statistically significant variation
in either the mean state of the climate or in its variability. So, the change of climate can be the cause of
Climate Variability, it has consequences for sea level changes, plant life, and mass extinctions; it also
affects human societies.
The Mediterranean region has been identified as one of the most sensitive regions in the world to climate
change (Thiébault et al, 2016). Several studies have observed changes in precipitation and temperature
in the French Mediterranean, that can evolve to heavy precipitation, windstorms, heat waves and
In this study, we observe climate variability, including variation in precipitation and temperature in
French Mediterranean using reanalysis ERA5 and in situ data. The goal of the study is to determine
whether global warming will change climate variability of extreme events? And what trend is that
change ?
1.1 Heat wave
According to the World Meteorological Organization, Heatwave is an extreme weather event with
marked warming of the air over a large area that usually lasts from a few days to few weeks. The
Mediterranean basin is a region where hot conditions and heatwaves are popular in the summer. The
frequency of heat waves has been increasing in recent decades in Europe and specifically in the
Mediterranean region due to climate change (Thiébault et al, 2016). Since the 1960s, the Mediterranean
region has become warmer with the frequency, intensity and duration of heat waves has been increasing
(Kuglitsch et al. 2010) and it has a lot of impact on human health, especially in areas with vulnerable
1.2 Cevenol episode
On the Mediterranean rim, strong precipitation events take place during the autumn. These violent
events, lasting one to several days, are accompanied by significant floods and therefore present a major
hazard for the inhabitants of these regions (méteo France). These events take place when warm air
masses loaded with humidity from the Mediterranean in late summer move towards the interior of the
continent. These humid air masses cool rapidly as they approach the Cevenes (a small mountain range
in the south of France) and condense (Fig. 1).
Figure 1. Operation of cevenol episode (Méteo France)
Figure 2. Annual distribution of
The large temperature difference between the continent and the sea induces a great instability which
allows the appearance of a significant and very intense stormy activity (precipitation often exceeding
100mm/h). We are, therefore, going to focus on the impacts of climate change on these extreme events,
in particular on the frequency of these events each year.
In the Languedoc-Roussillon region, most of bigger precipitation events occur during these episodes in
autumn, as shown in figure 2 (data from Meteo France). This figure shows the number of days per
month since 1958 when there was more than 160mm of precipitation. For all of the departments (in
particular for Gard and Hérault) these events take place in autumn, these are Cévennes episodes.
2.1 Analysis of temperature variations
For evaluating variation of temperature, data of European Climate Assessment & Dataset (ECA&D)
was used. We choose to restrict our study to the period 1990-2019 in order to highlight recent climatic
trends, the three variables analyzed are minimum/night temperatures (Tn), maximum/day temperatures
(Tx), average temperatures (Tn) for 4 stations at study area (Cepet, Montpellier, Marignane, Perpignan).
Figure 3. Study area and monitoring stations
2.1.1 Variation of annual average temperature
Basic calculations were carried out on the raw series in order to analyse the variations in temperature
over time:
- Check if the data is missing in some series;
- Calculate an average annual temperature for all stations. The purpose is to observe the changes
of temperature (Tn, Tx, Tg) with the time series in the study area.
2.1.2 The temperature anomaly
The temperature anomaly is the difference between the temperature measured in a place in degrees
Celsius, positive or negative, compared to the normal average temperature (calculated over a period of
at least 30 years) daily, seasonal or annual observed in a homogeneous geographic region. The
temperature anomaly is calculated by equation below:
    
2.1.3 Number of heat wave events
In several studies, heatwave definitions were tested by using percentiles (e.g., from the 75th to 99th
centile) of mean temperature with duration ≥2 days. In this study, we chose a threshold for the heatwave:
95th centile, lasting for 5 days. We counted the number of heatwave occurrences per year at each station
to observe extreme temperatures between 1990-2019.
2.2 Cevenol episode
2.2.1 Méteo France data
Precipitation in the Languedoc-Rousillon region has been measured since 1958 by Météo France. The
data used for this study are the time series of the number of days per year when precipitation exceeds a
certain threshold (here 60mm / day and 160mm / day).
These data will allow us to observe the evolution since 1958 in the number of days of heavy
precipitation. However, these data do not detail the measured quantity of precipitated water, only the
number of days exceeding the threshold is provided.
2.2.2 ERA5 data
To obtain the precipitation values over the study area and therefore to determine which precipitation
event can be considered extreme. It was necessary to have time series with good geographical coverage
of the area and a daily measurement step.
For reasons of data availability, the ERA 5 reanalysis of the European Center for Medium-Term
Weather Forecasts (ECMWF / ECMWF) was used. A meteorological reanalysis is a numerical method
of integrating meteorological data across the globe while performing interpolations between the data
using a model.
These reanalyzes have a spatial resolution of 30km and produce hourly estimates of many atmospheric
and climatic variables since 1979. The ERA points used are distributed along the Mediterranean coast,
as shown in figure 4.
Figure 4. ERA 5 data
3.1 Analysis of temperature variations
3.1.1 Variation of annual average temperature
In general, there was a gradual increase in temperature during the period 1990 - 2019. The trend of the
30-year lowest, average, and highest annual temperatures are shown in Figure 5. Since 1990, the lowest
temperature has increased from 11oC to 12oC until 2019 (p <0.01). These results show the clear impact
of climate change on temperature in the French Mediterranean. The lowest temperature increased
continuously for 29 years, about 0.3oC/10 years. Likewise, the annual average and highest temperature
also increased gradually during the survey period, 15.5oC - 16.3oC and 19.7oC - 21.2oC respectively (p
Figure 5. The annual lowest temperature (Tn), mean temperature (Tn), highest temperature (Tx) in
French Mediterranean from 1990 to 2019. The dash lines are trend lines, obtained by linear
The results of the average temperature anomalies in French Mediterranean between 1990 and 2019 were
divided into two phases (Figure 6). Negative anomalies frequently existed between 1990 and 2013,
while positive anomalies only existed from 2014 to 2019. In addition, in the first phase there existed
positive anomalies in some years, notably 1994, 1997, 2003, 2006, 2009 and 2011. In contrast, in the
second phase, there were only positive anomalies. This result shows that temperature anomalies
(increase trend) occurred strongly in recent years (since 2014).
Figure 6. Average temperature anomalies in French Mediterranean in period 1990 2019
Figure 7 depicts the number of heatwave events in four stations from 1990 - 2019. The annual number
of heatwaves at the stations had strong fluctuations, but most were less than 5 events per year except
for Cepet station. Cepet station was the station with the most heatwave events in recent years, more
than 20 events per year. Merignane and Perpignan stations had only one year with the number of events
more than 20 times while Montpellier station had only a maximum of 13 events in 2003. In addition,
the results also showed that 2003 was the year with the most heatwave events. This result is consistent
with the observation of the number of people affected by prolonged high temperature in France in 2003.
Besides, in recent years (from 2015), the number of events at stations has also tended to increase by
more than 5 events per year at Marignane and Cepet stations.
Figure 7. Number of heatwave events in Cepet, Montpellier, Marignane and Perpignan stations from
1990 2019.
3.2 Extreme precipitation
3.2.1 Evolution of major events since 1958
Using Meteo France data, we can determine the evolution of the number of days exceeding a
precipitation threshold for each year since 1958 (figure 8). The basic statistical data of these two series
are summarized in table 1.
Table 1. The basic statistical data
58.89 day/year
18.67 day/year
21 day/year
6.58 day/year
4.39 day/year
0 day/year
We notice a high variability in these data and the absence of a general trend (the rolling 20-year average
remains close to the 62-year average).
Figure 8. Evolution of extreme events, Météo France
These two graphs do not allow us to conclude on the impact of climate change on major precipitation
events in the study area. However, we are limited by Meteo France data which only shows the number
of days and not the amount of precipitation. Therefore, we cannot determine if the threshold of
60mm/day corresponds to the precipitation threshold that described the limit between extreme and
normal events.
So, we used the ERA5 data to answer this question. By plotting the histograms and cumulative
histograms for each of the ERA5 points (Appendix 1), we can determine the value of the threshold limit
which is defined as the amount of precipitation that only 0.5% (5 percentile) of events exceed.
We obtain an average threshold value of 41.79 ± 9.22 mm/day (1σ). This threshold value allows us to
determine the number of days per year corresponding to an extreme event. As shown in figure 8b
(number of days with extreme event per month since 1979), the events selected by this method
correspond for the most part to fall precipitation, which we will assume to be Cévennes episodes.
3.2.2 The evolution of the number of extreme events determined with the ERA5
Figure 9a shows the evolution of the number of extreme events determined with the ERA5 data. The
number of days per year is quite variable (as for France weather data) but a very slight trend can be seen
from these data.
Figure 9. Evolution of extreme events, Météo France
The linear regression on the time series shows us that every 10 years the number of extreme events
increases by 1.5. However, the p-value of the regression being equal to 0.42 (there is a 42% chance of
being wrong in accepting this regression), it is difficult to conclude on the existence of this trend.
4.1 Heat wave
After analyzing the temperature data of 4 stations in the French Mediterranean in the period 1990 -
2019, the results show that the temperature in this area has increased. Since 2014, temperature
anomalies tend to be 0.5 - 0.90C higher than the average temperature of the period, much higher in 2003
(the year with the historically high temperature in France). Heatwave events in 2015 also showed signs
of increase compared to previous years, especially at Cepet station. It can be concluded that the
temperature in French Mediterranean has changed and tends to increase in the future.
4.2 Cevenol episode
Using weather data from France and ERA5, we were able to observe that the occurrence of a Cevennol
episode has not changed since the 1950s. To confirm this result, it would be necessary to use long times
series of measurements of precipitation.
Indeed, ERA5 data only cover the period with satellite observations (since 1979) and only allow us to
observe the absence of a trend over the last 35 years. Appendix 2 also shows that the annual precipitation
averages of these events have been relatively constant since 1979, so there is no change in the violence
of the event either. Another element was not taken into account in this study: the duration of these
events, which could also vary in a context of climate change.
[1] Thiébault, Stéphanie & Moatti, Jean-Paul & Eds, & Ducrocq, V. & Gaume, Eric & Dulac, François
& Hamonou, E. & Shin, Yunne-Jai & Joel, Guiot & Boulet, Gilles & Guégan, Jean-François & Barouki,
Robert & Annesi-Mae, Isabella & Marty, Pascal & Torquebiau, Emmanuel & Sou, Jean-François &
Aumeeruddy-Thomas, Yildiz & Chotte, Jean-Luc & Lacroix, Denis. (2016). The Mediterranean Region
Under Climate Change. A Scientific Update.
[2]. Kuglitsch, F. G., Toreti, A., Xoplaki, E., Della‐Marta, P. M., Zerefos, C. S., Türkeş, M., &
Luterbacher, J. (2010). Heat wave changes in the eastern Mediterranean since 1960. Geophysical
Research Letters, 37(4).
Appendix 1: Histogramme and cumulative histogramme of ERA5 precipitation
Appendix 2: Evolution of precipitation since 1979
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
Heat waves have discernible impacts on mortality and morbidity, infrastructure, agricultural resources, the retail industry, ecosystem and tourism and consequently affect human societies. A new definition of socially relevant heat waves is presented and applied to new data sets of high-quality homogenized daily maximum and minimum summer air temperature series from 246 stations in the eastern Mediterranean region (including Albania, Bosnia-Herzegovina, Bulgaria, Croatia, Cyprus, Greece, Israel, Romania, Serbia, Slovenia, Turkey). Changes in heat wave number, length and intensity between 1960 and 2006 are quantified. Daily temperature homogeneity analysis suggest that many instrumental measurements in the 1960s are warm-biased, correcting for these biases regionally averaged heat wave trends are up to 8% higher. We find significant changes across the western Balkans, southwestern and western Turkey, and along the southern Black Sea coastline. Since the 1960s, the mean heat wave intensity, heat wave length and heat wave number across the eastern Mediterranean region have increased by a factor 7.6 ±1.3, 7.5 ±1.3 and 6.2 ±1.1, respectively. These findings suggest that the heat wave increase in this region is higher than previously reported.