Climate, Climate Variability and Impacts in the Mediterranean Area: An Overview
39
The response time of the sea to these extreme events is of the order of years and
probably the event experienced in 1987 was not the beginning of the formation of
additional deep salty water. The 1981 event had a similar effect. Its impact can be
traced at 420 m depth until 1985 in the amplitude of time series of the first EOF
mode of the 1981 cooling (Korres et al 2000b).
Long term deep water warming was also measured in the Alghero-Provencial
basin between 1000 and 2700 m (Bethoux and Gentili, 1996). Between 1959 and
1995 a warming of O.13'C and an increase of salinity of 0.04 psu were measured.
Twice as high values are reported (see Zodiatis and Gasparini 1966) for the
intermediate waters in the Ligurian Sea and the Tyrrhenian Sea which may have
been imported from the eastern Mediterranean basin. There are different hypotheses
about the causes of the warming in the western deep waters. Temporarily the
beginning of these effects seem to coincide with the damming of the Nile and of
Black sea rivers as well as a decrease of rainfall around the northwestern
Mediterranean Sea.
Evaporation computed from latent heat fluxes obtained from ECMWF budgets
result in 1100 mm year,l which is out of the range suggested by Castellari. The
precipitation would be 450 mm year,l, which is smaller than the 590 mm year' I of
the climatology. The long term means of E - Pis 650 mm year I (climatology: 950
mm year,I). However this "climatology" was performed for two years only in which
a heat accumulation occurred in the Mediterranean basin (1987 and 88).
The Adriatic was in more detail studied by Artegiani et al. (1997) and Maggiore
et al. (1998). The obtained mean surface heat budget indicates a loss of 19 - 22 W
m,2. This should be compensated by heat advection through the channel of Otranto.
The surface water flux E - P varies between 60 and 520 mm year' I : E ranges between
1080 and 1340 mm year I , P between 820 and 1020 mm year' I , and the river run-off
is 1170 mm year,l. Thus there is a gain of water of the order of 650 to 1100 mm
year I which means that the Adriatic is a dilution basin with respect to salinity. Its
winter surface heat loss leads to deep water formation.
There is a lack of results to confirm that a relationship exists between the
variability of the sea surface heat fluxes over the sea and climate variability - such
as precipitation changes - over land. In rainfall statistics no marked effect of the
1981 and 1986 events have yet been found. Maracchi et al. (2000) investigated the
relationship between the SST of the northern Thyrrenian Sea according to COADS
data and the precipitation in Tuscany. They found a correlation e. g. between the
September SST and extreme rainfall events in October as well as a correlation
between extreme rainfall events in December and SSTs in June and conclude that
an increase of the SST by 1.8'C may cause an increase between 40% and 65% of
extreme monthly rainfalls. They investigated the time series of SSTs from 1950 to
1997 and came to interesting results. Firstly, monthly SSTs deviate from the 27
years mean by about ± 2C. Secondly, the trend varies considerably with the length
of the interval. For the whole period it is only 0.02CIl 00 years which is not
significant. But for shorter intervals, such as 1970 - 1997, it can amount to 3.3
CII 00 years with largest values in May and August. This is due to the fact that the
39
The response time of the sea to these extreme events is of the order of years and
probably the event experienced in 1987 was not the beginning of the formation of
additional deep salty water. The 1981 event had a similar effect. Its impact can be
traced at 420 m depth until 1985 in the amplitude of time series of the first EOF
mode of the 1981 cooling (Korres et al 2000b).
Long term deep water warming was also measured in the Alghero-Provencial
basin between 1000 and 2700 m (Bethoux and Gentili, 1996). Between 1959 and
1995 a warming of O.13'C and an increase of salinity of 0.04 psu were measured.
Twice as high values are reported (see Zodiatis and Gasparini 1966) for the
intermediate waters in the Ligurian Sea and the Tyrrhenian Sea which may have
been imported from the eastern Mediterranean basin. There are different hypotheses
about the causes of the warming in the western deep waters. Temporarily the
beginning of these effects seem to coincide with the damming of the Nile and of
Black sea rivers as well as a decrease of rainfall around the northwestern
Mediterranean Sea.
Evaporation computed from latent heat fluxes obtained from ECMWF budgets
result in 1100 mm year,l which is out of the range suggested by Castellari. The
precipitation would be 450 mm year,l, which is smaller than the 590 mm year' I of
the climatology. The long term means of E - Pis 650 mm year I (climatology: 950
mm year,I). However this "climatology" was performed for two years only in which
a heat accumulation occurred in the Mediterranean basin (1987 and 88).
The Adriatic was in more detail studied by Artegiani et al. (1997) and Maggiore
et al. (1998). The obtained mean surface heat budget indicates a loss of 19 - 22 W
m,2. This should be compensated by heat advection through the channel of Otranto.
The surface water flux E - P varies between 60 and 520 mm year' I : E ranges between
1080 and 1340 mm year I , P between 820 and 1020 mm year' I , and the river run-off
is 1170 mm year,l. Thus there is a gain of water of the order of 650 to 1100 mm
year I which means that the Adriatic is a dilution basin with respect to salinity. Its
winter surface heat loss leads to deep water formation.
There is a lack of results to confirm that a relationship exists between the
variability of the sea surface heat fluxes over the sea and climate variability - such
as precipitation changes - over land. In rainfall statistics no marked effect of the
1981 and 1986 events have yet been found. Maracchi et al. (2000) investigated the
relationship between the SST of the northern Thyrrenian Sea according to COADS
data and the precipitation in Tuscany. They found a correlation e. g. between the
September SST and extreme rainfall events in October as well as a correlation
between extreme rainfall events in December and SSTs in June and conclude that
an increase of the SST by 1.8'C may cause an increase between 40% and 65% of
extreme monthly rainfalls. They investigated the time series of SSTs from 1950 to
1997 and came to interesting results. Firstly, monthly SSTs deviate from the 27
years mean by about ± 2C. Secondly, the trend varies considerably with the length
of the interval. For the whole period it is only 0.02CIl 00 years which is not
significant. But for shorter intervals, such as 1970 - 1997, it can amount to 3.3
CII 00 years with largest values in May and August. This is due to the fact that the