 in 1940 [Béthoux and Gentili, 1997]. The increased latent heat flux  (1.6
  in 1940 [Béthoux and Gentili, 1997]. The increased latent heat flux  (1.6  in 
1995) corresponds  to a surplus of water loss by evaporation of 2  cm , which may explain, in 
part, the increased salinity of Mediterranean waters.
Known changes in freshwater input are decreases
in the outflow of the Nile river since the closing of High Dam in 1964 [Wadie, 1984],
and the Ebro river since the 1960s [Martin and Milliman, 1997],
about  2
   in 
1995) corresponds  to a surplus of water loss by evaporation of 2  cm , which may explain, in 
part, the increased salinity of Mediterranean waters.
Known changes in freshwater input are decreases
in the outflow of the Nile river since the closing of High Dam in 1964 [Wadie, 1984],
and the Ebro river since the 1960s [Martin and Milliman, 1997],
about  2  
  
   and 1
 
and 1  
  
   ,
respectively.
These two man-made 
changes correspond to a diminished layer of freshwater input at the sea surface of about  3.6 cm
 ,
respectively.
These two man-made 
changes correspond to a diminished layer of freshwater input at the sea surface of about  3.6 cm  
 . Another anthropogenic cause of salinity increase may be an increasing flow of 
saline water from the Red Sea into the Mediterranean.
At its
opening in 1869,
the cross-sectional area of
the Suez Canal was 304
 . Another anthropogenic cause of salinity increase may be an increasing flow of 
saline water from the Red Sea into the Mediterranean.
At its
opening in 1869,
the cross-sectional area of
the Suez Canal was 304  ,
compared to 3600
 ,
compared to 3600  in 1981,
after different deepenings and widenings
[Halim, 1990].
The changes favour the natural inflow of Red Sea water towards the 
Mediterranean Sea. With a rough estimate of a  mean current of about 0.25 m
  in 1981,
after different deepenings and widenings
[Halim, 1990].
The changes favour the natural inflow of Red Sea water towards the 
Mediterranean Sea. With a rough estimate of a  mean current of about 0.25 m   in the Suez 
Canal, this salinity input may be equivalent to the effect of an evaporation increase of about 1.3 
cm  over the entire sea.
  in the Suez 
Canal, this salinity input may be equivalent to the effect of an evaporation increase of about 1.3 
cm  over the entire sea.
The other probable change in freshwater budget corresponds to a decrease in precipitation, observed around the western basin since at least the 1940s and also reported from some observations in the eastern basin [Bradley et al., 1987]. From calculated evaporation increase, known river runoff decreases and simulated freshwater deficit change, the estimated precipitation decrease in 1995 amount either to 4.4 cm or to 3.1 cm , if the estimated effect from Suez Canal inflow is taken into account. These values correspond to a 14% or a 10% decrease of precipitation over the whole sea; the previous estimate from records from the north-western basin suggested a 11% decrease [Béthoux and Gentili, 1996]. Regional changes of river runoff (Nile and Ebro rivers) and the link with the Red Sea may correspond to an increase of the freshwater deficit of about 4.9 cm in 1995, i.e. about 50% of the estimated change from salinity increase. Increase in evaporation and decrease in precipitation, due to global climate and probably to GHE change, may correspond to the remaining 50% salinity change.