150
1. Luterbacher and E. Xoplaki
dry and cool during the second part of the 19 th century. The distinct trend towards
Mediterranean winter dryness during the last few decades of the 20 th century is
unique in the whole 500 year precipitation time series.
The reconstruction quality is generally better for temperature than for
precipitation. Seasonal reconstructions are less trustworthy than the monthly
estimates, but they still show some skill over various Mediterranean areas allowing
qualitatively and quantitatively statements about the temperature and precipitation
course over the last few centuries.
The patterns of the coldest (1891), the warmest (1772), the driest (1609) and the
wettest (1684) Mediterranean winters over the last few centuries have been shown
together with the departures from the 20 th long-term mean and the quality of the
reconstructions. Clear temperature and precipitation anomalies have been found
with inverse sign for Southern Europe/Northern Africa and Northern Europe.
Although only a few predictors were available for the larger Mediterranean for the
driest and wettest winters, there is potential for skillful reconstructions for some
Mediterranean regions.
The reconstructed sea level pressure fields (Luterbacher et al. 2001a) related to
these extremes clearly showed the expected anomalies with cold/warm, and
moist/dry air advection towards the Mediterranean.
We found that the relationship between the reconstructed winter NAO and the
mean winter Mediterranean precipitation remained stable over the last 5 centuries.
In periods with stronger westerlies over the eastern North Atlantic, the
Mediterranean area, as a whole, obtained less precipitation and vice versa.
The North Atlantic Oscillation it is not the most important signal of atmospheric
variability concerning temperature over the averaged larger Mediterranean region.
This mainly reflects the fact that the averaged mean temperature value includes both
areas with a strong link to the NAO and others with no connection at all. Their
competitive effects may be the reason for the low correlations over the last 500 years.
Preliminary results dividing the Mediterranean into sub-regions indicate close
relation between the winter NAO and winter temperature over the Eastern
Mediterranean whereas for the Western and Central parts this is not the case.
The presented continuous temperature and precipitation reconstructions over
Europe, North Africa and the Near East are of great importance for all kind of
climatic analyses dealing with natural variabilities on time scales from months to
decades or centuries. They can be used to study the low and high frequency
variability and the characteristics and extremes of climate and the relationship with
the atmospheric circulation and can be compared with model-produced
reconstructions of natural and forced (external and internal) variability for the last
centuries.
Further investigations will include new predictor data in order to improve the
reconstruction skill, to provide time-dependent uncertainty ranges of our estimates,
to study the influence of forcing factors, and to study the low frequency climate
variability for different climate sensitive regions along the Mediterranean for all the
four seasons over the last centuries and the comparison with reconstructions from
1. Luterbacher and E. Xoplaki
dry and cool during the second part of the 19 th century. The distinct trend towards
Mediterranean winter dryness during the last few decades of the 20 th century is
unique in the whole 500 year precipitation time series.
The reconstruction quality is generally better for temperature than for
precipitation. Seasonal reconstructions are less trustworthy than the monthly
estimates, but they still show some skill over various Mediterranean areas allowing
qualitatively and quantitatively statements about the temperature and precipitation
course over the last few centuries.
The patterns of the coldest (1891), the warmest (1772), the driest (1609) and the
wettest (1684) Mediterranean winters over the last few centuries have been shown
together with the departures from the 20 th long-term mean and the quality of the
reconstructions. Clear temperature and precipitation anomalies have been found
with inverse sign for Southern Europe/Northern Africa and Northern Europe.
Although only a few predictors were available for the larger Mediterranean for the
driest and wettest winters, there is potential for skillful reconstructions for some
Mediterranean regions.
The reconstructed sea level pressure fields (Luterbacher et al. 2001a) related to
these extremes clearly showed the expected anomalies with cold/warm, and
moist/dry air advection towards the Mediterranean.
We found that the relationship between the reconstructed winter NAO and the
mean winter Mediterranean precipitation remained stable over the last 5 centuries.
In periods with stronger westerlies over the eastern North Atlantic, the
Mediterranean area, as a whole, obtained less precipitation and vice versa.
The North Atlantic Oscillation it is not the most important signal of atmospheric
variability concerning temperature over the averaged larger Mediterranean region.
This mainly reflects the fact that the averaged mean temperature value includes both
areas with a strong link to the NAO and others with no connection at all. Their
competitive effects may be the reason for the low correlations over the last 500 years.
Preliminary results dividing the Mediterranean into sub-regions indicate close
relation between the winter NAO and winter temperature over the Eastern
Mediterranean whereas for the Western and Central parts this is not the case.
The presented continuous temperature and precipitation reconstructions over
Europe, North Africa and the Near East are of great importance for all kind of
climatic analyses dealing with natural variabilities on time scales from months to
decades or centuries. They can be used to study the low and high frequency
variability and the characteristics and extremes of climate and the relationship with
the atmospheric circulation and can be compared with model-produced
reconstructions of natural and forced (external and internal) variability for the last
centuries.
Further investigations will include new predictor data in order to improve the
reconstruction skill, to provide time-dependent uncertainty ranges of our estimates,
to study the influence of forcing factors, and to study the low frequency climate
variability for different climate sensitive regions along the Mediterranean for all the
four seasons over the last centuries and the comparison with reconstructions from