Heat waves and dry spells are analyzed (i) at eight stations in south Morav
ia (Czech Republic), (ii) in the control ECHAM3 GCM run at the gridpoint cl
osest to the study area, and (iii) in the ECHAM3 GCM run for doubled CO2 co
ncentrations (scenario A) at the same gridpoint (heat waves only). The GCM
outputs are validated both against individual station data and areally repr
esentative values. In the control run, the heat waves are too long, appear
later in the year, peak at higher temperatures and their numbers are under-
(over-) estimated in June and July (in August). The simulated dry spells a
re too long, and the annual cycle of their occurrence is distorted. Mid-tro
pospheric circulation, and heat waves and dry spells are linked much less t
ightly in the control climate than in the observed. Since mid-tropospheric
circulation is simulated fairly successfully, we suggest the hypothesis tha
t either the air-mass transformation and local processes are too strong in
the model or the simulated advection is too weak. In the scenario A climate
, the heat waves become a common phenomenon: warming of 4.5 degrees C in su
mmer (difference between scenario A and control climates) induces a five-fo
ld increase in the frequency of tropical days and an immense enhancement of
extremity of heat waves. The results of the study underline the need for (
i) a proper validation of the GCM output before a climate impact study is c
onducted and (ii) translation of large-scale information from GCMs into loc
al scales using downscaling and stochastic modelling techniques in order to
reduce GCMs' biases.