ORIGINAL PAPER
Variability of droughts in the Czech Republic, 1881–2006
R. Brázdil & M. Trnka & P. Dobrovolný & K. Chromá &
P. Hlavinka & Z. Žalud
Received: 26 November 2007 /Accepted: 24 May 2008 / Published online: 11 November 2008
# Springer-Verlag 2008
Abstract We analyze droughts in the Czech Republic from
1881–2006 based on the Palmer drought severity index
(PDSI) and the Z-index using averaged national temperature
and precipitation series for the calculations. The
standardized precipitation index (SPI), PDSI and Z-index
series show an increasing tendency towards longer and
more intensive dry episodes in which, for example,
droughts that occurred in the mid-1930s, late 1940s–early
1950s, late 1980s–early 1990s and early 2000s were the
most severe. Cycles at periods of 3.4–3.5, 4.2–4.3, 5.0–5.1
and 15.4 years exceeded 95% confidence levels in
application of maximum entropy spectral analysis. These
are expressed at different intensities throughout the period
studied. The occurrence of extremely dry and severely dry
months is associated with a higher frequency of anticyclonic
situations according to the classification employed
by the Czech Hydrometeorological Institute. Principal
component analysis documents the importance of the ridge
from the Siberian High over Central Europe when extreme
and severe droughts in months of the winter half-year are
considered in terms of sea-level pressure. In the summer
half-year, the ridge of the Azores High over Central Europe
is the most important. Drought episodes have a profound
effect on national and regional agricultural production, with
yields being consistently lower than in normal years, as is
documented through the example of spring barley, winter
wheat, forage crops on arable land, and hay from meadows.
Seasons with pronounced drought during the April–June
period (e.g., 1947 and 2000) show the most significant
yield decreases. Forests appear to be very vulnerable to
long-term drought episodes, as it was the case during the
dry years of 1992–1994. This study clearly confirms the
statistically significant tendency to more intensive dry
episodes in the region, driven by temperature increase and
precipitation decrease, which has already been suggested in
other studies.
1 Introduction
Droughts may be considered, apart from floods, as the most
disastrous natural events in the Czech Republic (see e.g.,
Brázdil et al. 2005, 2007b). Shortage of water during
drought episodes has important consequences for agriculture,
forestry, water management, and other human activities,
as well as for remaining semi-natural ecosystems.
The term ‘drought’ expresses a negative deviation of water
balance from the climatological normal over a given area.
Deficiency of precipitation over an extended period of time
(several weeks to several years) is the primary cause of
drought in Central Europe, while other meteorological
elements (such as temperature, wind, and humidity) frequently
intensify its impacts. Any realistic definition of drought
must be specific with respect to region and application. Four
interrelated categories of drought – meteorological, agricultural,
hydrological and socio-economic – are usually distinguished,
based on time scales and impacts (e.g., Heim 2002).
Agricultural drought impacts are mainly associated with time
scales in terms of weeks to 6–9 months, while hydrological
and socioeconomic impacts usually became apparent after
longer time intervals. Individual drought categories and their
Theor Appl Climatol (2009) 97:297–315
DOI 10.1007/s00704-008-0065-x
R. Brázdil (*) :P. Dobrovolný :K. Chromá
Institute of Geography, Masaryk University,
Brno, Czech Republic
e-mail: brazdil@geogr.muni.cz
M. Trnka :P. Hlavinka :Z. Žalud
Institute of Agrosystems and Bioclimatology,
Mendel University of Agriculture and Forestry,
Brno, Czech Republic
time-scales obviously overlap. The occurrence of meteorological
drought, however, precedes the onset of specific
impacts, so it is extremely important to understand the
regional characteristics of meteorological drought before
studying the specific impacts of this phenomenon.
Steadily increasing awareness of the effects of drought has
led to a number of regional and large-scale studies (see e.g.,
Dai et al. 1998, 2004; Szinell et al. 1998; Rebetez 1999;
Wilhite 2000; Domonkos et al. 2001; Heim 2002; LloydHughes
and Saunders 2002; Panu and Sharma 2002;
Svoboda et al. 2002; Pongrácz et al. 2003; Stefan et al.
2004; Sönmez et al. 2005; van der Schrier et al. 2006, 2007;
Blenkinsop and Fowler 2007; Livada and Assimakopoulos
2007; Marsh et al. 2007). The temporal and spatial aspects of
drought climatology have also been the subjects of several
papers that include the territory of the Czech Republic (see
e.g., Možný 2004; Blinka 2005; Dufková and Šťastná 2005;
Brázdil et al. 2007b; Dubrovský et al. 2007b; Trnka et al.
2007a). However, a number of problems remain to be
addressed.
The current paper concentrates on certain new climatological
aspects and on the statistical analysis of the longterm
temporal variability of drought over the territory of
the Czech Republic in the context of observed climate
fluctuations in the period 1881–2006. An overview of
information on drought indices and the available data is
given, and a general description of climate fluctuations in
the Czech Republic follows. Temporal drought patterns
are studied, using various indices, in terms of trends and
fluctuations, cyclicity and synoptic causes. The results
obtained are discussed in broader climatological and
European contexts. Finally, several examples of drought
impact are presented.
2 Drought indices and data
Several indices may be used for the analysis of drought (for
an overview, see e.g., Heim 2002). The Palmer drought
severity index (PDSI) and the Z-index (Palmer 1965) were
employed in this study. They are among the most
comprehensive and widely applied methods for the quantification
of droughts all over the world (e.g., Szinell et al.
1998; Lloyd-Hughes and Saunders 2002; Ntale and Gan
2003; Dai et al. 2004; van der Schrier et al. 2007). A
comprehensive overview of the calculation procedures
required to derive the PDSI and Z-index may be found,
for example, in Palmer (1965, 1968), Alley (1984) or van
der Schrier et al. (2006, 2007). Both Z-index and PDSI can
be used to describe drought climatology (e.g., Tolasz et al.
2007; Trnka et al. 2007a) and the Z-index is also used as a
good indicator of agricultural drought (e.g., Quiring and
Papakryiakou 2003; Trnka et al. 2007b).
In general, PDSI is based on a supply-and-demand
approach to the water-balance equation. It thus incorporates
antecedent precipitation, moisture supply, and demand at
the surface, as calculated by the Thornthwaite (1948)
method. The PDSI applies a two-layer bucket-type model
for soil moisture computations and makes three assumptions
relating to soil profile characteristics: (1) the waterholding
capacity of the surface layer is set at a maximum of
25 mm; (2) the water-holding capacity of the underlying
layer has a maximum value dependent on the soil type; and
(3) water transfer to or from the lower layer occurs only
when the surface layer is full or empty. The PDSI itself may
be described as an accumulative departure relative to local
mean patterns of atmospheric moisture supply and demand
at the surface (Palmer 1965, 1968), and it is thought to
represent episodes of prolonged drought as well.
PDSI calculation also includes an intermediate term
known as the Palmer moisture anomaly index (Z-index),
which is a measure of surface moisture anomaly for a given
week/month, without consideration of the antecedent conditions
characteristic of PDSI. The Z-index may therefore
be employed to track short-term deviations of soil moisture
from the normal range and is frequently used to estimate,
for example, agricultural droughts, since the Z-index
responds relatively quickly to changes in soil moisture
(Karl 1986; Quiring and Papakryiakou 2003; Trnka et al.
2007b). Because of its capacity to rank the dryness or
wetness of individual months, the Z-index has been used as
one of the indicators for short-term drought spells. Dai et al.
(2004) have demonstrated that the monthly Z-index and
PDSI correlate rather well, during the warm season, with
observed soil moisture departures across a number of
regions on the northern hemisphere.
In this study, self-calibrated versions of the PDSI and
Z-index (scPDSI and scZ-index, respectively) were used
(Wells et al. 2004). The scPDSI improves upon the
‘traditional’ PDSI. Lacking computational resources, Palmer
calculated a set of empirical weighting factors used in the
PDSI algorithm by averaging the values from only a few
locations, mainly from the American Midwest. These
averaged values for the weighting factors have become a
fixed part of the PDSI computation, regardless of the
climate in which it is used. Wells et al. (2004) improve the
performance of the PDSI by automating the calculations
leading to the weighting factors. These weighting factors
are determined for each location, and this makes them
uniquely appropriate to only that location. A detailed
description on how scPDSI is computed may be found in
Wells et al. (2004) and, more concisely, in van der Schrier
et al. (2006). As the applied versions of scPDSI and scZindex
assumed all the precipitation fell in the form of the rain,
the index values during months with a significant proportion
of snow precipitation (i.e., December to February) might not
298 R. Brázdil et al.
reflect actual changes in soil moisture. However, most of the
subsequent analysis was focused on the vegetation season and
we believe that the resulting overall effect is relatively small.
This claim is supported by the results of van der Schrier et al.
(2007) who found relatively a small effect on the overall
drying trend when a modification of scPDSI, accounting
for snow formation and melting, was used in a region
where snow precipitation is much more important than in
the Czech Republic.
As well as the scPDSI and scZ-index, the “relative”
version of both indices (Dubrovský et al. 2007b) was used
in some parts of the study. The relative indices differ from
the self-calibrated indices by using two different weather
series in a two-step process. In the first step, the model of
the drought index is calibrated using the reference weather
series, which relates to a certain reference station in order to
allow for between-station comparisons. Having calibrated
the model, it is then applied to the second series, hereafter
called the tested series. The tested series might relate either
to a different station (to compare the drought conditions in
that station with respect to the reference station) and/or to a
different period (to compare drought conditions in that
period with respect to the reference period). In analyzing
drought impact on crop production in various regions of the
Czech Republic, we used the reference series created by
aggregating data from a set of 233 climatological stations in
the Czech Republic which had been compiled for the study
of Tolasz et al. (2007). In this case, the reference series
represents a wider spectrum of precipitation-temperature
situations, which makes the model applicable to a wider
spectrum of climatic conditions and allows comparison of
drought intensities in different parts of the country. From
now on, we shall denote the two relative drought indices as
rZ-index and rPDSI.
As drought episodes in Central Europe are closely
associated with precipitation deficits, the standardized
precipitation index (SPI) was also calculated alongside the
scPDSI and scZ-index series. The SPI is in fact a
transformation of a given precipitation total aggregated
over a selected period (commonly 1 to 24 months, where
the shorter time scales may represent agricultural drought
and the longer time scales relate better to hydrological
drought) into a standardized normal distribution (http://
drought.mssl.ucl.ac.uk/spi.html). In this study we used a
12-month SPI significantly correlated (r=0.78) with the
scPDSI series and not influenced by annual distribution of
precipitation.
The analysis of droughts in the Czech Republic for the
period 1881–2006 is based on following mean monthly series:
a) monthly temperature series for the Czech Republic
calculated by the method of arithmetical means from
data provided by 174 homogenized meteorological
series consisting of at least 30 years of observations
in the period 1848–2000 (Štěpánek 2004) and extended
to 2006 by the same author; temperature series used for
calculation were homogenized using the standard
normal homogeneity test (SNHT) by Alexandersson
(1986)
b) monthly precipitation series for Bohemia (the western
part of the Czech Republic) was originally created by
the method of isohyets for the period 1876–1956 (Jílek
1957) and extended later by calculations done by the
Czech Hydrometeorological Institute
c) monthly precipitation series for Moravia and Silesia
(the eastern part of the Czech Republic), calculated by
the method of weighted averaging from data provided
by 143 rain-gauge stations for the period 1881–1980
(Brázdil et al. 1985), then extended to 1988 and
onwards and completed by calculations done by the
Czech Hydrometeorological Institute
d) from precipitation series for Bohemia and Moravia and
Silesia (testing of the relative homogeneity of two
series using the SNHT by Alexandersson 1986, showed
no inhomogeneities) a new mean precipitation series
for the whole Czech Republic was calculated by the
method of weighted means, where the areas of both
parts of the Czech Republic were taken into consideration
as weights (66.9 and 33.1%, respectively).
Both mean temperature and precipitation series for the
whole Czech Republic were used for further analysis and
calculations of drought indices, i.e., 12-month SPI, scZindex,
and scPDSI.
In addition to meteorological data, the crop yields of
several widely grown crops in the Czech Republic were
used to demonstrate large-scale drought impacts. In the case
of spring barley and winter wheat, the yield series for the
period 1961–2000 were available for individual districts
(i.e., administrative units with areas of approximately
1,000 km2
). The mean national yields of spring barley,
winter wheat, forage crops on arable land and hay from
permanent grasslands were available for the period 1918–
2006 (with the exception of the Second World War period).
The district yields were detrended using the average annual
yield against year-of-harvest for each crop district. The two
driest and two wettest seasons were always excluded from
trend calculations. The detrending procedure resulted in
values of non-standardized residuals (hereafter referred to
as yield deviation), calculated for each district, and used in
the subsequent analyses. In the 1918–2006 period, the yield
deviations were calculated as a difference between the
seasonal yield and 5-year running averages. Detrending was
essential to avoid inclusion of intensification and technological
advances in agricultural production during the 20th
century. The same approach was used (e.g., by Quiring and
Variability of droughts in the Czech Republic, 1881–2006 299
Papakryiakou 2003) and tested for the Czech Republic by
Trnka et al. (2007b).
Meteorological data were complemented by values of
the maximum soil water holding capacity (MSWHC) in the
rooting zone (up to 1.5 m deep). This parameter, at a
resolution of 1 km2
, was estimated using a combination of
digitalized maps of soil types (Tomášek 2000) and detailed
soil physics data from 1,073 soil pits collected by the Czech
National Soil Survey. For each of the 25 soil types, a mean
value of MSWHC was determined as an average of the
maximum water-holding capacities of all soil pits of a
particular soil type in the database. The MSWHC of the
individual soil types ranges from 137 to 302 mm and was
determined at each soil pit by the weighted soil-waterholding
capacities of individual soil horizons, up to the
maximum rooting depth. For the calculation of scPDSI and
scZ-index time series (1881–2006), a spatial average of
MSWHC representing whole territory of the Czech Republic
was used (i.e., 182 mm).
3 Climate fluctuations in the Czech Republic
Climate fluctuations in the Czech Republic have been
evaluated in several papers taking into consideration either
meteorological data since 1961 (e.g., Brázdil et al. 2001,
2007a; Huth and Pokorná 2004, 2005; Moliba et al. 2006;
Chládová et al. 2007) or long-term series (see e.g., Pišoft et
al. 2004; Štěpánek 2004; Brázdil et al. 2007b).
Series of annual air temperatures in the Czech Republic
show a clear rising trend over the period 1881–2006, with a
statistically significant linear trend of 0.082°C/10 years
(Fig. 1a). This is consistent with the global temperature rise,
which has reached a value of 0.74°C over the past 100 years
(1906–2005) (IPCC 2007). Annual precipitation totals
exhibit a statistically insignificant negative trend of
−2.34 mm/10 years (Fig. 1b).
These observed fluctuations in temperature and precipitation,
as well as the process of global warming, raise the
question as to what extent climate continentality has
changed in the Czech Republic. Two indices have been
chosen for further processing from among the several
available:
(a) The index of temperature continentality KT, after
Gorczyński (1920) is calculated from the formula
KT ¼ 1:7=sinϕð Þ A À 12 sin ϕð Þ
in which A is the annual temperature range (mean monthly
maximum minus mean monthly minimum) and ϕ is the
latitude of the given station.
(b) The index of precipitation seasonality, after Markham
(1970), which is calculated as the vector sum of the
individual monthly totals with the size of the resulting
vector then divided by the annual precipitation sum and
expressed as a percentage (a lower value of the index marks
a more balanced annual variation of precipitation, and vice
versa – for more details see Brázdil 1978).
In terms of temperature continentality in the Czech
Republic (Fig. 1c), there are no significant changes over the
period 1881–2006, despite year-by-year fluctuations (maximum
value in 1929 with an extremely cold February,
minimum in 1974 with very mild winter months). On the
other hand, the index of precipitation seasonality shows a
clear negative linear trend but its value, at −0.4% per
10 years, is not statistically significant for the significance
level of α = 0.05 (Fig. 1d). This indicates a long-term
tendency towards a more balanced distribution of precipitation
over the course of the year.
4 Long-term variability of droughts in the Czech Republic
4.1 Fluctuations of drought in the Czech Republic
Long-term fluctuations of 12-month SPI, scZ-index, and
scPDSI smoothed by 5-year Gauss filter for the territory of
the Czech Republic in the period 1881–2006 are illustrated
in Fig. 2. All three drought indices show a clear decreasing
trend in the period studied, i.e., an increasing tendency to
longer and more intensive dry episodes. Values of statistically
significant decreasing trends reach −0.024/10 years
for 12-month SPI, −0.048/10 years for the scZ-index and
−0.192/10 years for scPDSI. Although several extremely
dry and severely dry patterns emerged, the droughts in the
mid-1930s, late 1940s–early 1950s, late 1980s–early 1990s
and early 2000s were the best expressed. Turning to the
scPDSI values, the episode in the early 1990s was the most
severe. The drought from the late 1940s to the early 1950s
is comparable with the previous episode only from the
point of view of its duration. A prevailing trend towards
more intensive and prolonged drought in Central Europe
has been confirmed in a study by Trnka et al. (2008), based
on data from selected stations. As the SPI allows tracking
of drought tendencies driven only through precipitation
changes we consider the scZ-index and scPDSI (that also
take into account changes in the air temperature) as more
suitable tools for evaluating long-term drought trends.
4.2 Cyclicity of droughts
Analysis of drought cyclicity in the Czech Republic based
on scZ-index and scPDSI series was performed by the
method of maximum entropy spectral analysis (MESA, see
e.g., Olberg 1982; Olberg and Rákóczi 1984). ME-spectra
for spring (MAM), summer (JJA), autumn (SON) and
300 R. Brázdil et al.
vegetation period (April–September) are shown in Fig. 3.
Although the spectral density of different cycles exceeded
the 95% confidence level, they concentrate mainly at
lengths of 3.4–3.5, 4.2–4.3, 5.0–5.1 and 15.4 years. A
causal explanation of the cycles revealed is beyond the
scope of this paper, although it may be noted that cycles
longer than 10 years are usually associated with fluctuations
in solar activity (see e.g., Benestad 2003) and shorter
b
400
500
600
700
800
900
1000
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
c
0
10
20
30
40
50
60
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
d
0
10
20
30
40
50
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
a
4
5
6
7
8
9
10
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
Index-Gorczynski(%)Precipitation(mm)Temperature(ºC)Index-Markham(%)´
Fig. 1 Climate fluctuations in the Czech Republic in the period 1881–2006 (with linear trends and smoothed by Gauss filter over 10 years):
a Annual air temperature (°C). b Annual precipitation totals (mm). c Index of temperature continentality after Gorczyński (%). d Index of
precipitation seasonality after Markham (%)
Variability of droughts in the Czech Republic, 1881–2006 301
cycles, in the range of 3–7 years, may follow from the
internal variability of the atmosphere-ocean system
reflected, for example, in the El Niño–Southern Oscillation
(ENSO) phenomenon (for Central Europe, see e.g., Brázdil
and Bíl 1998).
The method of dynamic MESA (Olberg 1982) may also
be used for the study of the temporal stability of the cycles
revealed. Figure 4 documents basic differences between the
two drought indices used. Although some time intervals
with significant short cycles (e.g., between 3 and 3.5 years)
appear in the first and last thirds of the period studied for
the scZ-index, longer cycles are typical of the scPDSI.
4.3 Synoptic climatology of droughts
The synoptic causes of droughts in the Czech Republic
were studied with respect to severely dry and extremely dry
months selected in the light of the scZ-index (Table 1). The
first approach takes into consideration the frequency of
occurrence of synoptic situations after the classification
created by Brádka et al. (1961) and continuously extended
by a team from the Czech Hydrometeorological Institute
(Katalog 1967), but only for days after 1946. The basic idea
was to classify synoptic processes according to the
direction of airflow and the character of circulation
-6
-4
-2
0
2
4
6
8
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
scPDSI
-6
-4
-2
0
2
4
6
8
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
scZ-index
-3
-2
-1
0
1
2
3
1881 1891 1901 1911 1921 1931 1941 1951 1961 1971 1981 1991 2001
SPI
Fig. 2 Fluctuations in 12-month SPI, scZ-index, and scPDSI over the territory of the Czech Republic during the period 1881–2006, including
corresponding linear trends and smoothing of the series by 5-year Gauss filter
302 R. Brázdil et al.
(cyclonic or anticyclonic) during the natural synoptic
period. Types of synoptic situations were derived from the
basic features of pressure systems, the frontal zones, and
frontal systems from the greater set of similar situations.
The catalog lists 28 types of synoptic situation (for their list
and schemes, see e.g., Brázdil et al. 2007b; Řezníčková et
al. 2007). This classification was used rather than the
better-known classification by Hess and Brezowsky developed
for Germany (e.g., Hess and Brezowsky 1977;
Gerstengarbe and Werner 1993) which is territorially
shifted for the territory of the Czech Republic.
Differences between relative frequencies of situations
occurring in selected extremely dry (severely dry) months
and the reference period 1961–1990 (Fig. 5) show a generally
expected decrease in frequencies of wetter cyclonic situations
and an increase in drier anticyclonic ones. The most
spectacular is expressed for negative departures in frequencies
of situation Wc (western cyclonic – only winter halfyear),
A (anticyclone over Central Europe – only winter halfyear)
and B (trough over Central Europe) on the one hand
and positive departures of the situation SEa (south-eastern
anticyclonic), Ea (eastern anticyclonic), Sa (southern anticyclonic)
and Wa (western anticyclonic) in the winter half-year
on the other. These differences are higher for extremely dry
months than for severely dry ones.
The second approach towards characterizing macro-scale
circulation patterns during drought episodes is based on
analysis of sea-level pressure (SLP) conditions using principal
component analysis (PCA – see e.g., Barry and Carleton
2001). Mean monthly SLP fields provided by the Met. Office
Hadley Centre (Allan and Ansell 2006) were used for the
analysis. The spatial resolution of the HadSLP2 database is 5
50.0
12.5
5.1
4.3 3.4
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
trend
15.4
10.0
5.0
4.2 2.0
3.4
scPDSIME-spectrum
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
SON
trend
7.1
5.3
4.2
3.4
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
trend
16.7
7.4
5.0
3.13.4
scPDSI
ME-spectrum
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
2.0
14.3
6.3
5.3
3.3
2.4
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
trend
15.4
8.7
5.0
3.4
scPDSI
ME-spectrum
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
2.2
4.3
6.7
Apr−Sep
trend
14.3
6.1
3.4
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
trend
15.4
9.1
5.0
4.3
3.5
scPDSI
ME-spectrum
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
6.9
95%
95%
95%
95%
95%
95%
95% 95%
MAM
JJA
0.5
0.5
0.5
0.5
Fig. 3 Normalized ME-spectra
of scZ-index and scPDSI
(MAM, JJA, SON, April–
September) for the Czech Republic
in the period 1881–2006.
The numbers express cycles (in
years) which exceed the 95%
confidence level. Peaks in MEspectra
for the frequency of 0.0
indicate the linear trend in series
Variability of droughts in the Czech Republic, 1881–2006 303
degrees of latitude-longitude and the data covers the period
from 1850 to 2004. Using 250 grid-points extending from
70°W, 70°N to 50°E, 25°N, mean SLP fields for the reference
period 1961–1990 were calculated (Figs. 6a, 7a). Further
mean SLP fields for extremely and severely dry months were
constructed. Circulation patterns typical of dry periods
selected by means of scZ-index values were highlighted as
departures from normal conditions (Figs. 6b, 7b). Because of
significant differences between summer and winter circulation
in the Atlantic-European area, drought patterns were
studied separately for the winter half-year (October–March)
and the summer half-year (April–September). Further, typical
dry period circulation patterns were characterized by interpreted
component scores, calculated through PCA.
In extremely dry months of the winter half-year (six
cases), there is an important pressure increase in Eastern
Europe, i.e., an intensification of the ridge of the Siberian
High in Central Europe with a flow of dry continental air
(Fig. 6). The Icelandic Low therefore deepens. The first
principal component (PC) explains 92.4% of the total
variance, i.e., the character of the SLP field is very uniform
in all the months studied. This confirms the importance of
the influence of high pressure in Central Europe associated
with the ridge of the Siberian High.
In extremely dry months of the summer half-year (22
cases), absolute changes in the SLP field are not as strongly
expressed as in the previous case. An increase in SLP for
Western and Central Europe is reflected in a more
pronounced Azores High and its ridge to Central Europe,
also expressed by the first PC explaining 89.1% of the total
variance (Fig. 7).
Severely dry months tend to reflect similar situations,
although the first PC explains a smaller share of
variance − 82.6% for the months of the winter half-year
(36 cases) and 78.7% for months of the summer half-year
(40 cases). However, in these cases, the share of other
PCs is also very small. For example, the second PC
explains only 6.6 and 6.1% of total variance, the third
PC 3.6 and 5.7%, etc. As in extremely dry events, drought
in the months of the winter half-year is associated with
the ridge of the Siberian High (Fig. 8a) and in the summer
half-year with the ridge of the Azores High (Fig. 8b).
5 Discussion and conclusions
The results of drought analysis in the territory of the Czech
Republic during the period 1881–2006 show a clear
tendency towards prolongation and greater severity of
drought episodes. This is mainly due to the process of
global warming (IPCC 2007), also reflected in an air
temperature rise in the Czech Republic; changes in
precipitation totals do not appear so crucial, although an
insignificant decrease also appears here. The coherency of
drought indices series with temperatures (Fig. 9) in the
Fig. 4 Dynamic ME-spectra of scZ-index and scPDSI (April–
September) for the Czech Republic in the period 1881–2006 (length
of the basic period 30 years, step 1 year). Only the cycles exceeding
the 95% confidence level are indicated
Table 1 Category of droughts according to the Palmer scZ-index, drought severity index (scPDSI) and SPI
scZ-index scPDSI SPI Drought category
≥3.50 ≥4.00 ≥2.00 Extremely wet
3.49 to 2.50 3.99 to 3.00 1.99 to 1.50 Severely wet
2.49 to 1.00 2.99 to 2.00 1.49 to 1.00 Wet
0.99 to −1.24 1.99 to −1.99 0.99 to −0.99 Normal
−1.25 to −1.99 −2.00 to −2.99 −1.00 to −1.49 Dry
−2.00 to −2.74 −3.00 to −3.99 −1.50 to −1.99 Severely dry
≤−2.75 ≤−4.00 ≤−2.00 Extremely dry
304 R. Brázdil et al.
vegetation period (April–September) exceeds 95 and 99%
confidence levels only for some cycles (trend, 12.5, 3.3,
and 3.1, 2.6, and 2.5 years), while coherency coefficients
with precipitation totals exceed both confidence levels for
all cycles in the case of scZ-index and for cycles longer
than 5 years for scPDSI (for calculation of coherency
coefficients see Schönwiese 1985).
The increasing air temperature leads to higher potential
evapotranspiration estimates in the water-balance model of
both scZ-index and scPDSI, which in combination with an
evident (although not significant) decrease of precipitation
leads to a higher probability of soil moisture deficit. This is
the primary cause of the high coherency between the
statistically significant trend for air temperature and trends
in both drought indices. The results for shorter cycles,
however, indicate that precipitation is the principal meteorological
element driving soil moisture dynamics in Central
Europe. This is in line with the commonly held view that
precipitation deficit is the primary cause of individual
drought episodes in the region, with the temperature seen as
a moderating or intensifying factor.
The frequency of anticyclonic situations, together with
the temperature and precipitation patterns of the two most
important drought episodes in the Czech Republic selected
according to scPDSI (April 1947–June 1954 and April
1988–December 1994) were studied with respect to the
reference period 1961–1990 (Fig. 10). The first drought
episode shows a clear prevalence of above-normal anticyclonic
situations in many months, more very dry months
(18 months with precipitation totals lower than 50% of the
corresponding mean) and relatively smaller positive temperature
anomalies (but in 2 months higher than 4.0°C).
During this period, short sequences of wet months also
appeared, mainly from November 1947 to February 1948
and from September to November 1952, but their occurrence
in the months of the winter half-year did not
significantly affect the dryness of the whole episode,
largely due to comparatively lower precipitation totals
-15
-10
-5
0
5
10
15
20
Wc
Wcs
Vfz
NWc
Nc
NEc
Ec
SEc
SWc1
SWc2
SWc3
B
Bp
C
Cv
Wa
Wal
NWa
NEa
Ea
SEa
Sa
SWa
A
Ap1
Ap2
Ap3
Ap4
1 2
Cyclonic situations Anticyclonic situations
-15
-10
-5
0
5
10
15
20
Wc
Wcs
Vfz
NWc
Nc
NEc
Ec
SEc
SWc1
SWc2
SWc3
B
Bp
C
Cv
Wa
Wal
NWa
NEa
Ea
SEa
Sa
SWa
A
Ap1
Ap2
Ap3
Ap4
1 2
Cyclonic situations Anticyclonic situations
Synoptic situations
Differences(%)
Extremely dry months
Severely dry months
Fig. 5 Differences in relative frequencies (%) of the occurrence of synoptic situations (Katalog 1967) in extremely dry and severely dry months
(see Table 1 for definition) selected in the Czech Republic according to scZ-index (1946–2006) with respect to normal patterns (1961–1990):
1 Winter half year (October–March), 2 Summer half-year (April–September)
Variability of droughts in the Czech Republic, 1881–2006 305
during the winter half-year. In the second drought episode,
in the late 1980s–early 1990s, the frequency of anticyclonic
situations was close to normal, which is surprising given
the duration and intensity of drought during this period.
The dryness of this episode was related to the frequent
occurrence of dry months (wet months with totals ≥150%
of the corresponding mean appeared almost exclusively
during winter half-year) and higher temperatures, often with
departures of more than 2.0°C.
The existence of drying trends in the Czech Republic
corroborates the findings of Szinell et al. (1998), who
reported regionally specific, but nevertheless significant,
trends in scPDSI values towards a drier climate at 15
stations in Hungary, and of Horváth (2002), who reported
similar results for the River Tisza catchment. Dai et al.
(1998) noted that the area under the influence of drought
increased in Europe between 1960 and 1995; however, this
rise was found to be of the same magnitude that Europe had
already experienced in the early 1920s and late 1940s. In
subsequent work applying the scPDSI to the 1870–2002
time series, Dai et al. (2004) described a notable drying
trend throughout Central Europe (including the Czech
Republic), beginning at the start of the 20th century, which
was linked to increasing temperatures within the same time
frame. Similarly, van der Schrier et al. (2006) noted quite
strong drying trends in Central Europe based on scPDSI for
the summer months (between 1950 and 2002) in the area
located east of 15°E. On the other hand, studies by Frich et
al. (2002), Moberg and Jones (2005) and Schmidli and Frei
(2005) did not indicate significant drying trends in this
region. This is understandable, since in these studies the
degree of dryness was evaluated in terms of consecutive
Fig. 6 Winter half-year: a Mean
SLP in the Atlantic-European
area in the reference period
1961–1990. b Deviations (hPa)
of composite mean SLP of extremely
dry months (see Table 1
for definition) from mean SLP
in the reference period 1961–
1990. c Interpreted component
scores of the first PC for SLP in
extremely dry months; H High,
L Low
306 R. Brázdil et al.
days without precipitation rather than in terms of soil
moisture anomaly. Using the number of dry days takes only
one component of landscape water balance (frequency of
rainfall) into account but completely omits changes on the
demand side, as well as changes in precipitation totals.
Most of the studies based on observed data in the Central
European region show a statistically significant increase in
temperatures during the 20th century, while a tendency
towards higher precipitation was found only during the
winter. The changes that occurred in the precipitation
parameters evaluated during the summer half-year were
mainly insignificant (Frich et al. 2002; Moberg and Jones
2005; Schmidli and Frei 2005; Hundecha and Bárdossy
2005; Chládová et al. 2007). In addition, analysis of the
annual cycle of trends of selected climatic elements in the
Czech Republic shows a rather pronounced tendency
towards higher temperatures and lower precipitation during
most of the April–June period (Moliba et al. 2006).
According to the projections of global circulation models
for the Czech Republic (e.g., Dubrovský et al. 2005), an
increase in potential evapotranspiration (driven by the
temperature increase) will not be met by an adequate
increase in precipitation, which will inevitably lead to a
higher frequency of droughts, as the studies of Dubrovský
et al. (2007a, b) have indicated.
As has been shown in various papers (e.g., Quiring and
Papakryiakou 2003; Scian 2004; Trnka et al. 2007b), the
impacts of meteorological drought are very important to
agriculture in much of temperate zone, including Central
Europe. Agriculture in this region is most vulnerable to
prolonged droughts during early spring, since it reduces
survival and stand density leading, in some cases, to
Fig. 7 Summer half-year: a
Mean SLP in the AtlanticEuropean
area in the reference
period 1961–1990. b Deviations
(hPa) of composite mean SLP of
extremely dry months (see Table
1 for definition) from mean SLP
in the reference period 1961–
1990. c Interpreted component
scores of the first PC for SLP in
extremely dry months; H High,
L Low
Variability of droughts in the Czech Republic, 1881–2006 307
complete crop failure (as was the case in the droughts of
1947 and 2000). Drought during anthesis and/or grain
filling periods adversely affects grain numbers and mass.
This occurred in some regions in 1947, when the
combination of a long winter, which led to late sowing,
and severe drought, peaking in May and then September,
were the main causes of poor emergence of spring crops
and of slow growth with unsatisfactory tillering. The
drought impacts were most severe in the south-east of the
Czech Republic and on sandy soils; they resulted in a
reduction of up to 70% in regional yields of cereals
(especially winter wheat) and forage crops (Minář 1947).
The drought of 2000 lasted from April to June in the
Czech Republic and was the primary reason for a poor
spring crop, particularly in southern Czech Republic
(similar to 1947), where the losses compensated to farmers
Fig. 8 Interpreted component
scores of the first PC for SLP
in the Atlantic-European area in
severely dry months (see Table
1 for definition) of the winter
half-year (a) and the summer
half-year (b): H High, L Low
12.5
trend
3.3
2.6
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
trend
12.5 2.53.1
scPDSI
Coherencycoefficient
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5
Precipitation
scZ-index
Period (years)
Frequency
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
trend 12.5
3.3
scPDSI
22.53456122496
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4
6.3
Coherencycoefficient
95%
99%
95%
99%
95%
99%
95%
99%
0.5
0.5
Air temperatureFig. 9 Coherency coefficients
of scZ-index and scPDSI series
with air temperature and precipitation
series of the Czech Republic
in the vegetation period
(April–September) 1881–2006.
The numbers indicate cycles (in
years) for which the coherency
coefficients exceed the 95% and
99% confidence levels
308 R. Brázdil et al.
from the state budget were as high as around 5 billion
Czech crowns (approximately 185 million Euros).
Figure 11a shows the effect of the intense drought on
spring barley yields in 2000. The mean value of the relative
Z-index (e.g., Dubrovský et al. 2007b; Trnka et al. 2007b)
was calculated for arable land in 62 districts of the Czech
Republic over the period 1961–2000 using a database of
233 meteorological stations (Tolasz et al. 2007). The
districts with the highest drought intensity between April
and June also show the highest negative deviations from the
long-term yield mean. The intensity of drought during
April–June may explain 64.8% of the variability in spring
barley yields at district level. On the other hand, the
drought of 2000 had much lighter impacts on winter wheat
production (Fig. 11b). A far better developed root system
and earlier onset of growth in the spring led to the reduction
of wheat yields being less severe and widespread than that
of spring barley. The drought intensity between April and
June can explain only 25.2% of the variability in winter
wheat yields between individual districts.
The effect of drought on agricultural production may also
be observed on the national level, despite differences between
the individual regions. For example, areas located at higher
elevations with higher precipitation totals tend to be less
a
-60
-40
-20
0
20
40
60
80
1947 1948 1949 1950 1951 1952 1953 1954
Frequency(%)
b
-6
-4
-2
0
2
4
6
1947 1948 1949 1950 1951 1952 1953 1954
Temperature(°C)
c
0
50
100
150
200
250
1947 1948 1949 1950 1951 1952 1953 1954
Precipitation(%)
April 1947 − June 1954
Fig. 10 Characteristics of two main drought episodes in the Czech Republic, selected according to scPDSI: a, d Departures in relative frequencies
of anticyclonic situations (%). b, e Monthly temperature anomalies (°C). c, f Monthly precipitation anomalies (%). Reference period 1961–1990
Variability of droughts in the Czech Republic, 1881–2006 309
vulnerable to drought. There is an obvious tendency to belowaverage
yields at Czech Republic level in almost all dry years
(Fig. 12). This holds true not only for annual crops (both
spring and winter) but also for perennials (grasslands used
for hay production). The results presented in Fig. 12a match
the findings of Arora et al. (1987), Petr (1987), Quiring and
Papakryiakou (2003) and Zimolka (2006), who determined
that spring cereal yield is sensitive to moisture stress during
emergence, shooting, heading and early soft dough stages,
developments that usually take place in the April–June
period. Similar but less pronounced relationships were found
between drought severity during October (wheat emergence)
and April–June (shooting, heading and early soft dough) in
the case of winter wheat (Fig. 12b). On the other hand, the
seasonal water balance (April–July) can explain between 40
and 42% of variability in the national mean yields of forage
crops on arable land and hay from meadows (Fig. 12c–d).
Relationships between drought indices for the growing
season and yields of all four crops are well represented by
a second-order polynomial. This closely approximates the
nature of the crop-yield water relationship (Ash et al. 1992)
as crop yields may be inhibited not only through water stress
but also by low global radiation, below-normal temperatures,
root anoxia, and higher infestation pressure of fungal
diseases, all factors that tend to be associated with unusually
wet seasons. Further tests showed no improvement in model
d
-60
-40
-20
0
20
40
60
80
1988 1989 1990 1991 1992 1993 1994
Frequency(%)
April 1988 − December 1994
e
-6
-4
-2
0
2
4
6
1988 1989 1990 1991 1992 1993 1994
Temperature(°C)
f
0
50
100
150
200
250
1988 1989 1990 1991 1992 1993 1994
Precipitation(%)
Fig. 10 (continued)
310 R. Brázdil et al.
results when higher-order polynomials were used, which is
in accord with the findings of Yamoah et al. (2000) and
Quiring and Papakryiakou (2003). The highest influence of
drought on Czech national crop yields may be found in the
years 1922, 1934, 1947, 1976, 1988, 1992, 1993, 2000, and
2003. The extraordinarily hot and dry summer of 2003 (for
the Czech Republic see Pavlík et al. 2003) had not only
agricultural consequences; between 22,000 and 35,000 heatrelated
deaths were recorded across Europe (e.g., Schär and
Jendritzky 2004; Schönwiese et al. 2004; Trigo et al. 2005;
Fischer et al. 2007). The 2 years with the wettest vegetation
season (1926 and 1965) in the Czech Republic were
associated with significant yield decreases for cereals and
yield stagnation for forage crops and hay production.
Forestry is another sector vulnerable to droughts in the
Czech Republic. For example, forest damage attributable to
drought and air pollution grew from 1.8% in the period 1901–
1950 to 18.1% in the next 30 years (Forst et al. 1985). Salvage
felling of timber arising out of droughts reached 7.2% during
the 1963–1999 period. The impacts of the early-1990s
droughts can be particularly well documented.
Corresponding damage from 1992–1994 was reflected, with
a 1-year delay, in high values of salvage felling arising out of
drought (Fig. 13). In 2001–2005, the percentage share of
drought damage in the whole volume of salvage felling in
the Czech Republic achieved ca. 11%, but in some years (e.
g., 2004, 2005) it was as high as 20% (Brázdil et al. 2007b).
Observed drying trends in Central Europe are confirmed by
model projections. For example, Pal et al. (2004) basing their
work on regional climate model projections, identified
Central Europe as being among the areas particularly
vulnerable to an increase in summer droughts (and also
floods), following the observed drying trend in Western/
Central Europe over the past 25 years. According to the
products of GCM HadCM3 for the SRES scenario A2 (see
Houghton et al. 2001), a rise in mean temperature by 2–4°C
in 2060–2099 may be anticipated in the Central European
region, accompanied by only a small change in annual
precipitation totals (Dubrovský et al. 2005). Annual variation
of precipitation should also change, with a conspicuous
increase in winter (by as much as 25%) and a reduction in
summer. This could result in the desertification of the climate
of Central Europe with inevitable consequences for the
Czech Republic, where precipitation is the main source of
water. One consequence may well be a multiple increase in
the probability of occurrence of intense drought episodes
(including the devastating) and also consequent changes in
the water economy of the landscape, as has been documented
by Dubrovský et al. (2007a, b).
Most of the scenarios employed for climatic change
indicate a pronounced increase in territory endangered by
drought, affecting the most productive agricultural regions.
Such a development would mostly affect southern Moravia,
the driest south-eastern part of the Czech Republic. In terms of
forest economy, this represents a considerable risk of forest
fires in combination with higher summer temperatures,
reduction of reserves of soil water and a deficit in precipitation.
Original and partly changed ecosystems, or the remnants
thereof, will also be endangered because conspicuous
aridation will considerably change the conditions for many
stands. This is one example of the crucial importance of
studies devoted to drought, developing a better scientific
understanding of events and preventive mitigation of the
anticipated impacts of the climate of the future.
Fig. 11 Relationships between relative Z-index (Dubrovský et al.
2007b; see Section 2 for explanation) and yield deviation from longterm
means (1961–2000) at 62 (spring barley) and 65 (winter wheat)
districts of the Czech Republic during the drought year 2000 for
spring barley (a) and winter wheat (b). The dotted line indicates the
slope of regression function. In both cases, the yield departures are
significantly correlated with the regional drought intensity, with a
Pearson correlation coefficient of 0.79 and 0.47, respectively
(significance level α=0.01). N.B: Of a total of 77 Czech Republic
districts 15 (12) were excluded completely as yield data were available
either for only a small portion of the evaluated period or over an
insignificant acreage
Variability of droughts in the Czech Republic, 1881–2006 311
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1963 1968 1973 1978 1983 1988 1993 1998 2003
Volume(mil.m3)
Fig. 13 Volume of salvage felling (million m3
) due to drought in the Czech Republic during the period 1963–2005 (source: Research Institute of
Forestry Economy and Gamekeeping, Jíloviště-Strnady; data after 1990 cover approximately three-quarters of the forest area in the Czech
Republic) (Brázdil et al. 2007b)
NORMAL RANGE
WETDRY
-50
-40
-30
-20
-10
0
10
20
30
40
50
-3.0 -2.0 -1.0 0.0 1.0 2.0 3.0
scZ-index (April-July)
Yielddeviation(%)
2000
1934
1976
2003
1947
1992
NORMAL RANGE
-50
-40
-30
-20
-10
0
10
20
30
40
50
-3.0 -2.0 -1.0 0.0 1.0 2.0 3.0
scZ-index (April-July)
Yielddeviation(%)
2000
1934
19762003
1947
1992
WETDRY
scZ-index (April-June)
NORMAL RANGE
WETDRY
-25
-20
-15
-10
-5
0
5
10
15
20
25
-3.0 -2.0 -1.0 0.0 1.0 2.0 3.0
Yielddeviation(%)
1993
19921976
1988
1950
1922
2003
2000
1947
Yielddeviation(%)
NORMAL RANGE
WETDRY
-25
-15
-5
5
15
25
35
-3.0 -2.0 -1.0 0.0 1.0 2.0 3.0
scZ-index (October and April-June)
2000
1993
1976
1988
1992
2003
1947
1922 1950
a b
c d
Fig. 12 Relationships between scZ-index and yield deviations from a 5-year running average of selected crops of a spring barley, b winter wheat,
c forage crops on arable land and d hay from meadows in different months for the Czech Republic in the period 1918–2006 (with the exception of
1939–1944). In all cases, the yield departures are significantly correlated with drought intensity, with a Pearson correlation coefficient of a 0.55;
b 0.43; c 0.63 and d 0.65 respectively (significance level α=0.01). The dotted line indicates the course of a second-order polynomial function
312 R. Brázdil et al.
Acknowledgements This work was financially supported by project
No. 521/08/1682 of Grant Agency of the Czech Republic. Rudolf
Brázdil, Petr Dobrovolný and Kateřina Chromá were supported by
research plan MŠM0021622412 (INCHEMBIOL), Miroslav Trnka,
Zdeněk Žalud and Petr Hlavinka by research plan MŠM6215648905
“Biological and technological aspects of sustainability of controlled
ecosystems and their adaptability to climate change”; both projects are
financed by the Ministry of Education, Youth and Sports of the Czech
Republic. Petr Štěpánek (Czech Hydrometeorological Institute, Brno)
is acknowledged for providing his mean temperature series for the
Czech Republic, as well as the AnClim software used for part of the
calculations; Jan Honner and Iveta Horáková (Czech Statistical Office,
České Budějovice, and Prague) for compilation and quality control of
national and district crop yields data; and Tony Long (Svinošice) for
English language corrections. Brian Fuchs (National Drought Mitigation
Center, University of Nebraska-Lincoln) and the second anonymous
reviewer are also acknowledged for their useful and constructive
criticism of the manuscript.
References
Alexandersson H (1986) A homogeneity test applied to precipitation
data. J Climatol 6:661–675
Allan RJ, Ansell TJ (2006) A new globally complete monthly
historical mean sea level pressure data set (HadSLP2): 1850–
2004. J Clim 19:5816–5842
Alley WM (1984) The Palmer drought severity index: limitations and
assumptions. J Clim Appl Meteorol 23:1100–1109
Arora VK, Prihar SS, Gajri PR (1987) Synthesis of a simplified water
use simulation model for predicting wheat yields. Water Resour
Res 23:903–910
Ash GHB, Shaykewich CF, Raddatz RL (1992) Moisture risk
assessment for spring wheat on the eastern prairies: a water-use
simulation model. Climatol Bull 26:65–78
Barry RG, Carleton AM (2001) Synoptic and dynamic climatology.
Routledge, London, p 620
Benestad RE (2003) Solar activity and earth’s climate. Springer,
Berlin Heidelberg New York, p 287
Blenkinsop S, Fowler HJ (2007) Changes in European drought
characteristics projected by the PRUDENCE regional climate
models. Int J Climatol 27:1595–1610
Blinka P (2005) Klimatologické hodnocení sucha a suchých období na
území České republiky v letech 1876–2002 (Climatological
evaluation of drought and dry periods on the territory of the
Czech Republic in the years 1876–2002). Meteorol Zpr 58:10–18
Brádka J, Dřevikovský A, Gregor Z, Kolesár J (1961) Počasí na území
Čech a Moravy v typických povětrnostních situacích (Weather on
the territory of Bohemia and Moravia in typical weather
situations). Hydrometeorologický ústav, Praha, p 31
Brázdil R (1978) Stupeň nerovnoměrnosti ročního chodu srážek (The
degree of irregularity of the annual variation of precipitation).
Sbor Českoslov Spol Zeměp 83:91–103
Brázdil R, Bíl M (1998) Jev El Niño–Jižní oscilace a jeho možné
projevy v polích tlaku vzduchu, teploty vzduchu a srážek v
Evropě ve 20. století (El Niño–Southern Oscillation and its
effects on air pressure, air temperature and precipitation in
Europe in the 20th century). Geografie-Sbor ČGS 103:65–87
Brázdil R, Kolář M, Žaloudík J (1985) Prostorové úhrny srážek na
Moravě v období 1881–1980 (Areal precipitation sums in
Moravia in the period 1881–1980). Meteorol Zpr 38:87–93
Brázdil R, Štěpánek P, Květoň V (2001) Temperature series of the
Czech Republic and its relation to Northern Hemisphere temperatures
in the period 1961–1999. In: Brunet India M, López
Bonillo D (eds) Detecting and modelling regional climate
change. Springer, Berlin Heidelberg New York, pp 69–80
Brázdil R, Dobrovolný P, Elleder L, Kakos V, Kotyza O, Květoň V,
Macková J, Müller M, Štekl J, Tolasz R, Valášek H (2005)
Historické a současné povodně v České republice (Historical and
recent floods in the Czech Republic). Masarykova univerzita,
Český hydrometeorologický ústav, Brno Praha, p 370
Brázdil R, Chromá K, Dobrovolný P, Tolasz R (2007a) Climate
fluctuations in the Czech Republic during the period 1961–2005.
Int J Climatol doi:10.1002/joc.1718
Brázdil R, Kirchner K, Březina L, Dobrovolný P, Dubrovský M,
Halásová O, Hostýnek J, Chromá K, Janderková J, Kaláb Z,
Keprtová K, Kotyza O, Krejčí O, Kunc J, Lacina J, Lepka Z,
Létal A, Macková J, Máčka Z, Mulíček O, Roštínský P, Řehánek T,
Seidenglanz D, Semerádová D, Sokol Z, Soukalová E, Štekl J,
Trnka M, Valášek H, Věžník A, Voženílek V, Žalud Z (2007b)
Vybrané přírodní extrémy a jejich dopady na Moravě a ve Slezsku
(Selected natural extremes and their impacts in Moravia and Silesia).
Masarykova univerzita, Český hydrometeorologický ústav, Ústav
geoniky Akademie věd České republiky, Brno Praha Ostrava,
p 432
Chládová Z, Kalvová J, Raidl A (2007) The observed changes of
selected climate characteristics in the period 1961–2000. Meteorol
Čas 10:13–19
Dai A, Trenberth KE, Karl TR (1998) Global variations in droughts
and wet spells: 1900–1995. Geophys Res Lett 25:3367–3370
Dai A, Trenberth KE, Qian T (2004) A global data set of Palmer
Drought Severity Index for 1870–2002: relationship with soil
moisture and effects of surface warming. J Hydrometeorol
5:1117–1130
Domonkos P, Szalai S, Zoboki J (2001) Analysis of drought severity
using PDSI and SPI indices. Idöjárás 195:93–107
Dubrovský M, Nemešová I, Kalvová J (2005) Uncertainties in climate
change scenarios for the Czech Republic. Clim Res 29:139–156
Dubrovský M, Hayes M, Trnka M, Svoboda M, Wilhite DA, Žalud Z,
Semerádová D (2007a) Projection of future drought conditions
using drought indices applied to GCM-simulated weather series.
In: 7th European Meteorological Society Meeting, Madrid,
EMS2007-A-00355
Dubrovský M, Svoboda M, Trnka M, Hayes M, Wilhite D, Žalud Z,
Hlavinka P (2007b) Application of relative drought indices to
assess climate change impact on drought conditions in Czechia.
Theor Appl Climatol doi:10.1007/s00704-008-0020-x
Dufková J, Šťastná M (2005) Determination of drought occurrence
trend in the area of Southern Moravia and its influence on soil
erosion. Meteorol Čas 8:131–136
Fischer EM, Seneviratne SI, Vidale PL, Lüthi D, Schär C (2007) Soil
moisture–atmosphere interactions during the 2003 European
summer heat wave. J Clim 20:5081–5099
Forst P, Caban J, Michalík P (1985) Ochrana lesů a přírodního
prostředí (Protection of forests and natural environment). Státní
zemědělské nakladatelství, Praha, p 416
Frich P, Alexander LV, Della-Marta P, Gleason B, Haylock M, Klein
Tank AMG, Peterson T (2002) Observed coherent changes in
climatic extremes during the second half of the 20th century.
Clim Res 19:193–212
Gerstengarbe FW, Werner PC (1993) Katalog der Grosswetterlagen
Europas nach Paul Hess und Helmuth Brezowsky 1881–1992.
Offenbach am Main, Deutscher Wetterdienst, p 249
Gorczyński W (1920) Sur le calcul du degré de continentalisme et son
application dans la climatologie. Geogr Ann 2:324–331
Heim RR (2002) A review of twentieth-century drought indices used
in the United States. Bull Am Meteorol Soc 83:1149–1165
Hess P, Brezowsky H (1977) Katalog der Grosswetterlagen Europas
1881–1976. Berichte des Deutschen Wetterdienstes 113.
Deutscher Wetterdienst, Offenbach am Main, p 14
Variability of droughts in the Czech Republic, 1881–2006 313
Horváth S (2002) Spatial and temporal patterns of soil moisture
variations in a sub-catchment of River Tisza. Phys Chem Earth
27:1051–1062
Houghton JT, Ding Y, Griggs DJ, Noguer M, van der Linden PJ, Dai X,
Maskell K, Johnson CA (eds) (2001) Climate Change 2001: The
Scientific Basis. Cambridge University Press, Cambridge, p 881
Hundecha Y, Bárdossy A (2005) Trends in daily precipitation and
temperature extremes across western Germany in the second half
of the 20th century. Int J Climatol 25:1189–1202
Huth R, Pokorná L (2004) Trendy jedenácti klimatických prvků v
období 1961–1998 v České republice (Trends in eleven climatic
elements in the Czech Republic in the period 1961–1998).
Meteorol Zpr 57:168–178
Huth R, Pokorná L (2005) Simultaneous analysis of climatic trends in
multiple variables: an example of application of multivariate
statistical methods. Int J Climatol 25:469–484
IPCC (2007) Climate Change 2007: The Physical Science Basis.
Summary for Policymakers. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change (online). IPCC: Paris, p 18. Available at http://
www.ipcc.ch/SPM2feb07.pdf (accessed 2 April 2007)
Jílek J (1957) Atmosférické srážky v Čechách (1876–1956) (Atmospheric
precipitation in Bohemia, 1876–1956). Meteorol Zpr
10:133–134
Katalog povětrnostních situací pro území ČSSR (Cataloque of weather
situations over the territory of the C.S.S.R). Hydrometeorologický
ústav, Praha 1967, p 94
Karl TR (1986) The sensitivity of the Palmer Drought Severity Index
and Palmer’s Z-index to their calibration coefficients including
potential evapotranspiration. J Clim Appl Meteorol 25:77–86
Livada I, Assimakopoulos VD (2007) Spatial and temporal analysis of
drought in Greece using the Standardized Precipitation Index
(SPI). Theor Appl Climatol 89:143–153
Lloyd-Hughes B, Saunders MA (2002) A drought climatology for
Europe. Int J Climatol 22:1571–1592
Markham CG (1970) Seasonality of precipitation in the United States.
Ann Assoc Am Geogr 60:593–597
Marsh T, Cole G, Wilby R (2007) Major droughts in England and
Wales, 1800–2006. Weather 62:87–93
Minář M (1947) Vliv počasí na zemědělství v jednotlivých měsících
(Weather influence on agriculture in individual months). Meteorol
Zpr 1:33, 36, 60, 84, 92
Moberg A, Jones PD (2005) Trends in indices for extremes in daily
temperature and precipitation in central and western Europe,
1901–99. Int J Climatol 25:1149–1171
Moliba JC, Huth R, Beranová R (2006) Roční chod trendů klimatických
prvků v České republice (Annual cycle of trends in climatic
elements in the Czech Republic). Meteorol Zpr 59:129–134
Možný M (2004) Vymezení a intenzita sucha na území ČR v letech
1891–2003 (Delimitation and intensity of drought over the Czech
Republic between 1891 and 2003). Český hydrometeorologický
ústav, Praha, p 88
Ntale HK, Gan TY (2003) Drought indices and their application to
East Africa. Int J Climatol 23:1335–1357
Olberg M (1982) Statistische Analyse meteorologisch-klimatologischer
Zeitreihen. Abh Meteorol Dienstes DDR 128:129–141
Olberg M, Rákóczi F (1984) Informationstheorie in Meteorologie und
Geophysik. Mit besonderer Berücksichtigung der MaximumEntropie-Spektralschätzung.
Akademie-Verlag, Berlin, p 181
Pal JS, Giorgi F, Bi X (2004) Consistency of recent European summer
precipitation trends and extremes with future regional climate
projections. Geophys Res Lett 31:L13202
Palmer WC (1965) Meteorological drought. Office of climatology
research paper 45. Weather Bureau, Washington, D.C., p 58
Palmer WC (1968) Keeping track of crop moisture conditions,
nationwide: the Crop Moisture Index. Weatherwise 21:156–161
Panu US, Sharma TC (2002) Challenges in drought research: some
perspectives and future directions. Hydrolog Sci J 47(S):S19–S30
Pavlík J, Němec L, Tolasz R, Valter J (2003) Mimořádné léto roku
2003 v České republice (Extraordinary summer 2003 in the
Czech Republic). Meteorol Zpr 56:161–165, and I–VIII
Petr J (ed) (1987) Počasí a výnosy (Weather and yields). Státní
zemědělské nakladatelství, Praha, p 368
Pišoft P, Kalvová J, Brázdil R (2004) Cycles and trends in the Czech
temperature series using wavelet transforms. Int J Climatol
24:1661–1670
Pongrácz R, Bogardi I, Duckstein L (2003) Climatic forcing of
droughts: a Central European example. Hydrolog Sci J 48:39–50
Quiring SM, Papakryiakou TN (2003) An evaluation of agricultural
drought indices for the Canadian prairies. Agr Forest Meteorol
118:49–62
Rebetez M (1999) Twentieth century trends in droughts in southern
Switzerland. Geophys Res Lett 26:755–758
Řezníčková L, Brázdil R, Tolasz R (2007) Meteorological singularities
in the Czech Republic in the period 1961–2002. Theor Appl
Climatol 88:179–192
Schär C, Jendritzky G (2004) Hot news from summer 2003. Nature
432:559–560
Schmidli J, Frei C (2005) Trends of heavy precipitation and wet and
dry spells in Switzerland during the 20th century. Int J Climatol
25:753–771
Schönwiese CD (1985) Praktische Statistik für Meteorologen und
Geowissenschaftler. Gebrüder Borntraeger, Berlin, Stuttgart, p 231
Schönwiese CD, Staeger T, Trömel S (2004) The hot summer 2003 in
Germany. Some preliminary results of a statistical time series
analysis. Meteorol Z 13:323–327
Scian BV (2004) Environmental variables for modeling wheat yields
in the southwest pampa region of Argentina. Int J Biometeorol
48:206–212
Sönmez FK, Kömüscü AÜ, Erkan A, Turgu E (2005) An analysis of
spatial and temporal dimension of drought vulnerability in Turkey
using Standardized Precipitation Index. Nat Hazards 35:243–264
Stefan S, Ghioca M, Rimbu N, Boroneant C (2004) Study of
meteorological and hydrological drought in southern Romania
from observational data. Int J Climatol 24:871–881
Svoboda M, LeComte D, Hayes M, Heim R, Gleason K, Angel J,
Rippey B, Tinker R, Palecki M, Stooksbury D, Miskus D,
Stephens S (2002) The drought monitor. Bull Am Meteorol Soc
83:1181–1190
Štěpánek P (2004) Homogenizace teploty vzduchu na území České
republiky v období přístrojových pozorování (Homogenisation of
air temperature over the territory of the Czech Republic in the
instrumental period). Práce a studie, sešit 32. Český hydrometeorologický
ústav, Praha, p 56
Szinell CS, Bussay A, Szentimrey T (1998) Drought tendencies in
Hungary. Int J Climatol 18:1479–1491
Thornthwaite CW (1948) An approach towards a rational classification
of climate. Geogr Rev 38:55–94
Tolasz R, Míková T, Valeriánová A (2007) Atlas podnebí Česka
(Climatic atlas of Czechia). Český hydrometeorologický ústav,
Praha Olomouc, p 256
Tomášek M (2000) Půdy České republiky (Soils of the Czech
Republic). Česká geologická služba, Praha, p 67
Trigo RM, García-Herrera R, Díaz J, Trigo IF, Valente MA (2005)
How exceptional was the early August 2003 heatwave in France?
Geoph Res Lett 32:L10701
Trnka M, Dubrovský M, Svoboda M, Semerádová D, Hayes M, Žalud Z,
Wilhite D (2007a) Developing a regional drought climatology for
the Czech Republic. Int J Climatol doi:10.1002/joc.1745
Trnka M, Hlavinka P, Semerádová D, Dubrovský M, Žalud Z, Možný
M (2007b) Agricultural drought and spring barley yields in the
Czech Republic. Plant Soil Environ 53:306–316
314 R. Brázdil et al.
Trnka M, Kyselý J, Možný M, Dubrovský M (2008) Changes in
Central-European soil-moisture availability and circulation patterns
in 1881–2005. Int J Climatol doi:10.1002/joc.1703
van der Schrier G, Briffa KR, Jones PD, Osborn TJ (2006) Summer
moisture variability across Europe. J Clim 19:2818–2834
van der Schrier G, Efthymiadis D, Briffa KR, Jones PD (2007) European
Alpine moisture variability 1800–2003. Int J Climatol 27:415–427
Wells N, Goddard S, Hayes M (2004) A self-calibrating Palmer
drought severity index. J Clim 17:2335–2351
Wilhite D (ed) (2000) Drought: a global assessment. Vol. 1. Routledge
Publishers, London, p 422
Yamoah CF, Walters DT, Shapiro CA, Francis CA, Hayes MJ (2000)
Standardized precipitation index and nitrogen rate effects on crop
yields and risk distribution in maize. Agr Ecosyst Environ
80:113–120
Zimolka J (ed) (2006) Ječmen – formy a užitkové směry v České
republice (Barley – types and products used in the Czech
Republic). Profi-Press, Praha, p 200
Variability of droughts in the Czech Republic, 1881–2006 315