Content uploaded by Christoph Aubrecht
Author content
All content in this area was uploaded by Christoph Aubrecht
Content may be subject to copyright.
ORIGINAL ARTICLE
Assessment of regional climate change impacts
on Hungarian landscapes
Gabor Mezo
¨si •Burghard C. Meyer •
Wolfgang Loibl •Christoph Aubrecht •
Peter Csorba •Teodora Bata
Received: 21 January 2012 / Accepted: 1 June 2012 / Published online: 7 July 2012
Springer-Verlag 2012
Abstract The assessment of regional climate change
impacts combined with the sensitivity of landscape func-
tions by predictive modelling of hazardous landscape
processes is a new fundamental field of research. In par-
ticular, this study investigates the effects of changing
weather extremes on meso-regional-scale landscape vul-
nerability. Climatic-exposure parameter analysis was per-
formed on a predicted climate change scenario. The
exposure to climate change was analysed on the basis of
the original data of the meso-scale IPCC A1B climate
scenario from the REMO and ALADIN regional models
for the periods of 2021–2050 and 2071–2100, and the
regional types of climate change impacts were calculated
by using cluster analysis. Selected climate exposure
parameters of the REMO and ALADIN models were
analysed, in particular, for extreme events (days with pre-
cipitation greater than 30 mm, heat waves, dry periods, wet
periods) and for daily temperature and precipitation. The
landscape functions impacted by climate change are
proxies for the main recent and future problematic pro-
cesses in Hungary. Soil erosion caused by water, drought,
soil erosion caused by wind, mass movement and flash
floods were analysed for the time periods of 1961–1990,
2021–2050 and 2071–2100. Based on the sensitivity
thresholds for the impact assessments, the landscape
functional sensitivity indicators were interpreted, and an
integrative summary of the five indicators was made, dif-
ferentiating the regions facing only a few or multiple
sensitivities. In Central Hungary, the increasing exposure
and sensitivity to droughts will be a serious problem when
following the REMO scenario. In several regions, most
indicators will change the sensitivity threshold from a
tolerable risk to an increased or very high risk.
Keywords Climate change Sensitivity of landscape
Impact of hazardous landscape processes Carpathian
basin Regionalisation
Introduction
The assessment of regional climate change impacts is a
new fundamental field of research, especially when
investigating the effects of changing weather extremes,
exposure at a landscape scale, sensitivity and vulnerability.
G. Mezo
¨si (&)T. Bata
Department of Physical Geography and Geoinformatics,
University of Szeged, Egyetem str. 2, Szeged 6722, Hungary
e-mail: mezosi@geo.u-szeged.hu
T. Bata
e-mail: batateodora@gmail.com
B. C. Meyer
Institut fu
¨r Geographie, Universita
¨t Leipzig, Johannisallee 19a,
04103 Leipzig, Germany
e-mail: burghard.meyer@uni-leipzig.de
W. Loibl C. Aubrecht
Austrian Institute of Technology, Donau-City-Str. 1,
1220 Wien, Austria
e-mail: wolfgang.loibl@ait.ac.at
C. Aubrecht
e-mail: christoph.aubrecht@ait.ac.at
C. Aubrecht
University of Southern California, Donau-City-Str. 1,
1220 Wien, Austria
P. Csorba
Department of Landscape Protection and Environmental
Geography, University of Debrecen, Egyetem te
´r1,
Debrecen 4010, Hungary
e-mail: csorbap@tigris.unideb.hu
123
Reg Environ Change (2013) 13:797–811
DOI 10.1007/s10113-012-0326-1
To assess regional climate change impacts, this study uses
the IPCC greenhouse-gas emission scenario A1B that was
calculated for Central Europe by Bartholy et al. (2008) and
Pongra
´cz et al. (2009) by applying the REMO, ALADIN or
PRECIS regional climate model. These results on the cli-
mate change exposure for Hungary predict (1) an increased
average temperature, mainly in the summer, and a higher
variability of rainfall, especially in the summer and winter;
(2) an increased number of tropical days; (3) an increased
length of heat periods; (4) changes in the summer and
winter precipitation; and (5) a slightly increased number of
days with heavy rains (more than 30 mm).
Rannow et al. (2010) have developed a multifunctional
assessment framework to create a model for Germany by
varying the regional priorities of adaptation activities in spa-
tial planning, using 11 ‘‘impacts of primary relevance for
spatial planning on a regional level’’. The methodology used
by Meyer et al. (2009) and Rannow et al. (2010) is applied for
the regionalisation of the impacts when calculating the simple
predictive risk assessment and combining the modelled
changes of the climate variables with maps and statistics. The
results are potential regional hazard sensitivities at the scale of
German administrative districts and natural regions.
It is obvious that several scientific problems should be
considered when using coarse-grained pixel information
from regional climate modelling on a landscape or eco-
system level with the aim to apply a sound approach and to
investigate future climate change impacts that are relevant
for planning and risk prevention, especially in the case of
extreme events. A multitude of interesting research dis-
cussions and investigations have occurred in recent years;
some notable examples include the following: Jentsch and
Bayerkuhnlein (2008) on the effects of extreme meteoro-
logical events on ecosystems; Opdam and Washer (2004)
on the interlinkage of climate change and habitat frag-
mentation; Martens et al. (2010) on the institutional and
multisectoral perspectives of abrupt and extreme climate
changes; Pielke et al. (2007) on the changing role of
agriculture in the climate system; and Metzger et al. (2008)
with a spatially explicit and quantitative vulnerability
assessment of the ecosystem changes in Europe.
The investigation of the multifunctional assessment of
climate change impacts on Hungarian landscapes should
help meet the public demand for research results regarding
sensitivity assessments using landscape functions and
landscape processes as indicators to clarify ‘‘relevant direct
and indirect impacts of climate change with a special focus
on the regional differentiation of its effects’’ (Meyer et al.
2010). Information about the regional and local impacts of
climate change is of interest to several branches of plan-
ning and risk prevention (Meyer et al. 2009).
For Hungary, the coarse-grained climate information of
the available regional modelling (of 11 min 911 min or
approx. 25 km 925 km) leads to a regional approach. The
overlay of the A1B climate scenario by the REMO and
ALADIN regional climate models with the regional land-
scape types in Hungary is employed to assess the crucial
changes in exposure, sensitivity and landscape vulnerabil-
ity. Problematic landscape processes are investigated in
this study by analysing the changes in soil erosion by
water, droughts, wind erosion, mass movements and flash
floods by comparing three periods: 1961–1990, 2021–2050
and 2071–2100. The aim of the research is to determine the
main climate change factors (exposure), the affected earth-
surface processes (sensitivity) and the critical impacts on
the landscapes in the next century as a result of the pre-
dicted climatic tendencies.
In the following, the methodological and data-driven
steps of the analysis regarding (1) the climate exposure in
Hungary based on regional models emphasising extreme
events, (2) the landscape types of Hungary and (3) the
methods of sensitivity risks analysis and assessments are
described. This part is followed by an integrative impact
assessment to predict the climate change–induced changes
of the landscape sensitivity for the mentioned time periods.
The discussion leads to open questions about the vulnera-
bility assessment and the application of the results in policy
and adaptive land management in the face of the problems
of the data, scale and accuracy of the scenario methods.
Data and methods
Landscape types of Hungary
The usage of landscape units has a long scientific history in
Hungary. Pe
´csi and Somogyi (1967) have differentiated a
map of landscape micro-regions. These micro-regions are
defined by using the geomorphologic characteristics of the
land forms for the regional and local typifications. The
regional climatic models used in this study did not include
enough data for a micro-region (sometimes only one or two
data points fall into one micro-region). Therefore, it was
practical to analyse at the aggregated hierarchical level of
meso-scalic landscape units (see Fig. 1). Two main meth-
ods are generally suitable for the delimitation of landscape
units when scaling up microscale information to the meso-
scale: integration or segmentation techniques. A segmen-
tation process was used in LANMAP (European Landscape
Typology and Map) by Mu
¨cher et al. (2010). The Hun-
garian part of the LANMAP map is not scientifically val-
idated. The segmentation process employed in LANMAP
on raster information resulted in meso-scale landscape
units that are not yet interpretable by the Hungarian
microscale data, especially because of the resulting small
units that are not explainable on a microscale level.
798 G. Mezo
¨si et al.
123
Therefore, the integration of the micro-regions was used to
aggregate the meso-regions based on the map of Pe
´csi and
Somogyi (1967; see Fig. 1). It was derived from geomor-
phologic (mountain, hills, plains) and land cover-/land-
scape-type characteristics. Because the present paper is
focused on the determination of land use and the hazardous
consequences of climate change usable in regional plan-
ning, it was reasonable to define 18 meso-regions with an
area of several thousands of square kilometres each, made
up of an integration of 230 micro-regions. These regions
are defined as homogeneous at the regional-scale level.
Climate change in Hungary
It is difficult to sketch a valid picture of climate change for
the entire area of Hungary. Pinning down tendencies is
hindered by multiple factors. One initial basic problem was
the choice of the climate change scenario to be considered
(IPCC 2007; Bartholy and Pongra
´cz 2010). In our analysis,
only the data from the A1B scenario are used following a
balanced storyline that does not belong to a single energy-
source change as described by IPCC (2007). Another
essential aspect is the selection of the regional climate
change model that is employed. These models are different,
not only in terms of the applicability of scale (global,
regional) but also in their basic model assumptions. The
results have been tested from four regional models (AL-
ADIN, REMO, PRECIS, RegCM), out of which the REMO
and ALADIN models are employed here because these
models are the most reliable for Central Europe.
Based on these prerequisites in the choice of scenario
and model and by following a similar rate and tendency as
observed between 1961 and 1990, our investigation resul-
ted in a continuous but uneven temperature increase
focused on the summer period in the Carpathian Basin for
this century (Table 1). The temporal distribution of pre-
cipitation will become increasingly variable, likely more
than the annual sum in the scenario periods of 2021–2050
and 2071–2100 predicted (Table 1). The analysis of the
REMO and ALADIN model scenario runs resulted in daily
averages, monthly extreme temperatures and similar trends
for the precipitation parameters and daily precipitation
rates higher than 30 mm or in the length of heat waves. In
Table 1, the differences in the predictions in the tempera-
ture and precipitation of the two models are given as
averages of REMO and ALADIN, including the range of
the predicted outcomes. These results are the main con-
clusions of the Hungarian Meteorological Service for our
study. The details are also described for the other models,
such as those by Pieczka et al. (2011) and Szabo
´et al.
(2011), which are predicted by similar trends.
Fig. 1 The 18 meso-regions of
Hungary (made up of 230
micro-regions) differentiated
into plains, hilly and
mountainous landscapes. The
meso-region is an integration of
geomorphologic
characterisation and landscape
types
Table 1 The average change in temperature (C) and precipitation
(mm) and the variation of the predictions in the Carpathian Basin
using the REMO and ALADIN models’ A1B scenario compared to
period 1961–1990, Szabo
´et al. (2011)
Period Years Spring Summer Autumn Winter
2021–2050 1.4–1.9 1.1–1.6 1.4–2.6 1.6–2.0 1.3
2071–2100 3.5 2.3–3.1 4.1–4.9 3.6–3.8 2.5–3.9
Assessment of regional climate change impacts on Hungarian landscapes 799
123
The model’s estimate for 2021–2050 precipitation is not
significant, but the results indicate a significant increase in
the single-event intensity. Bartholy et al. (2007) have
emphasised a potential autumn/wintertime precipitation
increase but no increase for the summer.
The simulations of the weather extremes project more
intense rainfall as well as longer warm and more frequent
dry periods in the Carpathian Basin. The number of frost
days will decrease by 30 and 50 % in the middle and end of
the century, respectively. In the meantime, the number of
hot days will double and triple (Sze
´pszo
´2008). The pro-
jection of the precipitation is, however, less straightforward
because the models predict changes at different scales
following the several incorporated methodological weak-
nesses in the modelling architecture. A sensitivity analysis
was applied in detail describing the ranges of the change of
the different climate change scenarios of the REMO and
ALADIN models for the Carpathian Basin by Csima and
Hora
´nyi (2008) and Sze
´pszo
´and Hora
´nyi (2008).
Distinguishing climate-region types for Hungary
The REMO and ALADIN simulations have provided
results that show differences in certain regions (particularly
in the east and south), depicting the uncertainty range by
applying different climate models. Multivariate statistical
analysis and classification have been performed consider-
ing the climate change trends in the Hungarian climate
regions and including the uncertainty range of the model
simulations. The objective was to distinguish the region
types with similar climate change characteristics. Hungary
has a rather homogenous terrain, whereas larger or moun-
tainous/alpine countries show a broader variation of local
climates. As the landscape diversity in Hungary is—when
described by statistics—relatively narrow, distinguishing
the climate regions was a somewhat challenging task that
was performed by applying factor and cluster analyses
(Loibl and Aubrecht 2011). Factor analysis was used for
reducing the number of variables and generating a few
distinct and integrated ‘‘super-indicators’’ out of (via factor
loadings) weighted input variables—the factor coefficients.
Cluster analysis was used to group the Hungarian regions
by the climate characteristics of the current and future
climate by applying multivariate statistics. To conduct the
clustering task, climate and climate change indicators were
extracted from raster sets of precipitation, temperature and
extreme-event indicators by averaging the indicators for
the 18 Hungarian meso-regions.
Therefore, the 20 km 920 km grid-cell results as
derived from the climate simulations were transferred to
the small Hungarian meso-regions by calculating the spa-
tially weighted spatial averages, ranges and standard
deviations of the climate data for those regions. A set of
factor analyses were performed to derive the appropriate
integrated ‘‘super-indicators’’ for the regions.
Finally, the factor analysis results were obtained for the
temperature and precipitation data subsets individually.
The factor analysis using the temperature data integrates
the ‘‘current temperature’’ (1961–1990), the temperature
from 2021 to 2050 and the temperature from 2071 to 2100
of the ALADIN and REMO scenarios and the current-to-
future temperature change. The temperature subsets con-
tain absolute numbers for the current and future tempera-
ture (the regional average and range within the region) and
the averages of frost days (B0C Tmin) and summer days
([25 C Tmax) by scenario version. The factor analysis
using the precipitation data integrates the ‘‘current pre-
cipitation’’ (1961–1990), the precipitation from 2021 to
2050 and the precipitation from 2071 to 2100 of the
ALADIN and REMO scenarios. The precipitation subsets
contain current and future rainfall absolutes (totals and
range within the regions), the numbers of extreme rainfall
days ([20 mm and [30 mm per day) and the average
daily rainfall sum on precipitation days ([1mm) by
scenario.
The factor analyses deliver the factor coefficients for the
meso-regions as ‘‘super-indicators’’. The coefficients of
those factors whose eigenvalues explain more than 10 % of
all variables’ variance (usually 2 factors per analysis) were
selected. These factor coefficients of the few important
factors describe the current and future climates from the
ALADIN and REMO scenario results.
Using those factor coefficients, cluster analyses were
performed with alternative linkage approaches and met-
rics to identify the regions of similar characteristics by
detecting the natural groupings in the data. Hierarchical
clustering, which records the tightness of linkages
between the factor coefficients by observing the similarity
or ‘‘distance’’ between the values by (region) case, is
typically applied. Several distance metrics and linking
methods are available with hierarchical clustering. Ward’s
linkage method was applied, which averages all the dis-
tances between pairs of objects in different clusters, with
adjustments for the covariance, to determine how far apart
the clusters are. As a distance metric, the normalised
Euclidean distance (root-mean-squared distance) was
used.
The cluster analysis results that are ultimately selected
to delineate the climate regions are based on the factor
coefficients of the 2 highest factors of the 4 different factor
analyses for the temperature and the precipitation consid-
ering the scenario results for 2021–2050 and 2071–2100
and integrating the temperature and precipitation ranges of
the meso-regions.
800 G. Mezo
¨si et al.
123
The analysis delivers four main climate-region types
(Fig. 2):
•Region type 1 covers 2 meso-regions contained in
cluster 1 in the western hilly area. This type is character-
ised by lower temperatures, less temperature increase and
less change in the temperature extremes. The type is
expected to be more humid, with higher precipitation
totals but smaller precipitation-change ratios and smaller
change rates regarding heavy rain events.
•Region type 2 indicates a moderate temperature
increase and more distinct changes in extreme temper-
ature events. The precipitation totals are moderate; the
future precipitation increase is expected at higher rates
but with moderate changes in extreme rainfall events.
The meso-regions of this type are located along a west
central corridor from the north to the south of Hungary.
•Region type 3 is characterised by a flat topography, with
the highesttemperatures, the highest temperature increase
and significant changes in the extreme temperature events
(increase of the summer days and decline of the frost
days). This type suffers from the lowest annual precipi-
tation totals and can expect the highest precipitation
decline (or, at least, the lowest precipitation-increase
ratios). The increase in heavy-rainfall days indicates a
growing concentration of rainfall and thus longer drought
periods. This region type covers a large area of Hungary,
ranging from the centre to the south-east.
•Region type 4 covers the north-eastern meso-regions
along the Slovakian border, with the lowest annual mean
temperatures, the highest intraregional temperature
variation and moderate precipitation totals resulting in
more humidity when compared with the other continental
regions. The moderate increase in the precipitation sum
and the (few) extreme-event days may have positive
effects on agriculture and nature evolution.
In Table 2, the regional differentiation of climate
change exposure is summarised for the regional landscapes
of Hungary, including the data for the periods of
1961–1990, 2021–2050 and 2071–2100 (projections based
on the REMO & ALADIN simulations). Differentiation
refers to the temperature and precipitation totals for sum-
mer days, tropical days and days with precipitation greater
than 30 mm.
Investigations of the landscape sensitivity in Hungary
due to climate change assessments of landscape hazards
In order of their actual importance in Hungary, the fol-
lowing natural processes were considered during the anal-
ysis: soil erosion by water, droughts, soil erosion by wind,
flash floods and mass movements. These processes repre-
sent the most important environmental hazards for land use
in Hungary (Szabo
´et al. 2008). The flood problem was not
included into our analysis because the increase in this pro-
cess is not primarily a consequence of the climate impacts
(influencing factors are the capacity of water-deduction of
the floodplain, the land use of the catchment and the con-
struction of dams in the upper section), and the floods are
difficult to predict in Carpathian Basin at a regional scale
without an enclosure of the surrounding mountains.
Fig. 2 The regional types of climate change exposure as a result of cluster analysis using all factors
Assessment of regional climate change impacts on Hungarian landscapes 801
123
Table 2 The landscape regions of Hungary and the impacts of climate change (REMO and ALADIN) for the periods of 1961–1990, 2021–2050 and 2071–2100
Code Name Temperature (C)
1961–1390
Temperature (C)
2021–2050
Temperature (C)
2071–2100
Precipitation (mm)
1961–1990
Precipitation (mm)
2021–2050
Precipitation (mm)
2071–2100
SU
1961–1990
SU
2021–2050
1 Western part of the
Carpathian basin
9.9 11.4 13.3 603 626 626 49.8 68.9
2 Loess and sand plain of
Nyı
´rse
´g and Hajdu
´sa
´g
10.0 11.7 13.6 571 552 549 55.3 79.4
3 Western part of the North
Hungarian Mountains
9.6 11.1 13.1 536 535 516 48.6 71.2
4 Little Hungarian Plain 10.3 11.7 13.7 530 552 545 52.8 70.6
5 Marcal basin and Koma
´rom
plain
10.1 11.6 13.5 541 559 555 51.7 70.7
5 Zala hills 10.0 11.5 13.5 656 671 659 60.5 80.9
7 Hilly region of inner Somogy 10.4 11.5 13.5 625 648 637 70.5 90.1
8Go
¨do
¨ll}
o hills 10.4 12.1 14.0 506 500 487 61.9 83.7
9 Dra
´va plain and Mecsek
Mountains
10.7 12.4 14.3 576 569 569 77.3 98.7
10 Mez}
ofo
¨ld plain 10.5 12.1 14.0 524 527 529 60.9 81.3
11 Transdanubian hills 10.5 12.1 14.0 567 577 572 66.4 87.0
12 Danube-Tisza Interfluve 10.7 12.4 14.3 493 478 47 72.7 94.3
13 Central part of die Great
Hungarian Plain
10.4 12.1 14.0 497 474 466 65.4 88.7
14 Ko
¨ro
¨s-Maros Interfluve 10.6 12.3 14.2 488 461 458 75.1 98.1
15 Plain of Upper Tisza 9.7 11.3 13.2 659 653 641 50.8 74.3
16 Transdannubian Mountains 10.1 11.6 13.5 556 569 567 53.3 74.4
17 Eastern part of hie North
Hungarian Mountains
9.1 10.7 12.6 564 565 549 44.8 67.9
18 Danube plain 10.7 12.4 14.3 508 503 501 69.6 90.2
Code Name SU
2071–2100
HEAT
1961–1990
HEAT
2021–2050
HEAT
2071–2100
RR30
1961–1990
RR30
2021–2050
BR30
2071–2100
1 Western part of the Carpathian basin 95.9 4.6 15.3 40.0 1.1 1.5 1.7
2 Loess and sand plain of Nyı
´rse
´g and Hajdu
´sa
´g 103.8 7.1 23.4 47.4 0.5 0.6 0.9
3 Western part of the North Hungarian Mountains 98.9 3.4 14.1 36.4 0.6 1.0 1.2
4 Little Hungarian plain 96.5 7.4 21.1 45.4 0.6 1.1 U
5 Marcal basin and Koma
´rom plain 96.0 6.2 13.4 43.7 0.6 1.0 1.3
5 Zala hills 106.0 6.1 13.6 46.6 1.2 1.5 1.6
7 Hilly region of inner Somogy 113.1 10.0 27.8 55.4 0.7 1.3 1.4
8Go
¨do
¨ll}
o hills 107.8 9.2 25.2 51.3 0.4 0.7 1.0
9 Dra
´va plain and Mecsek Mountains 119.5 11.9 32.3 59.2 0.6 0.8 1.1
802 G. Mezo
¨si et al.
123
Following these processes, a number of landscape
function–based sensitivities have been spatially assessed
using predictive models and diverse geo-data (Table 3)
from multiple sources. The main aims are (1) the explo-
ration of the status quo of the indicators’ assessment on the
regional scale for Hungary and (2) the assessment of the
usage of the climate change parameter predictions mod-
elled by REMO and ALADIN and typified by cluster
analysis in the sensitivity assessments.
In the following, the main landscape processes, the
models employed to predict the landscape indicators and
the thresholds used for the sensitivity assessment are
described (Usher 2001). An overview of the parameters,
the calculation steps of the predictive models and the
methods used to model the landscape hazard indicators,
including the thresholds for the sensitivity assessment, are
given in Table 4.
Soil erosion by water considers the physical soil deg-
radation processes at today’s largest spatial extent in
Hungary. It affects 2 million ha of productive land on Late
Tertiary and Quaternary alluvial and lacustrine clays, silty
sediments and loess (Stefanovits 1981). The erosion sen-
sitivity was calculated on a micro-regional scale following
the universal soil loss equation (USLE) of Wischmeier and
Smith (1978) adapted for Hungary. The parameters deter-
mining the soil sensitivity (K), length of slope (L) and
steepness (S) are relatively stable; the rainfall erosivity
factor (R) has the closest link to climate change. The
vegetation and crop factor (C) and the measures against
erosion (P) have a high degree of unpredictability because
of changing land-use systems and also due to potential
protective adaptive measures in the future. In our model-
ling example, the average RR30 value in the winter half-
years was used for the calculations of extreme rainfall
events (the RR30 values were calculated from REMO and
ALADIN model).
Droughts, in general, are natural phenomena, though
human activity can have a high indirect influence on them,
especially via land use. As a consequence, the estimates of
the drought sensitivity include high data uncertainties
compared with the other natural disaster predictions. The
sensitivity map in this study was developed using the Pa
´lfai
Drought Index (PaDI). The PaDI was calculated as the ratio
of the mean temperature of the summer months (April–
August) and the annual sum of precipitation. As a proxy for
the sensitivity changes induced by climate change expo-
sure, the regional types of climate change exposure are
applied (Fig. 2) as a result of the cluster analysis using all
climatic factors.
Wind erosion in the Carpathian Basin is not usually an
issue in sand-covered areas but frequently has an effect on
the degradation of arable soils. The wind erosion sensitivity
is primarily determined by the granulometry of the soil; for
Table 2 continued
Code Name SU
2071–2100
HEAT
1961–1990
HEAT
2021–2050
HEAT
2071–2100
RR30
1961–1990
RR30
2021–2050
BR30
2071–2100
10 Mez}
ofo
¨ld plain 104.7 9.2 25.4 51.3 0.5 0.9 1.1
11 Transdanubian hills 109.9 3.2 24.1 52.7 0.6 1.0 1.2
12 Danube-Tisza Interfluve 115.0 14.1 34.3 60.4 0.4 0.5 0.7
13 Central part of die Great Hungarian Plain 112.0 11.0 30.0 56.0 0.5 0.6 0.8
14 Ko
¨ro
¨s-Maros Interfluve 119.2 14.4 36.1 62.4 0.4 0.4 0.7
15 Plain of Upper Tisza 99.6 5.6 19.8 43.3 0.8 0.8 1.4
16 Transdannubian Mountains 100.0 5.3 17.0 41.2 0.5 1.0 1.2
17 Eastern part of hie North Hungarian Mountains 96.3 2.9 12.3 33.0 0.7 1.0 1.2
18 Danube plain 111.9 12.9 31.6 58.1 0.4 0.7 1.0
SU summer days, HEAT tropic days, RR30 days with precipitation greater than 30 mm, code region number
Assessment of regional climate change impacts on Hungarian landscapes 803
123
example, if a great amount of fine sediment is available on
the surface, deflation will be more intense as lower and
more frequent wind velocities will be adequate to reach the
critical entrainment force. The characteristic yearly aver-
age wind velocities close to the surface are 3 m/s in
Hungary, though the values are 15–20 % higher in the NW
and central parts of the basin. The Lo
´ki (2011) map of the
potential wind erosion hazard was used in this study, and
the climatic parameter affected by climate change for the
prediction of the sensitivity was associated with the
regional types of climate changes as result of the cluster
analysis shown in Fig. 2.
Flash flooding is one of the most frequent hazards in
Hungary (Cziga
´ny et al. 2010; Estrela et al. 2001). In the past
decades, the most important hydrometeorological parame-
ters have been clarified (Grundfest and Rips 2000). Moni-
toring systems have been set up to help mitigate this hardly
predictable hazard (Carpenter et al. 1999). A point-based
calculation was applied in this study by using the classifi-
cation given in Table 3. The climatic parameter that was
taken as the climatic-exposure climate change for the flash
flood prediction is the temporal frequency of the extreme
precipitation events higher than 30 mm (Sze
´pszo
´2008).
Mass movements were evaluated according to their
present-day activity. The endangered areas were predicted
in the mountains and hilly regions where the natural con-
ditions are able to mobilise the processes by using the
recorded information about recent significant landslide
events (Juha
´sz 2004; Fodor and Kleb 1986), geology, the
granulometric type of the sediment, relief and the actual
precipitation data for Hungary. Hilly regions with only
ancient quaternary mass movement have not been consid-
ered. The climatic parameter affected by climate change is
the sum of the precipitation in the winter season.
Thresholds for sensitivity assessment
The sensitivity was assessed by using the threshold values
for the classification of the hazard for each of the indicators
in the sensitivity classes of the regional landscape in
qualitative terms of low/tolerable (class 1), increased (class
2) and high (class 3). The thresholds are identified for each
indicator as described in Table 4. The usage of qualitative
classes is a standard procedure for an equal-weighted
integration of different factors in impact assessment. The
applied method is simple because the uncertainty of the
models (climatic and hydrological) and the limited amount
of verification do not enable highly precise calculations.
Practically, a matrix-based assessment was made to link the
climatic exposure of the region to its sensitivity to the
problematic processes.
Results: regionalised climate change impacts
and sensitivity assessments
The aim of the study was not to verify the quality or scaling
of different regional climate modelling. The data of REMO
Table 3 The main data sources for the calculation of the sensitivity indicators for Hungary
Sensitivity indicator Source of basic data
Soil erosion by water USLE map (Kerte
´sz and Centeri 2006)
Results of cluster analysis (sensitivity values were confronted to climate change indicating
parameter clusters (Loibl and Aubricht 2011)
Drought Daily/monthly average temperature (from REMO and ALADIN models)
Daily/monthly precipitation from REMO and ALADIN models)
Results of cluster analysis (sensitivity values were confronted to climate change indicating
parameter clusters (Loibl and Aubricht 2011)
Wind erosion Map on potential wind erosion risk (Lo
´ki 2011)
Soil type (Agrotopographical database 1991)
Results of cluster analysis (sensitivity values were confronted to climate change indicating
parameter clusters (Loibl and Aubricht 2011))
Flash floods Slope map (SRTM 2000)
Soil type (Agrotopographical database 1991)
Forest categories (European Environment Agency 2000)
Results of cluster analysis (sensitivity values were confronted to climate change indicating
parameter clusters (Loibl and Aubricht 2011))
Mass movements Mass movements of last decades (1965–2005) (Juha
´sz 2004, Fodor and Kleb 1986)
The precipitation sum of the winter season (from REMO and ALADIN models)
Results of cluster analysis (sensitivity values were confronted to climate change indicating
parameter clusters (Loibl and Aubricht 2011))
804 G. Mezo
¨si et al.
123
Table 4 The parameters, calculation steps, methods and sensitivity thresholds used for the climate change impact ON the landscape hazard indicators
Indicator Parameters Calculation Method/climate change affection Indicator sensitivity threshold
Soil erosion by water Rainfall erosivity factor (R),
Soil erodibility factor (K),
Length of slope (L),
Steepness (S),
Crop factor (C)
Protective measures (P)
Universal soil loss equation (USLE)
E=RK(LS)CP (t/ha)
E=soil erosion by water
USLE; Wischmeier and Smith
(1978) was used for risk
calculation
The climatic parameter effected by
climate change: Extreme rainfall
events and winter precipitation
were considered for the
assessment of climatical effect
(R-Faktor change)
1. Tolerable: 0–2 t/ha
2. Increased: 2–8 t/ha
3. High: [8 t/ha
Drought Precipitation
Temperature
PaDI =P
(i=Apr–Aug)
T
i
90.05/
P
(Oct–Sept)
P
i
T—monthly average temperature
P—monthly precipitation
PaDI—Palfai drought index
Pa
´lfai (2004) drought index
calculation (PaDI) was used for
drought sensitivity mapping
The climatic parameter affected by
climate change: sensitivity
values were confronted to
climate change indicating
parameter clusters (Loibl and
Aubricht 2011)
1. Tolerable: \6
2. Increased: 6–8
3. High: [8
Wind erosion Granulometry of the sediment/soil cover
Climatologic characters (starter velocity,
precipitation)
Vegetation cover;
Wind speed
GIS based calculation using WEQ
(2000), Klik (2004) and Lo
´ki (2011)
map on potential wind erosion risk
The wind erosion calculations,
based on Chepil formula, and on
Agro-topographical maps (1991)
The climatic parameter affected by
climate change: The sensitivity
values were confronted to
regional types of climate changes
as result of the cluster analysis
Fig. 2(Loibl and Aubricht 2011)
1. Tolerable: WE processes starts
on loamy and silty soil by a wind
speed of 8.6 and 10.5 m/s
2. Increased: WE starts at sandy
loamy soils at wind speed
between 6.5 and 8.5 m/s
3. High: WE starts on sandy and
peat bog soils with high amount
of organic material by a wind
speed of around 6.5 m/s
Flash flood Slope categories (area based)
Granulometry (area based)
Forest surfaces (area based)
Classification:
Slope: \0.1 % 0; 0.1–5 % 1; 5–30 %
2; 30 % \3
Silt/clay: 0–40 % 1; 40–80 % 2;
80 % \3
Forest: 0–20 % 3; 20–50 % 2;
50 % \1
Calculation: FF index =[(Class of
Slope 92) ?Class of
SiltClay ?(Class of Forest/2)]/3.5
Point-based calculation was used
by GIS using on classification
and formula
The climatic parameter affected by
climate change: temporal
frequency of extreme
precipitation events higher
30 mm was taken as the climatic
indicator
1. Tolerable: \1.43
2. Increased: 1.44–2.21
3. High: [2.22–(3.00)
Mass movement Most recent and significant landslide
events
Geological, mechanical type of the
sediment
Relief
Precipitation
Assessment: If the region is affected
by mass movements less than 5 % of
the total area, then a value of 1 is
given; if the affected area is between
5 and 25 %, then a value of 2 is
given; if the mass movements
occurred on more than 25 % of the
total area; then a value of 3 is given
Additive sensitivity calculation
based on observed actual mass
movements (Juha
´sz 2004; Fodor
and Kleb 1986) earlier events
was used
The climatic parameter affected by
climate change: the precipitation
sum of the winter season
1. Tolerable: \5%
2. Increased: 5–25 %
3. High: [25 %
Assessment of regional climate change impacts on Hungarian landscapes 805
123
and ALADIN have been used in our study at the meso-
regional level. These data on the climate change scenarios
have been employed by using single parameters on the
temperature, precipitation or extreme events and also by
typifying the four main categories of regional climate
change exposure by cluster analysis in the different
assessments on the regional sensitivities. The data scale
levels of the climate models and the regional landscape
units fit together well due to the data accuracy. The com-
bination is a step towards breaking down the climate
change exposure variations to a local-scale level.
Hungary is affected by several naturally driven envi-
ronmental hazards, which can modify the functioning and
the households of the regions and limit the use of resources.
The relevance of the different hazards is assessed in this
study by the sensitivity to the main problematic processes.
A balance or a resilient status of the processes, in terms of
the economy and society, relates mostly to damages of
material value. The present investigation used an envi-
ronmental hazard assessment perspective of nature-driven
hazards. To a certain extent, these hazards are also affected
by human factors.
In the following, we describe some of the outcomes of
the sensitivity change analysis from a large number of
results of the sensitivity assessments on soil erosion by
water, drought, wind, flash floods and mass movement. A
resilient status of the assessments and sensitivity changes
and the results calculated by the models using the thresh-
olds shown in Table 4for the three periods of investigation
are given in Table 5. A three-category system (tolerable,
increased, high) was used for an integration analysis;
therefore, 3?and 3?? were taken into consideration as 3
when the sensitivity indicators have strongly or very
strongly increased and when the critical limits of change
were reached.
The soil erosion hazards in response to climate change
are expected to affect similar areas in both of the time
periods of 2021–2050 and 2071–2100. The USLE calcu-
lation shows an increasing sensitivity in both periods. The
only significant difference is found in the period of
2021–2050 in the regions of the Dra
´va Plain, the Mecsek
Mountains and the Go
¨do
¨ll}
o Hills, where a decrease in rains
higher than 30 mm is expected in winter. The Fig. 3a–c
gives examples of the results for the soil erosion sensitivity
assessments for the actual situation and the two periods of
prediction for the A1B scenario.
The future change in the drought PaDI is also estimated
by referring to similar trends in the periods of 2021–2050
and 2071–2100, independent of using the REMO or AL-
ADIN data for the calculation. Only the south-western and
northern peripheries of Hungary might not face a prob-
lematic drying tendency. The results show that mainly the
Table 5 The sensitivity assessment results for the Hungarian meso-regions based on the REMO scenario (1–18 =name of micro-region; see
code in Table 3)
Meso-region indicator 1 2 345678 9101112 1314 15161718
Soil erosion
Basic 2 1 312322 22 3 1 1 1 1 2 3 1
2021_2050_REMO/ALADIN 3 1 313331 12 3 1 1 1 2 2 3 1
2071_2100_ REMO/ALADIN 3 1 313332 22 3 1 1 1 2 3 3 1
Drought
Basic 1 2 122113 23 1 3 1 3 2 2 1 3
2021_2050_REMO/ALADIN 1 3 112113 23 1 3?13?2213
2071_2100_REMO/ALADIN 2 3?122113?2313?? 13?? 2213
Wind_erosion
Basic 2 3 123233 22 2 3 1 2 1 2 1 2
2021_2050_REMO/ALADIN 1 3 113133 22 2 3?23 1212
2071_2100_REMO/ALADIN 2 3?123233?2223?? 33?1212
Flash_flood
Basic 2 1 312322 32 3 1 1 1 1 3 3 1
2021_2050_REMO/ALADIN 3 1 323331 22 3 1 1 1 2 2 3 1
2071_2100_REMO/ALADIN 3 1 313332 32 3 1 1 1 2 3 3 3
Mass movement
Basic 2 1 312322 32 3 1 1 1 1 2 3 1
2021_2050_REMO/ALADIN 3 1 323331 22 3 1 1 1 2 2 3 1
2071_2100_REMO/ALADIN 3 1 323332 32 3 1 1 1 2 3 3 1
Sensitivity classes 1 =tolerable, 2 =increased, 3 =high; 3?(3??) indicates that the regional sensitivity has strongly (very strongly)
increased
806 G. Mezo
¨si et al.
123
south-eastern part of the country is to expect serious prob-
lematic changes in the level of drought hazard. In the first
period, the hazard assessment class increases in the Danube-
Tisza Interfluves and in the loess and sandplain of Nyı
´rse
´gand
Hajdu
´sa
´g by one class, and in the later period, by two classes
towards problematic levels of aridity. The consequences are
well known; in theory, when droughts increase due to a
decreasing water supply, the groundwater table drops, agri-
cultural productivity declines and soil degradation occurs.
The wind erosion sensitivity follows the drought sensi-
tivity changes in the spatial distribution of Hungary, also
with an increasing sensitivity. This is mainly caused by the
soil and vegetation cover characteristics. We predicted a
continued increase in the sensitivity and the hazards in both
periods in two south-eastern regions, the Danube-Tisza
Interfluve and the Ko
¨ro
¨s-Maros Interfluve. A reduced
drying of the soils in the first period will decrease the
sensitivity in some western Hungarian regions.
A climate change-driven flash flood hazard increase can
be expected in the Transdanubian Hills and in the Northern
Mountains. The most sensitive areas are the regions of the
Zala Hills and the Transdanubian Hills. The most critical
areas are the region of the Zala River and the territory of
the Transdanubian Mountains. With this phenomenon, the
temporal frequency of the extreme precipitation events
above 30 mm was taken as the climatic indicator.
An increased sensitivity is also given for the Mecsek
Mountains and the Dra
´va plain; despite the medium relief
and high vertical fragmentation, the mass movements on
the slopes here are not frequent. Their increase is not
highly probable. Hailstorms and sudden floods are already
frequent in the area. Their frequency can increase due to
high-intensity precipitation. Concerning the mass-move-
ment hazard, those areas can be endangered by climate
change-driven processes where the lithological, morpho-
logical and hydrological preconditions for mass
Fig. 3 a Sensitivity to soil erosion by water (actual assessment). bSensitivity to soil erosion by water (Scenario A1B; 2021–2050). cSensitivity
to soil erosion by water (Scenario A1B; 2071–2100)
Assessment of regional climate change impacts on Hungarian landscapes 807
123
movements are given. The expected regional distribution
patterns are similar to flash floods.
An integrative analysis of the five sensitivity assess-
ments of the landscape functional hazards in Hungary due
to climate change is summarised in the Fig. 4a, b. The
figures show the expected climate change impact on two
levels of interpretation: (1) by the number of indicators
changed for the scenarios for Hungary for the periods of
2021–2050 and 2071–2100 compared to the 1961–1990
period (out of the maximal 5 investigated in this study);
and (2) the summarised changes in the hazard assessment of
the meso-regional sensitivity for the same periods. The latter
shows the increasing sensitivity on a scale from 5 (very low) to
15 (very high). The highest problematic increases in the pro-
cesses are found in the Marcal Basin and Koma
´rom plain in
the north-west of Hungary for the first (and second) period
when summarising all the sensitivity indicators. In this region,
the actual processes are very active.
Considering the results of our study, we conclude that
the vulnerability to climate change-induced natural process
changes may not pose as serious and sudden a risk to
human life as expected by Tobin and Montz (1997)
because of the generally slow changes in the land-use
system and the chance of adaptation activities of the
society. Szabo
´et al. (2008) calculated that, on smaller units
(landscape micro-region), the factors of potential environ-
mental hazards are expected to be lower in the north-
western part of Hungary compared to the south-eastern
part. The calculation of Szabo
´et al. (2008) was based on
investigations regarding a number of hazardous parame-
ters, including floods and droughts. The natural hazards
studied by this investigation give higher hazard predictions
in the regions of Tisza, Danube and around the Ko
¨ro
¨s
River. Because of the diversity of the applied methods and
a different time horizon of the predicted changes, it is
difficult to compare the results of our study with the study
of Szabo
´et al. (2008). Our study’s results are interpreted as
a step forward for a downscaling of climate change-
induced impacts. The sensitivity assessments are strongly
based on data and highly suitable for planning when
compared to the overall nonspatialised results of the
Hungarian VAHAVA-Project (Change, Impact, Reaction),
which analysed the effects of climate change on the envi-
ronment in general (Farago
´et al. 2010).
Flooding and excess inland water are also significant in
Hungary but have not been a subject of our study. The
primary causes of flooding are linked to the mountains
around the Carpathian basin. The interpretation of the
changing climate parameters, for example, the increase of
extreme precipitation events in the wintertime and of
anthropogenic factors (the changes in land use by defor-
estation, crop rotation, and urbanisation impacts), is fairly
determined. Based on the present state of knowledge,
several local sites of future excess inland water problems
can already be determined; however, their occurrence is
still interpreted as random with regards to the climate
(Pa
´lfai 2004; van Leeuwen et al. 2008; Rakonczai 2011).
Discussion and conclusion
The integrative methods example resulted in some inter-
esting outcomes when comparing the regional climatic
models of REMO and ALADIN, in general and for coarse-
grained meso-regional-scale levels of a number of land-
scapes in Hungary. The original climate data used in the
regional models feature a resolution of 25 925 km, and a
further downscaling to the Hungarian micro-region-scale
Fig. 4 a The number of hazard class changes in the indicators and the
additive sensitivity assessment of the meso-regional hazard for
2021–2050 scenarios compared to period 1961–1990 for Hungary.
bThe number of hazard class changes in the indicators and the
additive sensitivity assessment of the meso-regional hazard for
2071–2100 scenarios compared to period 1961–1990 for Hungary
808 G. Mezo
¨si et al.
123
level is not scientifically appropriate because of the lack of
locally accurate data. Further investigation should aim to
break down the climate change scenario modelling to local-
scale applications when following the scientific potential of
the results to link the climate change parameters of the
temperature, precipitation and related extreme events (heat
waves, drought periods, heavy rains and storms) to the
scale level of land use and biotope types as proxies of the
local ecosystems. This highlights the potential for the
methods of landscape functions analysis and assessments to
clarify the expected changes at the local-site-specific scale
using planned and applied measures or projects proposed
for climate change adaptation.
The aim of the study was not to apply our explorative
methodological approach to the entire set of climate change
scenarios provided by the IPCC (2007). The stability of the
climate change variables of the scenario A1B chosen in this
investigation is strongly linked to the societal success of the
general environmental policy (for example, by the further
activities following the Kyoto protocol) and to the unpre-
dictable societal and economic changes in land uses fol-
lowing the same time horizon until 2100. The goal of our
work lies in the regional differentiation of the expected
functional sensitivities and the usage of a qualitative
approach to clarify the hazards at an ordinal-data-scale level
(from low to high). The qualitative entrance to the assess-
ment is suitable, especially because of the diverse uncer-
tainties of the climate change regional modelling data and
prognosis. The regional environmental data and statistics of
diverse sources have been used. The intrinsic uncertainty
problems of the sensitivity assessment methods have not
been subjects of our study. The aggregation of the hetero-
geneous site-specific data to the sensitivities of regional
aggregates follows rule-based assessment methods.
The regional approach is suitable for general scenario
applications, for example, in water management, in nature
conservation policy (for example, for the management of
FFH and Habitat networks) or for agricultural and forest
policy and programme development. For Hungary, the
hazard assessment scenario is compared with the expecta-
tions, needs and outputs of VAHAVA (Szabo
´et al. 2008—
a programme of the Hungarian Academy of Sciences and
the Ministry for Rural Development dealing with the future
hazards of climate change for usage in governance issues.
The main users of the regional-scale methodology for mass
movements, soil erosion hazards by wind and water, flash
floods and droughts are seen in regional and local land-use
(landscape) planning, agricultural and water management
planning, and governmental applications to reduce societal
risks by reducing hazards in general. In these medium- or
long-term planning measures, the potential climate change
impacts will become integrated. The results might also be
valuable in the insurance sector.
The authors applied a set of simple methods, appropriate
largely on the basis of publicly available environmental
and social data. The aggregation method is simple because
of the methodological problems to integrate complex sys-
tems. The uncertainty of the meteorological and hydro-
logical models and the limited validity of rules-based
predictive modelling used in this study result in a differ-
entiated view of the sensitivity of Hungarian regions to
climate change. A detailed, in-depth analysis should be
applied to further break down the climate change impact to
the micro-regional scale of adaptation measurements.
The assessment results are integrated by a summation of
assessment points—but the results are obviously also use-
ful for sectoral indicator applications. From the scientific
viewpoint, intensive future investigations based on inte-
grative landscape and land-use models are required to
deepen knowledge regarding the interlinkages of landscape
maintenance for the investigated functions (for example,
between wind erosion and drought, water erosion and flash
floods) or the further development of methods for use with
single landscape functions (for example, by an enhanced
modelling of the processes of heavy rains and the hazards
of mass movements).
Generally, the authors have investigated only a chosen
set of landscape functions/landscape hazards by linking the
climate change parameters of the temperature and precip-
itation to problems of drought, erosion, mass movement
and flash flooding. Further investigations are seen in a
widening of the lists of functions or ecosystem functions
included in integrative methods and by a better reflection of
the multifunctionality of land use, including its fast and
dynamic changes. Future land-use changes may be more
dynamic and led by the strong dynamics of the socioeco-
nomic systems and the variable impacts of new technolo-
gies and markets. Such recent changes include the
conflicting new usages of the landscape in the context of
emerging bioenergy in concurrence with further-globalised
food markets. In this context, the problem of the resilience
of regional land-use systems is an open field of scientific
methods development and investigation, as this resilience
is strongly related to vulnerability and the adaptation
capacity. The methods for hazard assessment of landscape
functions or ecosystem services applied in our study pro-
vide results in a long-time perspective. Today, they are not
applicable for integration of the daily and seasonal
dynamics of weather and land-use systems in terms of
farming system models. A step forward in this field could
be, for example, the application of SWAT (soil water
assessment tool) in sensitivity and vulnerability assess-
ments when detailing the climate change impacts and the
hazards of the extreme events due to local catchments for
agricultural and water management purposes (Wisner et al.
1994).
Assessment of regional climate change impacts on Hungarian landscapes 809
123
Acknowledgments This publication is part of a research project
‘‘Creating the Centre of Excellence at the University of Szeged’’,
which is funded and supported by the European Union and co-
financed by the European Regional Development Fund (TA
´MOP-
4.2.1/B-09/1/KONV-2010).
References
Agrotopographical Database (1991) Budapest, http://www.mta-taki.
hu/en/departments/gis-lab/agrotopo_en
Bartholy J, Pongra
´cz R (2010) Analysis of precipitation conditions for
the Carpathian Basin based on extreme indices in the 20th
century and climate simulations for the 21st century. Phys Chem
Earth 35:43–51
Bartholy J, Pongra
´cz R, Barcza Z, Haszpra L, Gelybo
´G, Kern A,
Hidy D, Torma C, Hunyady A, Kardos P (2007) The regional
consequences of the climate change: present status and expected
trends. Fo
¨ldrajzi Ko
¨zleme
´nyek 55:257–269 (In Hungarian)
Bartholy J, Pongra
´cz R, Gelybo
´G, Szabo
´P (2008) Analysis of
expected climate change in the Carpathian Basin using the
PRUDENCE results. Id}
oja
´ra
´s 112:249–264 (In Hungarian)
Carpenter TM, Sperfslage JA, Georgakakos KP, Sweeney T, Fread
DL (1999) National threshold runoff estimation utilizing GIS is
support of operational flash flood warning systems. J Hydrol
224:21–44
Csima G, Hora
´nyi A (2008) Validation of the ALADIN—Climate
regional climate model at the Hungarian Meteorological Service.
Id}
oja
´ra
´s 112(3–4):155–177
Cziga
´ny S, Pirkhoffer E, Balassa B, Bugya T, Bo
¨tko
¨s T, Gyenizse P,
Nagyva
´radi L, Lo
´czy D, Geresdi I (2010) Flash floods as a
natural hazard in Southern Transdanubia. Fo
¨ldrajzi Ko
¨zleme
´-
nyek 134:281–298 (In Hungarian)
Estrela T, Mene
´ndez M, Dimas M, Marcuello C, Rees G, Cole G,
Weber K, Grath J, Leonard J, Ovesen NB, Fehe
´r J (2001) EEA
Sustainable water use in Europe. Part 3: Extreme hydrological
events: floods and droughts. http://www.eea.europa.eu/
publications/Environmental_Issues_No_21
European Environment Agency (2000) CORINE land cover technical
guide—Addendum 2000. EEA, Copenhagen
Farago
´T, La
´ng I, Csete L (eds) (2010) Climate change and
Hungary: mitigating the hazard and preparing for the impacts
(The ‘‘VAHAVA’’ Report). http://www.unisdr.org/files/18582_
thevahavareport08dec2010.pdf
Fodor T, Kleb B (1986) Engineering geological overview of Hungary.
MA
´FI, Budapest, p 199
Grundfest E, Rips A (2000) Flash floods. In: Parker DJ (ed) Floods
2002, vol 1. Routledge, London, pp 377–390
IPCC 2007 Climate Change 2007: The physical science basis.
Working group I contribution to the fourth assessment report of
the IPCC. In: Solomon S, Qin D, Manning M, Chen Z, Marquis
M, Averyt KB, Tignor M, Miller HL (eds) Cambridge University
Press, New York, http://www.ipcc.ch
Jentsch A, Beierkuhnlein C (2008) Research frontiers in climate
change: effects of extreme meteorological events on ecosystems.
CR Geosci 340:621–628
Juha
´sz A
´(2004) Mass movement endangering human settlements and
structures along the high bluff flanking Lake Balaton. Fo
¨ldrajzi
Ko
¨zleme
´nyek 52:19–30 (in Hungarian)
Kerte
´sz A
´, Centeri C (2006) Hungary. In: Poesen J, Boardman J (eds)
Soil erosion in Europe. Wiley, London, pp 139–153
Klik A (2004) Wind erosion assessment in Austria using wind erosion
equation and GIS. In: OECD: Agricultural impacts on soil
erosion and soil biodiversity: developing indicators for policy
analysis, Paris, www.oecd.org/tad/env/indicators
Loibl W, Aubrecht Ch (2011) Hungary Climate Regions: Regional
statistical analysis of selected Climate Change related indicators
and classification of physiogeographic Meso-Regions to Climate
Change region types. AIT – Austrian Institute of Technology
Vienna, p 52
Lo
´ki J (2011) Research of the land forming activity of wind and
protection against wind erosion in Hungary. Riscuri Si
Catastrofe 10:1–13 http://riscurisicatastrofe.reviste.ubbcluj.ro/
Volume/X_Nr_9_2011/PDF/J%20Loki.pdf
Martens P, Aerts JCJH, Amelung SB, Bouwer LM, Huynen M, Chang
CT, Ierland EC, Koppen CSA, McEvoy D, Mol APJ, Tatenhove
JPM (2010) Imagining the unimaginable: synthesis of essays on
abrupt and extreme climate change. Curr Opin Environ Sustain
2:347–355
Metzger M, Schro
¨ter D, Leemans R, Cramer W (2008) A spatially
explicit and quantitative vulnerability assessment of ecosystem
service change in Europe. Reg Environ Change 8:91–107
Meyer BC, Rannow S, Greiving S, Gruehn D (2009) Regionalisation
of climate change impacts in Germany for the usage in spatial
planning. GeoScape 1:34–43
Meyer BC, Rannow S, Loibl W (2010) Climate change and spatial
planning. Landsc Urban Plan 98:139–140
Mu
¨cher CA, Klijn JA, Wascher DM, Schamine
´e JHJ (2010) A new
European landscape classification (LANMAP): a transparent,
flexible and user-oriented methodology to distinguish land-
scapes. Ecol Ind 10:87–103
Opdam P, Washer D (2004) Climate change meets habitat fragmen-
tation: linking landscape and biogeographical scale levels in
research and conservation. Biol Conserv 117:285–297
Pa
´lfai I (2004) Undrained runoff and droughts in Hungary (Hydrologic
Studies). Ko
¨zDoc, Budapest p 492
Pe
´csi M, Somogyi S (1967) Magyarorsza
´g terme
´szeti fo
¨ldrajzi
ta
´jai e
´s geomorfolo
´giai ko
¨rzetei. Fo
¨ldrajzi. Ko
¨zleme
´nyek 15:
285–304
Pieczka I, Pongra
´c R, Bartholy J, Kis A, Miklo
´s E (2011) Estimated
extreme climatical data in the Region of Carpathian-basin using
results of the ENSEMBLES project. OMSZ, Budapest, pp 76–85
(in Hungarian)
Pielke RA Sr, Adegoke JO, Chase TN, Marshall CH, Matsui T,
Niyogi D (2007) A new paradigm for assessing the role of
agriculture in the climate system and in climate change. Agric
For Meteorol 142:234–254
Pongra
´cz R, Bartholy J, Szabo
´P, Gelybo
´G (2009) Comparison of
observed trends and simulated changes in extreme climate
indices in the Carpathian basin by the end of this century. Int J
Global Warm 1:336–355
Rakonczai J (2011) Effects and consequences of global climate
change in the Carpathian Basin. In: Blanco J, Kheradmand H
(eds) Climate change—geophysical foundations and ecological
effects, pp 297–322
Rannow S, Loibl W, Greiving S, Gruehn D, Meyer BC (2010)
Potential impacts of climate change in Germany—identifying
regional priorities for adaptation activities in spatial planning.
Landsc Urban Plan 98:160–171
Stefanovits P (1981) Pedology. Mez}
ogazdasa
´gi Kiado
´, Budapest,
p 376 (In Hungarian)
Szabo
´J, Lo
´ki J, To
´th C, Szabo
´G (2008) Natural hazards in Hungary.
In: Kerte
´sz A
´, Kova
´cs Z (eds) Dimensions and trends in
Hungarian geography. MTA FKI, Budapest, pp 55–68
Szabo
´P, Hora
´nyi A, Kru
¨zselyi I, Sze
´pszo
´G (2011) The climate
modelling at Hungarian Meteorological Survey: ALADIN and
REMO. OMSZ, Budapest, pp 87–101 (In Hungarian)
Sze
´pszo
´G (2008) Regional change of extreme characteristics over
Hungary based on different regional climate models of the
PRUDENCE project. Id}
oja
´ra
´s 112(3–4):265–284
810 G. Mezo
¨si et al.
123
Sze
´pszo
´G, Hora
´nyi A (2008) Transient simulation of the REMO
regional climate model and its evaluation over Hungary. Id}
oja
´ra
´s
112(3–4):203–231
Tobin GA, Montz BE (1997) Natural hazards: explanation and
integration. Guilford Publishing, New York, p 388
Usher M (2001) Landscape sensitivity: from theory to practice.
Catena 42:375–383
Van Leeuwen B, Tobak Z, Szatma
´ri J (2008) Development of an
integrated ANN—GIS framework for inland excess water
monitoring. J Environ Geogr 1:1–6
WEQ 2000—http://www.weru.ksu.edu/nrcs/weqguidance062503.doc
Wischmeier WH, Smith DD (1978) Predicting rainfall erosion losses:
a guide to conservation planning. Agriculture Handbook No.
537. USDA/Science and Education Administration, US. Govt.
Printing Office, Washington, p 58
Wisner B, Blaikie P, Cannon T, Davis I (1994) At risk: natural
hazards. People’s vulnerability and disasters. Routledge, New
York, p 447
Assessment of regional climate change impacts on Hungarian landscapes 811
123
