- Email: [email protected]
Climate variability in the Carpathian Mountains Region over 1961–2010
Global and Planetary Change 118 (2014) 85–96
Contents lists available at ScienceDirect
Global and Planetary Change journal homepage: www.elsevier.com/locate/gloplacha
Climate variability in the Carpathian Mountains Region over 1961–2010 Sorin Cheval, Marius-Victor Birsan ⁎, Alexandru Dumitrescu National Meteorological Administration (Meteo Romania), 97, Sos. Bucuresti-Ploiesti, 013686 Bucharest, Romania
a r t i c l e
i n f o
Article history: Received 14 October 2013 Received in revised form 19 February 2014 Accepted 22 April 2014 Available online 29 April 2014 Keywords: Carpathian Mountains climate change monthly trends maximum temperature wind speed relative humidity vapor pressure snow depth
a b s t r a c t The Carpathian Mountains Region (CMR) lies over parts of the territories of seven Central and Southeastern European countries, and the mountain chain induces major changes in the temperate climate specific to the latitudes between 43° and 49°N. Different administrations govern the long-term meteorological networks; the infrastructure, collection protocols, and storage capacities are specific to each country, so that a comprehensive study on the climate of the area has met considerable difficulties along time. Climate of the Carpathian Region (CARPATCLIM) is a regional initiative developed between 2010 and 2013 aiming to enhance the climatic information in the area by providing comprehensive, temporally and spatially homogenous data sets of the main meteorological variables. Based on daily data aggregated to a monthly scale at 10-km resolution, this study exploits and promotes the results of the CARPATCLIM project, documenting the variability of the main climatic variables over 1961–2010. For each month, the significant increasing or decreasing trends were identified, mapped and placed in the context of previous studies and climate change perspectives. The study has revealed several patterns in the climatic variability, i.e., positive or negative trends prevailing over the entire area, very distinct delineation between various trends induced by the Carpathian Mountain chain, and pledges for further scientific approaches, i.e., causes of the variability and applications in other domains. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The Carpathian Mountains Region (CMR) extends in Central and Southeastern Europe, over parts of the territories of the Czech Republic, Slovakia, Hungary, Poland, Ukraine, Romania, and Serbia. The rectangular-shaped frame of the CMR also includes some areas of Croatia, Bosnia and Herzegovina, and Austria despite the fact that they have no direct connection with the geography of the Carpathians. The mountain chain dominates the geographical context with altitudes exceeding 2500 m, but other forms of relief also contribute to the variety of the landscape, i.e., intra-mountainous depressions, sub-mountainous hills and lowlands. The CMR is a European biodiversity hotspot (Björnsen Gurung et al., 2009) which congregates major economical interests and derives social and political attention, demographic and land cover changes, and the climatic diagnosis and future projections have always made the scientific agenda. The latest IPCC report mentions significant climatic changes in the 2071–2100 perspective (Christensen et al., 2007), so that the climate monitoring services may have an authentic practical meaning in the CMR. The administrative, economic, and political changes that took place in the region along the recent history left behind a heterogeneous meteorological network, with ⁎ Corresponding author at: National Meteorological Administration (Meteo Romania), Department of Climatology, 97, Sos. Bucuresti-Ploiesti, 013686 Bucharest, Romania. Tel.: + 40 748 105 045 (GSM). E-mail addresses: [email protected] (S. Cheval), [email protected], [email protected] (M.-V. Birsan), [email protected], [email protected] (A. Dumitrescu).
http://dx.doi.org/10.1016/j.gloplacha.2014.04.005 0921-8181/© 2014 Elsevier B.V. All rights reserved.
numerous and significant changes between countries. The density of the national networks, the instruments, and the timing of the measurements and observations have been different, and they are not fully harmonized yet. A large number of studies, projects, and initiatives have tackled environmental issues and climate-related topics in the area at regional or country scale (Ruffini et al., 2006; UNEP, 2007; Villarini, 2011), but no valid description of the climate of the Carpathian Region has been available until very recently (Szalai, 2012). As a consequence, the outputs have been generally limited to local or country scales (Björnsen Gurung et al., 2009), with inherent difficulties whenever an integrated view was necessary. Climate of the Carpathian Region (CARPATCLIM) is one of the most recent international projects over the area, developed with the joint effort of National Meteorological Services (NMS) from all the Carpathian countries. It was contracted by the European Union represented by the Joint Research Centre (JRC), it ran between 2010 and 2013, and it aimed at enhancing the climatic information in the region by providing comprehensive, temporally and spatially homogenous, data sets of the main meteorological variables, and the corresponding metadata. The project was an excellent opportunity to address the inventory of the data sets available in each country, their gaps, quality, and homogeneity. Some project outputs have already been published recently (Lakatos et al., 2013; Spinoni et al., 2013), while many others are prepared to be turned to good account. This paper promotes and places the CARPATCLIM results in the context of similar reports on the topic, so that brief overviews are presented. The main objective of the study is to document the climatic
86
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
variability in the CMR, based on the parameters considered in the project, in order to identify significant trends in the present climate as a base for further in-depth investigations. To our knowledge, this is the first endeavor tackling the subject unitarily, at the scale of the Carpathian region, based on high quality, homogenous monthly data sets, aggregated and validated by an international consortium including all the NMSs in the area. Since many publications report on the future climate projections according to various IPCC scenarios, and the most common control period is 1961–1990, this study also represents a good opportunity to evaluate the track of the estimations, as it refers to the period 1961–2010.
2. Regional setting Fig. 1 shows the altitudes within the area of interest. There is no unanimous agreement regarding the geographical extension and margins of the CMR. Ruffini et al. (2006) report a surface of about 210,000 km2 for the Carpathian Ecoregion and propose principles for delimitating the Carpathian Convention area. UNEP (2007) places the limits at 43°28′ and 49°47′N; 16°58′ and 26°38′E, summing up an area of about 161,000 km2. Ruffini and Ptaček (2008) define a Carpathian Macroregion using administrative criteria. Following the JRC Tender Specification, the spatial area of interest for the CARPATCLIM project covered “the area between latitudes 50°N and 44°N, and longitudes 17°E and 27°E, approximately”. As regards the main sub-units, UNEP (2007) divides the CMR in Northwestern, Northeastern, Eastern, and Southeastern Carpathians, around the Pannonian and Transylvanian Depressions. The CMR has a temperate climate, with a rather continental regime, increasingly intensive eastwards. The Carpathians are more humid than the surrounding lowlands, and the average annual precipitation amounts register about 700–800 mm in the western parts, 350– 400 mm in the south, and 1000–1200 mm in the mountain area (Ruffini and Ptaček, 2008). The altitude, the compact arrangement and the shape of the Carpathian Chain introduce important disturbances in the zonal climate and in the general atmospheric circulation (UNEP, 2007). For the Romanian Carpathians, Cheval et al. (2011) report variations of the thermal vertical lapse according to the aspect, slope and land cover, and one can assume those are reflected in the local conditions and in the other meteorological variables, such as relative humidity, wind speed, and snow cover.
The land cover and hydrology have little influence on the regional climate of the CMR. Deciduous and conifer forests, steppe and cropland dominate the land cover, while the Danube River practically collects the waters of all the Carpathian hydrographical basins. 3. Materials and methods This study exploits the results of the project CARPATCLIM. Details about the objectives, methodology, metadata, input and output data of the project are easily available at www.carpatclim-eu.org (accessed on February 2014). Spinoni et al. (2013) describe the concepts and methods used for building a reliable climatologic database in the Carpathian Region. For the sake of brevity, we provide here only general information on the data and methodology. All the countries covering the CMR contributed to CARPATCLIM with relevant data and expertise, except for Bosnia and Herzegovina. Daily data sets from ground based meteorological stations, covering the period 1961–2010, refer to the following variables: maximum and minimum air temperature (Tmax and Tmin); precipitation amount (PP); average and maximum wind speed (WSave) and Wmax); sunshine duration (SD); cloud cover (CC); global radiation (GR); relative humidity (RH); air pressure (AP); water vapor pressure (VP); and snow depth (SwD). Within the project, each data set was quality controlled (QC), harmonized at the national borders and homogenized using the Multiple Analysis of Series for Homogenization (MASH v3.03) method and software (Szentimrey, 1999, 2008, 2011; Lakatos et al., 2013), and further combined at approximately 10-km resolution using the Meteorological Interpolation based on the Surface Homogenized Data Basis software (MISH v1.03; Szentimrey and Bihari, 2007). Homogenization was one of the main objectives of the CARPATCLIM project, so that it is worthy to present here the main details of the homogenization process and the quality of the resulting data set. MASH v3.03 was selected due to its demonstrated performance (Venema et al., 2012), and to the project partners' knowledge and skills. It is a relative homogenization method developed by Tamas Szentimrey from the Hungarian Meteorological Service, which makes no a priori assumption regarding the data homogeneity, and it uses an exhaustive searching scheme to detect and adjust the most probable breaks, and regime shifting points in the data series from each weather station. The distribution of the examined meteorological element is taken into account for using a multiplicative or additive model, depending on the
Fig. 1. Digital elevation model of the Carpathian Mountains Region.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
specific climate variable, while corrections are applied to the inhomogeneous series until no break is found. The homogenization of daily data uses the parameterization results obtained from monthly data homogenization (Szentimrey and Bihari, 2007; Szentimrey, 2011). Each country that homogenized the data within the national borders completed the data within 50 km-width buffers from the neighboring countries. The final project report on quality control and data homogenization measures applied per country, including QC protocols and measures to determine the achieved increase in data quality is available at www. carpatclim-eu.org/docs/deliverables/D1_12.pdf (accessed on 4 February 2014). A summary of the daily data sets, including their retrieval procedure, acronyms and station density information is presented in Table 1. The meteorological stations used for all the variables and their spatial distribution may be retrieved at www.carpatclim-eu.org/pages/atlas/ (accessed on 4 February 2014). For this study, the daily data corresponding to each pixel of the CMR domain were aggregated at a monthly scale. CARPATCLIM has delivered to the public an extended climatic database that covers large territories in the Central and Eastern Europe, and contains a significant number of variables, at fine temporal and spatial resolutions, quality controlled, harmonized at country borders and homogenized. The project's results can generate noteworthy benefits for various scientific domains and practitioners in the Carpathian region, but some data and methodological limitations have to be acknowledged as well. Factors like the irregular coverage with raw data, differences in the meteorological networks of the participating countries (i.e., time of observations and reports, sensors), and inherent model approximations within the data filling, homogenization and gridding procedures, may introduce errors in the results. The non-parametric Mann–Kendall (MK) test statistics sustained the local trend significance examination for each variable (Mann, 1945; Kendall, 1975). The MK test is a rank-based procedure, particularly suitable for non-normally distributed data which contain outliers and non-linear trends (Salas, 1993). In this study, the significance level was 10% (two-tailed test). The slope estimate was conducted with the nonparametric Kendall–Theil method (also known as the Theil–Sen slope estimate) which is suitable for nearly linear trends and is less affected by non-normal data and outliers (Helsel and Hirsch, 1992). Pearson (r) and Spearman (rho) correlation coefficients (p b 0.01) between the Theil–Sen slope values and elevation were computed for all the variables which showed an altitudinal distribution of the trends. It has to be mentioned that the correlation values should be regarded in the context of the geographical background of the correlated variables, namely significant orographic fragmentation, and different synoptic influences on each Carpathian branch. The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend and slope N 5%.
87
4. Results and discussions An overview of the trend results is presented in Table 2. Figs. 2–10 show the monthly variability of each parameter, at a 10% level of statistical significance, revealing the increase, decrease and no trend areas, as well as the abrupt slopes (above 5%). While the climatic signal is mostly consistent over the area, results suggest that the Carpathian chain has an important role in the climate of the region, with variations from one parameter to the other. The particular characteristics of the CARPATCLIM database (e.g., spatial and temporal resolution, particular methodologies for QC, homogenization and gridding) make very difficult to compare the outputs with other studies or databases. Nevertheless, the results have already been checked for their reliability with previous publications, mainly qualitatively. 4.1. Air temperature Except for autumn, the maximum and minimum air temperature is either increasing or remaining stable. The most extended increase occurs in the late spring and summer months (May–August), covering the CMR almost completely (Fig. 2). A similar situation happens in January, when the Tmax has no significant trend yet along the Carpathian Mountains. However, it is interesting to mention that the inner depressions (Pannonian and Transylvanian) have similar positive tendencies as the low regions outside the Carpathian Chain, probably, as a result of the frequent thermal inversions occurring in this month. The correlation between the Theil–Sen slope of Tmax and elevation may be considered important in January and February (r = − 0.4) with a very high statistical significance (p b 0.01). Large territories from the western part of the CMR (e.g., most of Hungary, Slovakia) have significant positive Tmax tendency in April, while the regions outside the Carpathian Chain show an increasing pattern in February and March, with certain similarities for Tmin. A comparable seasonality of the trends was reported for Serbia by Božanić and Gasperič (2010), who reported an increase a bit higher than 0.4 °C/decade in the northern part of the country, including the area covered by the Carpathians. Analyzing the mean maximum temperature in Romania from 1960 to 1998, Tomozeiu et al. (2002) found significant and steep increasing trends for winter and summer, and a decreasing trend for autumn. Obviously, the first decade of the 21st century was warm enough to determine the shift of the autumn trend from negative to stable, as reflected in the CARPATCLIM results. Besides, Ioniţă et al. (2012) noted that warmer summers have become more frequent after 1990. Proxy information, such as increased timberline elevation and the replacement of coniferous forests with mixed ones, confirms the warming in the Southern Carpathians (Mihai et al., 2007). Based on “very reliable time series of daily data” from three stations between 115 and 2635 m, Lapin et al. (2005) remarked very significant increasing air temperatures in the high mountains of Slovakia in the April to August season after 1990. Concerning the future projections, based on long term series, Brázdil
Table 1 Meteorological variables, units, retrieval methodology, acronyms, and number of stations used. Meteorological variable
Unit
Methodology for retrieving the primary data (daily)
Acronym
Total number of stations and density (as km2/station)
Maximum air temperature Minimum air temperature Precipitation amount Average wind speed Maximum wind speed Cloud cover Sunshine duration Global radiation Relative humidity Surface air pressure Surface vapor pressure Snow depth
°C °C mm m/s m/s % hours MJ/m2 % hPa hPa cm
Standard daily measurements Standard daily measurements Standard daily measurements Computed from 3-hour measurements Computed from 3-hour measurements Computed from 3-hour measurements Standard daily measurements Computed from relative daily sunshine duration measurements and extraterrestrial radiation Standard daily measurements Standard daily measurements Computed from daily relative humidity and saturation vapor pressure Computed from daily grids of mean air temperature, precipitation and relative air humidity
Tmax Tmin PP VV Vmax CC SunD GR RH AP VP SD
542; 27.6 542; 27.6 1165; 19.6 510; 28.2 510; 28.2 498; 28.4 330; 36.3 333; 36.2 501; 28.6 405; 33.2 568; 27.4 N/A
88
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
et al. (2009) predict a warming trend in the Czech Republic, and Hirschi et al. (2011) found that a significant increase in the hot extremes is more likely in the Southeastern Europe, including the CMR. As for Romania, Busuioc et al. (2010) argued that under the A1B IPCC scenario the annual mean temperature will increase by 1.4 °C (±0.4 °C) for the period 2021–2050, and by 3.1 °C (±0.7 °C) for the period 2070–2100, relative to 1961–1990. The frequency and magnitude of temperature extremes will also increase.
periods for 2071–2100, compared to the 1961–1990 control period. The precipitation is very likely to decrease over the Romanian area of the CMR in summer (Busuioc et al., 2010). The monthly temperature and precipitation trend maps realized by the European Climate Assessment & Datasets (ECA&D; http://eca. knmi.nl/utils/mapserver/trend.php#bottom, accessed on February 2014) reveal for the CMR similar patterns as those retrieved by the CARPATCLIM project, although they report that for the period 1951– 2012, the homogenization procedure differs, and considerably fewer meteorological stations are considered.
4.2. Precipitation The Mann–Kendall statistics captured irrelevant patterns for the precipitation trends in the CMR, with randomly distributed spots of increasing or decreasing regime. One can remark some more spatial consistency only in September, with large increasing spots mainly inside the Carpathian Chain (Transylvanian Depression), and in October, with the same pattern, but outside the mountainous range (Fig. 3). Significant correlations were found between the September precipitation Theil–Sen slope and elevation; the correlation coefficients are relevant in the context of the study (r = 0.34; rho = 0.31). The analysis of 60 years of data (1949–2009) revealed comparable precipitation seasonality in Serbia, with the most visible increasing trend in autumn and less coherent variability in the other seasons (Božanić and Gasperič, 2010). A slight boost of the annual precipitation is mentioned for the mountains of Serbia, while the lower spots have become drier (Spasov et al., 2002). Kyselỳ (2009) found a cut-off between western and eastern parts of the Czech Republic in many precipitation characteristics, noted that the pattern of changes is gradually more inconsistent to the Eastern Europe. Brázdil et al. (2009) used data from 23 climatological stations, homogenized by means of Standard Normal Homogeneity Test (SNHT) (Alexandersson, 1986), and argued that lower precipitation amounts occurred in the Czech Republic. Domonkos (2003) emphasized a drier climate in Hungary, with 15–20% decrease in the annual precipitation total during the 20th century, while Hirschi et al. (2011) reached a similar conclusion for Southeastern Europe, based on the gridded E-Obs data set (Haylock et al., 2008), and highdensity station data from Austria, Czech Republic, Romania, and Bulgaria aggregated in the CECILIA project (www.cecilia-eu.org, accessed on February 2014). Tomozeiu et al. (2005) reported a decrease of the winter precipitation in Romania (1961–1996), partially – but firmly – confirmed by the CARPATCLIM results for February. As for the future precipitation in the CMR, Kundzewicz et al. (2006), and Christensen et al. (2007) assume decreasing amounts and longer dry
4.3. Wind Two wind variables were investigated in terms of their temporal variability and the possible territorial outline, namely the average wind speed WSave and the maximum wind value (WSmax). The results show a decrease in wind speed in all months, confirming that the terrestrial stilling is a globally spread phenomenon (McVicar et al., 2012). It is acknowledged that the wind speed is decreasing due to the increase in roughness of many terrains; however, one can remark that the VV decreasing signal is consistently spread over the entire CMR (Fig. 4), suggesting that larger mechanisms are more important than the local roughness. The WSave decreasing signal is consistently spread over the entire CMR (Fig. 4). A very similar pattern is valid for the WSmax (not shown). There are several months, i.e., December, January and August, when the Carpathian Chain establishes a fairly clear distinction between the ‘no trend’ western parts, and the ‘negative trend’ oriental regions. Previous findings show similar conclusions over Hungary (Tar et al., 2001), Czech Republic (Brázdil et al., 2009), and Romania (Vespremeanu-Stroe et al., 2012; Birsan et al., 2013a, 2013b), and it is likely that the decreasing trend of the average wind speed will maintain in the CMR through 2071–2011 (Christensen et al., 2007). Nevertheless, there are studies indicating that more intense wind events can be expected over the Central Europe in the next decades (Leckebusch et al., 2008). 4.4. Cloud cover, sunshine duration, and global radiation Generally, one can assume that cloudiness, sunshine duration and global radiation are reasonably correlated, as are the corresponding temporal trends. Except for some limited spots, the CC's variability emphasizes no relevant trend in most of the months, whereas the SD and GR are more or less increasing (Figs. 5–7). Excepting the Northwestern
Table 2 Summary of the trend results: percentages of significant trends at 10% level (two-tail test) for each parameter. Values below 5% are omitted. Variable
Up/Down
Tmax Tmin PP VV Vmax CC
Up Up Up Down Down Up Down Up Down Up Down Up Down Up Down Up Down
Significant trends as percentage of total grid cells (%) Dec
SunD GR RH AP VP SD
5.8 7.4
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
75.8 75.4
25.5
17.7 9.7
26.2
94.3 59.4
97.1 87.7
100 100
100 100
40.9 85.0
43.2 78.3 6.1
15.7 6.3
26.6
11.5
46.3 79.3
47.2 74.7
91.9 95.8
74.2 88.9
69.4 88.5
65.8 92.3
64.7 94.8
13.6 36.4
27.8
24.8
12.4 45.3
24.1
10.0
27.2
10.1
16.5
65.3
27.2
16.9 6.7 14.9
54.3
63.1
22.8 28.7
Oct
39.6 33.1 75.7 59.0
28.1 28.6 50.0 77.3 62.8
27.8
7.1
10.0
7.2
16.4
19.4 15.2
29.0 32.5
90.9 50.1
11.5 42.2
69.1
57.3
99.5
99.6
33.0
11.6
63.5 96.8
18.5
7.5 18.5 11.0
Nov
16.9 17.5
6.1 6.0
22.9
15.2
42.7
Sep
27.0
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
89
Fig. 2. Maximum temperature trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Fig. 3. Monthly precipitation trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend with slope N5%; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
90
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 4. Mean wind speed trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Fig. 5. Cloud cover trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
91
Fig. 6. Sunshine duration trends in the Carpathian Mountains Region (1961–2010).
and Northeastern Carpathians, in September and October, the cloud cover seems to follow an increasing trend over the area of interest (Fig. 5), and the global radiation is manifestly decreasing in the regions outside the Carpathian Chain (Fig. 7). Filipiak and Mietus (2009) confirm the light change signal in the cloud cover variability in Poland.
part of the year, with the highest values between January and March (r = 0.55; rho = 0.5). For the Slovakian part of the Carpathians, Lapin et al. (2005) remarked a decrease in the RH during the warm season (April to August) after 1990, while Brázdil et al. (2009) reported the decrease of the annual RH values over the Czech Republic.
4.6. Air pressure and water vapor pressure 4.5. Relative humidity The variability of the monthly relative humidity is unsystematic over the CMR, with spots of increasing or decreasing trends hardly grouped in consistent territorial patterns (Fig. 8). The positive trends apparently focus around the Carpathian Chain in December–January, and September–October, while negative trends virtually dominate the Pannonian and Transylvanian Depressions in February, and March. Moreover, the correlation coefficients between the Theil–Sen slope and elevations are quite important and highly significant (p b 0.01) in the most
December and April are the only 2 months with noticeable increasing trends in air pressure, with ‘no trend’ identified for the rest of the year (not shown). Brázdil et al. (2009) found no significant variations in the western part of the CMR for 1961–2005. Moreover, there are very little changes expected in the sea level pressure of the area until 2071–2100 (Christensen et al., 2007). Fig. 9 shows that the water vapor pressure has been either stable or increasing over the CMR throughout 1961–2010. The widest increase can be remarked in January and in May–August, probably as a result
92
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 7. Global radiation trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
of both large-scale processes and warming, and with little orographic interference. 4.7. Snow The snow cover trends in mountain regions of Europe are generally characterized by large regional and altitudinal variations (Lemke et al., 2007). In the CMR, snow depth values are decreasing over quite many areas, mainly in January–February (Fig. 10), but also in December (very few spots, mainly in the Northwestern Carpathians) and March. As regards the trends, variations between the mountainous chain and the lower territories can be observed, and they reflect into the high correlation coefficients between the Theil–Sen slopes and elevations (r from − 0.46 to − 0.54; rho from − 0.55 to − 0.67). Vojtek et al. (2003) reported a general decrease of the snow cover duration and solid precipitation in the Slovakian Carpathians along the period 1960–2000, but increasing trends above 2600 m. The drier and warmer
climate over 1950–2000 led to less snow in Serbia (Spasov et al., 2002), while in Romania a decreased frequency of the snowfalls defines the period 1961–2003 (Micu, 2009). The SwD decline resulted from the CARPATCLIM project is consistent with the findings of Birsan and Dumitrescu (2014) for Romania, very likely under the influence of the Atlantic large-scale circulation, but also due to the temperature increase, which led to an increase of the rain/snow ratio – which directly affects the hydrological regime during winter and spring (Birsan et al., 2012, 2013b). At the same time, the positive tendency of the snow cover mentioned by Falarz (2004) for the mountain regions in Poland has become indistinguishable while extending the analysis to 2010, very probably as a result of the warmer climate of the last decade. 5. Conclusions This study identifies the direction and magnitude of the decadal climate variability in different parts of the Carpathians, responding to one
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
93
Fig. 8. Relative humidity trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend with slope N5%. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
of the main scientific needs in the area (Kozak et al., 2011). The paper investigates the monthly climate trends in the Carpathian region in a unitary manner, covering the whole area, using high quality homogenous data sets, and tackling comprehensively the main climate variables. In brief, this study is based on the best and most complete available data sets in the Carpathian region, and it provides an overall perspective on the climate variability over the area in the last decades. The CARPATCLIM project (2010–2013) resulted in a flexible and robust database, very useful for climate researchers and stakeholders, addressing the quality control, border harmonization, homogenization, and gridding issues in a unitary manner, at feasible temporal (daily) and spatial resolutions (cca. 10 km), which is one of the primary exploitations of the results. We have overviewed the monthly trend of the main climatic variables over the Carpathian Mountain Region, throughout 1961–2010, as a background for more in-depth approaches, and we have provided very basic assumptions regarding the possible causes or relationships. This paper demonstrates that the climate of the CMR has an active
variability, very well emphasized at monthly scale for variables like air temperature, wind, sunshine duration, and relative humidity, and less distinct for others. The results complete and update previous investigations on climate variability in the area and emphasize the role of the permanent climate monitoring as a measure for preventing associate threats like changes in the heavy rainfall events, heat waves, or snow cover. The trends of each meteorological element were classified as significant ‘positive’, ‘negative’, or ‘relatively stable’ based on the Mann– Kendall statistics, at a 90% significance level, in each pixel of the CMR. The analysis has revealed the following situations: (1) ‘positive’ or ‘negative’ trend prevails over the largest area of the region; e.g., Tmax in May–August; WS in April; CC in October; and VP in July and August; (2) ‘positive’ or ‘negative’ trend is largely extended either over the inner or the outer parts of the Carpathian Chain, while the other parts have no definite trend; e.g., Tmax in February and
94
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 9. Water vapor pressure trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
(3)
(4)
(5)
(6)
(7)
March; PP in October; WS in December and January; and RG in September and October; discrete emphasize of the Carpathian Chain comparing to the surrounding areas, either with significant trend or ‘no trend’; e.g., Tmax in January; and RH in January; apparently random trends in most situations; e.g., Tmax in December; Tmin in January and October; PP in March–August; WS in May; CC in January–February; and GR in May; no statistical trends over the CMR; e.g., Tmax in autumn; Tmin in November; PP in November and December; CC in most of the months; GR in December and January; decreasing and increasing trends do not occur in parallel over large territories in the same month; the very few such situations refer to rather small areas; e.g., SD in many months; RH in December, February, and April; except for the GR and RH, either decreasing or increasing trend characterizes one variable over the whole year; e.g., increasing
for Tmax, Tmin, PP, CC, SD, AP, and VP; decreasing for WSave, WSmax, and SwD; (8) due to their specific characteristics, some variables exhibit different tendencies, well separated between the higher areas and lower territories; it is the case of maximum temperature in January, and snow cover in March, for example. It has to be mentioned that a systematic comparison of the outputs reported in the specialized literature tackling the area and topic, must consider all the specific methodologies applied (e.g., homogeneity, missing data volume accepted, time interval, variables, and metadata). However, there is quite a qualitative agreement of this study's results with previous publications, even if there are many differences in the data sets used, period of reference and methodology, indicating that proper quality of the climatic databases and adequate expertise are in place in the CMR countries.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
95
Fig. 10. Snow depth trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend with slope N5%. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Ultimately, this paper challenges a few directions for the climate change studies in the Carpathian region. There is a need to decipher the mechanisms controlling the variability of each element. Teleconnections, mainly the North Atlantic Oscillation, have often been linked to the variations in the atmospheric variables in the CMR (e.g., Tomozeiu et al., 2005; Brázdil et al., 2009), but other causes should certainly be considered. Based on the results presented in this paper, one can figure out that the interactions between the general atmospheric circulation and the major relief (Carpathian Mountains) run the climate variations in the CMR. In some cases, the general atmospheric circulation overcomes the orography (e.g., Tmax in May–August, Tmin in June–August, and water vapor pressure in July–August), while the role of the relief prevails in other situations (e.g., Tmax in February and March, precipitation in October). Finer temporal scales may be useful for various applications, i.e., extreme indices built upon daily data relevant for agriculture, drought or flood control; the CARPATCLIM database can eventually support such developments.
Acknowledgments We thank the two anonymous reviewers for their comments and suggestions which led to an overall improvement of the original manuscript. This study used data from the CARPATCLIM Database (Szalai et al 2013), available online at carpatclim-eu.org, © European Commission – Joint Research Center, 2013. We acknowledge the members of the CARPATCLIM consortium for their contribution in developing proper methodologies, and for their significant role in the data collection and computation of the gridded variables.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gloplacha.2014.04.005. References Alexandersson, H., 1986. A homogeneity test applied to precipitation data. J. Climatol. 6, 661–675. http://dx.doi.org/10.1002/joc.3370060607. Birsan, M.-V., Dumitrescu, A., 2014. Snow variability in Romania in connection to largescale atmospheric circulation. Int. J. Climatol. 34 (1), 134–144. http://dx.doi.org/10. 1002/joc.3671. Birsan, M.V., Zaharia, L., Chendes, V., Branescu, E., 2012. Recent trends in streamflow in Romania (1976–2005). Rom. Rep. Phys. 64 (1), 275–280. Birsan, M.V., Marin, L., Dumitrescu, A., 2013a. Seasonal trends in wind speed in Romania. Rom. Rep. Phys. 65 (4), 1479–1484. Birsan, M.V., Zaharia, L., Chendes, V., Branescu, E., 2013b. Seasonal trends in Romanian streamflow. Hydrol. Process. http://dx.doi.org/10.1002/hyp.9961. Björnsen Gurung, A., Bokwa, A., Chełmicki, W., Elbakidze, M., Hirschmugl, M., Hostert, P., Ibisch, P., Kozak, J., Kuemmerle, T., Matei, E., Ostapowicz, K., Pociask-Karteczka, J., Schmidt, L., van der Linden, S., Zebisch, M., 2009. Global change research in the Carpathian Mountain Region. Mt. Res. Dev. 28 (3), 282–288. http://dx.doi.org/10. 1659/mrd.1105. Božanić, D., Gasperič, M. (Eds.), 2010. Initial National Communication of the Republic of Serbia Under the United Nations Framework Convention on Climate Change. Ministry of Environment and Spatial Planning, Belgrade (151 pp.). Brázdil, R., Trnka, M., Dobrovolný, P., Chromá, K., Hlavinka, P., Žalud, Z., 2009. Variability of droughts in the Czech Republic, 1881–2006. Theor. Appl. Climatol. 97, 297–315. http://dx.doi.org/10.1007/s00704-008-0065-x. Busuioc, A., Caian, M., Cheval, S., Bojariu, R., Boroneant, C., Baciu, M., Dumitrescu, A., 2010. Variability and Climate Change in Romania. Pro Universitaria, Bucharest (226 pp., in Romanian). Cheval, S., Dumitrescu, A., Petrişor, A.I., 2011. The July Surface Temperature Lapse in the Romanian Carpathians. Carpath. J. Earth Environ. 6, 189–198. Christensen, J.H., Hewitson, B., Busuioc, A., Chen, A., Gao, X., Held, I., Jones, R., Kolli, R.K., Kwon, W.T., Laprise, R., Magaña Rueda, V., Mearns, L., Menéndez, C.G., Räisänen, J., Rinke, A., Sarr, A., Whetton, P., 2007. Regional climate projections. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.
96
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge (United Kingdom and New York, NY, USA). Domonkos, P., 2003. Recent precipitation trends in Hungary in the context of larger scale climatic changes. Nat. Hazards 29, 255–271. Falarz, M., 2004. Variability and trends in the duration and depth of snow cover in Poland in the 20th century. Int. J. Climatol. 24 (13), 1713–1727. http://dx.doi.org/10.1002/ joc.1505. Filipiak, J., Mietus, M., 2009. Spatial and temporal variability of cloudiness in Poland, 1971–2000. Int. J. Climatol. 29, 1294–1311. http://dx.doi.org/10.1002/joc.1777. Haylock, M.R., et al., 2008. A European daily high-resolution gridded dataset of surface temperature and precipitation for 1950–2006. J. Geophys. Res. 113, D20119. Helsel, D.R., Hirsch, R.M., 1992. Statistical Methods in Water Resources. Elsevier. Hirschi, M., Seneviratne, S.I., Alexandrov, V., Boberg, F., Boroneant, C., Christensen, O.B., Formayer, H., Orlowsky, B., Stepanek, P., 2011. Observational evidence for soilmoisture impact on hot extremes in southeastern Europe. Nat. Geosci. 4, 17–21. http://dx.doi.org/10.1038/NGEO1032. Ioniţă, M., Rîmbu, N., Chelcea, S., Pătruţ, S., 2012. Multidecadal variability of summer temperature over Romania and its relation with Atlantic Multidecadal Oscillation. Theor. Appl. Climatol. http://dx.doi.org/10.1007/s00704-012-0786-8. Kendall, M.G., 1975. Rank Correlation Methods. Charles Griffin, London. Kozak, J.., Björnsen Gurung, A., Ostapowicz, K. (Eds.), 2011. Research Agenda for the Carpathians: 2010–2015. Kraków (http://www.forumcarpaticum.org/FC-main/ Download/Research_agenda_for_the_Carpathians.pdf). Kundzewicz, Z.W., Radziejewski, M., Pinskwar, I., 2006. Precipitation extremes in the changing climate of Europe. Clim. Res. 31, 51–58. Kyselỳ, J., 2009. Trends in heavy precipitation in the Czech Republic over 1961–2005. Int. J. Climatol. 29, 1745–1758. http://dx.doi.org/10.1002/joc.1784. Lakatos, M., Szentimrey, T., Bihari, Z., Szalai, S., 2013. Creation of a homogenized climate database for the Carpathian region by applying the MASH procedure and the preliminary analysis of the data. Időjárás 117 (1), 143–158. Lapin, M., Štastny, P., Chmelík, M., 2005. Detection of climate change in the Slovak Mountains. Croat. Meteorol. J. 40 (40), 101–104. Leckebusch, G.C., Weimer, A., Pinto, J.G., Reyers, M., Speth, P., 2008. Extreme wind storms over Europe in present and future climate: a cluster analysis approach. Meteorol. Z. 17 (1), 067–082. http://dx.doi.org/10.1127/0941-2948/2008/0266. Lemke, P., Ren, J., Alley, R.B., Allison, I., Carrasco, J., Flato, G., Fujii, Y., Kaser, G., Mote, P., Thomas, R.H., Zhang, T., 2007. Observations: changes in snow, ice and frozen ground. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge (United Kingdom and New York, NY). Mann, H.B., 1945. Nonparametric tests against trend. Econometrica 13, 245–259. McVicar, T.R., Roderick, M.L., Donohue, R.J., Li, L.T., Van Niel, T.G., Thomas, A., Grieser, J., Jhajharia, D., Himri, Y., Mahowald, N.M., Mescherskaya, An.V., Kruger, A.C., Rehman, S., Dinpashoh, Y., 2012. Global review and synthesis of trends in observed terrestrial near-surface wind speeds: implications for evaporation. J. Hydrol. 416-417, 182–205. http://dx.doi.org/10.1016/j.jhydrol.2011.10.024. Micu, D., 2009. Snow pack in the Romanian Carpathians under changing climatic conditions. Meteorol. Atmos. Phys. 105 (1–2), 1–16. http://dx.doi.org/10.1007/s00703009-0035-6. Mihai, B., Savulescu, I., Sandric, I., 2007. Change detection analysis (1986–2002) of vegetation cover in Romania. A study of alpine, subalpine, and forest landscapes in the Iezer Mountains, Southern Carpathians. Mt. Res. Dev. 27 (3), 250–258. http://dx. doi.org/10.1659/mred.0645.
Ruffini, F.V., Ptaček, P. (Eds.), 2008. Atlas of the Carpathian Macroregion. EURAC Research Bolzano & Department of Geography, Palacky University of Olomouc (62 pp.). Ruffini, F.V., Streifeneder, T., Eiselt, B., 2006. Implementing an international mountain convention. An approach for the delimitation of the Carpathian Convention areaEuropean Academy, Bolzano/Bozen (119 pp.). Salas, J.D., 1993. Analysis and modeling of hydrologic time series. Chapter 19 in In: Maidment, D.R. (Ed.), Handbook of Hydrology, McGraw Hill. Spasov, P., Spasov, D., Petrović, P., 2002. Changes in drought occurrences in Serbia. Proceedings of International Conference on Drought Mitigation and Prevention of Land Desertification'. Bled, Slovenia, 21-25 April 2002 (CD-ROM). International Commission on Irrigation and Drainage (ICID), Ljubljana, Slovenia. Spinoni, J., Antofie, T., Barbosa, P., Bihari, Z., Lakatos, M., Szalai, S., Szentimrey, T., Vogt, J., 2013. An overview of drought events in the Carpathian Region in 1961–2010. Adv. Sci. Res. 10, 21–32. http://dx.doi.org/10.5194/asr-10-21-2013. Szalai, S., 2012. CARPATCLIM: A Project for Climate Atlas with Data Harmonization. Balwois. http://balwois.com/2012/USB/papers/355.pdf (accessed: February 2014). Szentimrey, T., 1999. Multiple Analysis of Series for Homogenization (MASH). Proceedings of the 2nd Seminar for Homogenization of Surface Climatological Data. WMO, Budapest, Hungary, pp. 27–46 (WCDMP-No. 41). Szentimrey, T., 2008. Development of MASH homogenization procedure for daily data. Proceedings of the Fifth Seminar for Homogenization and Quality Control in Climatological Databases, Budapest, Hungary, 2006, WCDMP-No. 71, pp. 123–130. Szentimrey, T., 2011. Manual of Homogenization Software MASHv3.03, Hungarian Meteorological Service. p. 64. Szentimrey, T., Bihari, Z., 2007. Mathematical Background of the Spatial Interpolation Methods and the Software MISH (Meteorological Interpolation based on Surface Homogenized Data Basis). Proceedings from the Conference on Spatial Interpolation in Climatology and Meteorology, Budapest, Hungary, 2004, COST Action, 719. COST Office, pp. 17–27. Tar, K., Makra, L., Horváth, S., Kircsi, A., 2001. Temporal change of some statistical characteristics of wind speed over the Great Hungarian Plain. Theor. Appl. Climatol. 69, 69–79. http://dx.doi.org/10.1007/s007040170035. Tomozeiu, R., Busuioc, A., Stefan, S., 2002. Changes in seasonal mean of maximum air temperature in Romania and their connection with large-scale circulation. Int. J. Climatol. 22 (10), 1181–1196. http://dx.doi.org/10.1002/joc.785. Tomozeiu, R., Stefan, S., Busuioc, A., 2005. Winter precipitation variability and large-scale circulation patterns in Romania. Theor. Appl. Climatol. 81 (3-4), 193–201. http://dx. doi.org/10.1007/s00704-004-0082-3. UNEP, 2007. Carpathian Environment Outlook. United Nations Environment Programme (UNEP), Geneva (236 pp.). Venema, V.K.C., Mestre, O., Aguilar, E., Auer, I., Guijarro, J.A., Domonkos, P., Vertacnik, G., Szentimrey, T., Stepanek, P., Zahradnicek, P., Viarre, J., Muller-Westermeier, G., Lakatos, M., Williams, C.N., Menne, M., Lindau, R., Rasol, D., Rustemeier, E., Kolokythas, K., Marinova, T., Andresen, L., Acquaotta, F., Fratianni, S., Cheval, S., Klancar, M., Brunetti, M., Gruber, C., Prohom Duran, M., Likso, T., Esteban, P., Brandsma, T., 2012. Benchmarking homogenization algorithms for monthly data. Clim. Past 8, 89–115. http://dx.doi.org/10.5194/cp-8-89-2012. Vespremeanu-Stroe, A., Cheval, S., Tătui, D., 2012. The wind regime of Romania—characteristics, trends and North Atlantic Oscillation influences. Forum Geogr. XI (2), 118–126. http://dx.doi.org/10.5775/fg.2067-4635.2012.003.d. Villarini, G., 2011. Analyses of annual and seasonal maximum daily rainfall accumulations for Ukraine, Moldova, and Romania. Int. J. Climatol. 32 (14), 2213–2226. http://dx.doi. org/10.1002/joc.3394. Vojtek, M., Fasko, P., Stastny, P., 2003. Some selected snow climate trends in Slovakia with respect to altitude. Acta Meteorol. Univ. Comenianae 32, 17–27.
Contents lists available at ScienceDirect
Global and Planetary Change journal homepage: www.elsevier.com/locate/gloplacha
Climate variability in the Carpathian Mountains Region over 1961–2010 Sorin Cheval, Marius-Victor Birsan ⁎, Alexandru Dumitrescu National Meteorological Administration (Meteo Romania), 97, Sos. Bucuresti-Ploiesti, 013686 Bucharest, Romania
a r t i c l e
i n f o
Article history: Received 14 October 2013 Received in revised form 19 February 2014 Accepted 22 April 2014 Available online 29 April 2014 Keywords: Carpathian Mountains climate change monthly trends maximum temperature wind speed relative humidity vapor pressure snow depth
a b s t r a c t The Carpathian Mountains Region (CMR) lies over parts of the territories of seven Central and Southeastern European countries, and the mountain chain induces major changes in the temperate climate specific to the latitudes between 43° and 49°N. Different administrations govern the long-term meteorological networks; the infrastructure, collection protocols, and storage capacities are specific to each country, so that a comprehensive study on the climate of the area has met considerable difficulties along time. Climate of the Carpathian Region (CARPATCLIM) is a regional initiative developed between 2010 and 2013 aiming to enhance the climatic information in the area by providing comprehensive, temporally and spatially homogenous data sets of the main meteorological variables. Based on daily data aggregated to a monthly scale at 10-km resolution, this study exploits and promotes the results of the CARPATCLIM project, documenting the variability of the main climatic variables over 1961–2010. For each month, the significant increasing or decreasing trends were identified, mapped and placed in the context of previous studies and climate change perspectives. The study has revealed several patterns in the climatic variability, i.e., positive or negative trends prevailing over the entire area, very distinct delineation between various trends induced by the Carpathian Mountain chain, and pledges for further scientific approaches, i.e., causes of the variability and applications in other domains. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The Carpathian Mountains Region (CMR) extends in Central and Southeastern Europe, over parts of the territories of the Czech Republic, Slovakia, Hungary, Poland, Ukraine, Romania, and Serbia. The rectangular-shaped frame of the CMR also includes some areas of Croatia, Bosnia and Herzegovina, and Austria despite the fact that they have no direct connection with the geography of the Carpathians. The mountain chain dominates the geographical context with altitudes exceeding 2500 m, but other forms of relief also contribute to the variety of the landscape, i.e., intra-mountainous depressions, sub-mountainous hills and lowlands. The CMR is a European biodiversity hotspot (Björnsen Gurung et al., 2009) which congregates major economical interests and derives social and political attention, demographic and land cover changes, and the climatic diagnosis and future projections have always made the scientific agenda. The latest IPCC report mentions significant climatic changes in the 2071–2100 perspective (Christensen et al., 2007), so that the climate monitoring services may have an authentic practical meaning in the CMR. The administrative, economic, and political changes that took place in the region along the recent history left behind a heterogeneous meteorological network, with ⁎ Corresponding author at: National Meteorological Administration (Meteo Romania), Department of Climatology, 97, Sos. Bucuresti-Ploiesti, 013686 Bucharest, Romania. Tel.: + 40 748 105 045 (GSM). E-mail addresses: [email protected] (S. Cheval), [email protected], [email protected] (M.-V. Birsan), [email protected], [email protected] (A. Dumitrescu).
http://dx.doi.org/10.1016/j.gloplacha.2014.04.005 0921-8181/© 2014 Elsevier B.V. All rights reserved.
numerous and significant changes between countries. The density of the national networks, the instruments, and the timing of the measurements and observations have been different, and they are not fully harmonized yet. A large number of studies, projects, and initiatives have tackled environmental issues and climate-related topics in the area at regional or country scale (Ruffini et al., 2006; UNEP, 2007; Villarini, 2011), but no valid description of the climate of the Carpathian Region has been available until very recently (Szalai, 2012). As a consequence, the outputs have been generally limited to local or country scales (Björnsen Gurung et al., 2009), with inherent difficulties whenever an integrated view was necessary. Climate of the Carpathian Region (CARPATCLIM) is one of the most recent international projects over the area, developed with the joint effort of National Meteorological Services (NMS) from all the Carpathian countries. It was contracted by the European Union represented by the Joint Research Centre (JRC), it ran between 2010 and 2013, and it aimed at enhancing the climatic information in the region by providing comprehensive, temporally and spatially homogenous, data sets of the main meteorological variables, and the corresponding metadata. The project was an excellent opportunity to address the inventory of the data sets available in each country, their gaps, quality, and homogeneity. Some project outputs have already been published recently (Lakatos et al., 2013; Spinoni et al., 2013), while many others are prepared to be turned to good account. This paper promotes and places the CARPATCLIM results in the context of similar reports on the topic, so that brief overviews are presented. The main objective of the study is to document the climatic
86
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
variability in the CMR, based on the parameters considered in the project, in order to identify significant trends in the present climate as a base for further in-depth investigations. To our knowledge, this is the first endeavor tackling the subject unitarily, at the scale of the Carpathian region, based on high quality, homogenous monthly data sets, aggregated and validated by an international consortium including all the NMSs in the area. Since many publications report on the future climate projections according to various IPCC scenarios, and the most common control period is 1961–1990, this study also represents a good opportunity to evaluate the track of the estimations, as it refers to the period 1961–2010.
2. Regional setting Fig. 1 shows the altitudes within the area of interest. There is no unanimous agreement regarding the geographical extension and margins of the CMR. Ruffini et al. (2006) report a surface of about 210,000 km2 for the Carpathian Ecoregion and propose principles for delimitating the Carpathian Convention area. UNEP (2007) places the limits at 43°28′ and 49°47′N; 16°58′ and 26°38′E, summing up an area of about 161,000 km2. Ruffini and Ptaček (2008) define a Carpathian Macroregion using administrative criteria. Following the JRC Tender Specification, the spatial area of interest for the CARPATCLIM project covered “the area between latitudes 50°N and 44°N, and longitudes 17°E and 27°E, approximately”. As regards the main sub-units, UNEP (2007) divides the CMR in Northwestern, Northeastern, Eastern, and Southeastern Carpathians, around the Pannonian and Transylvanian Depressions. The CMR has a temperate climate, with a rather continental regime, increasingly intensive eastwards. The Carpathians are more humid than the surrounding lowlands, and the average annual precipitation amounts register about 700–800 mm in the western parts, 350– 400 mm in the south, and 1000–1200 mm in the mountain area (Ruffini and Ptaček, 2008). The altitude, the compact arrangement and the shape of the Carpathian Chain introduce important disturbances in the zonal climate and in the general atmospheric circulation (UNEP, 2007). For the Romanian Carpathians, Cheval et al. (2011) report variations of the thermal vertical lapse according to the aspect, slope and land cover, and one can assume those are reflected in the local conditions and in the other meteorological variables, such as relative humidity, wind speed, and snow cover.
The land cover and hydrology have little influence on the regional climate of the CMR. Deciduous and conifer forests, steppe and cropland dominate the land cover, while the Danube River practically collects the waters of all the Carpathian hydrographical basins. 3. Materials and methods This study exploits the results of the project CARPATCLIM. Details about the objectives, methodology, metadata, input and output data of the project are easily available at www.carpatclim-eu.org (accessed on February 2014). Spinoni et al. (2013) describe the concepts and methods used for building a reliable climatologic database in the Carpathian Region. For the sake of brevity, we provide here only general information on the data and methodology. All the countries covering the CMR contributed to CARPATCLIM with relevant data and expertise, except for Bosnia and Herzegovina. Daily data sets from ground based meteorological stations, covering the period 1961–2010, refer to the following variables: maximum and minimum air temperature (Tmax and Tmin); precipitation amount (PP); average and maximum wind speed (WSave) and Wmax); sunshine duration (SD); cloud cover (CC); global radiation (GR); relative humidity (RH); air pressure (AP); water vapor pressure (VP); and snow depth (SwD). Within the project, each data set was quality controlled (QC), harmonized at the national borders and homogenized using the Multiple Analysis of Series for Homogenization (MASH v3.03) method and software (Szentimrey, 1999, 2008, 2011; Lakatos et al., 2013), and further combined at approximately 10-km resolution using the Meteorological Interpolation based on the Surface Homogenized Data Basis software (MISH v1.03; Szentimrey and Bihari, 2007). Homogenization was one of the main objectives of the CARPATCLIM project, so that it is worthy to present here the main details of the homogenization process and the quality of the resulting data set. MASH v3.03 was selected due to its demonstrated performance (Venema et al., 2012), and to the project partners' knowledge and skills. It is a relative homogenization method developed by Tamas Szentimrey from the Hungarian Meteorological Service, which makes no a priori assumption regarding the data homogeneity, and it uses an exhaustive searching scheme to detect and adjust the most probable breaks, and regime shifting points in the data series from each weather station. The distribution of the examined meteorological element is taken into account for using a multiplicative or additive model, depending on the
Fig. 1. Digital elevation model of the Carpathian Mountains Region.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
specific climate variable, while corrections are applied to the inhomogeneous series until no break is found. The homogenization of daily data uses the parameterization results obtained from monthly data homogenization (Szentimrey and Bihari, 2007; Szentimrey, 2011). Each country that homogenized the data within the national borders completed the data within 50 km-width buffers from the neighboring countries. The final project report on quality control and data homogenization measures applied per country, including QC protocols and measures to determine the achieved increase in data quality is available at www. carpatclim-eu.org/docs/deliverables/D1_12.pdf (accessed on 4 February 2014). A summary of the daily data sets, including their retrieval procedure, acronyms and station density information is presented in Table 1. The meteorological stations used for all the variables and their spatial distribution may be retrieved at www.carpatclim-eu.org/pages/atlas/ (accessed on 4 February 2014). For this study, the daily data corresponding to each pixel of the CMR domain were aggregated at a monthly scale. CARPATCLIM has delivered to the public an extended climatic database that covers large territories in the Central and Eastern Europe, and contains a significant number of variables, at fine temporal and spatial resolutions, quality controlled, harmonized at country borders and homogenized. The project's results can generate noteworthy benefits for various scientific domains and practitioners in the Carpathian region, but some data and methodological limitations have to be acknowledged as well. Factors like the irregular coverage with raw data, differences in the meteorological networks of the participating countries (i.e., time of observations and reports, sensors), and inherent model approximations within the data filling, homogenization and gridding procedures, may introduce errors in the results. The non-parametric Mann–Kendall (MK) test statistics sustained the local trend significance examination for each variable (Mann, 1945; Kendall, 1975). The MK test is a rank-based procedure, particularly suitable for non-normally distributed data which contain outliers and non-linear trends (Salas, 1993). In this study, the significance level was 10% (two-tailed test). The slope estimate was conducted with the nonparametric Kendall–Theil method (also known as the Theil–Sen slope estimate) which is suitable for nearly linear trends and is less affected by non-normal data and outliers (Helsel and Hirsch, 1992). Pearson (r) and Spearman (rho) correlation coefficients (p b 0.01) between the Theil–Sen slope values and elevation were computed for all the variables which showed an altitudinal distribution of the trends. It has to be mentioned that the correlation values should be regarded in the context of the geographical background of the correlated variables, namely significant orographic fragmentation, and different synoptic influences on each Carpathian branch. The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend and slope N 5%.
87
4. Results and discussions An overview of the trend results is presented in Table 2. Figs. 2–10 show the monthly variability of each parameter, at a 10% level of statistical significance, revealing the increase, decrease and no trend areas, as well as the abrupt slopes (above 5%). While the climatic signal is mostly consistent over the area, results suggest that the Carpathian chain has an important role in the climate of the region, with variations from one parameter to the other. The particular characteristics of the CARPATCLIM database (e.g., spatial and temporal resolution, particular methodologies for QC, homogenization and gridding) make very difficult to compare the outputs with other studies or databases. Nevertheless, the results have already been checked for their reliability with previous publications, mainly qualitatively. 4.1. Air temperature Except for autumn, the maximum and minimum air temperature is either increasing or remaining stable. The most extended increase occurs in the late spring and summer months (May–August), covering the CMR almost completely (Fig. 2). A similar situation happens in January, when the Tmax has no significant trend yet along the Carpathian Mountains. However, it is interesting to mention that the inner depressions (Pannonian and Transylvanian) have similar positive tendencies as the low regions outside the Carpathian Chain, probably, as a result of the frequent thermal inversions occurring in this month. The correlation between the Theil–Sen slope of Tmax and elevation may be considered important in January and February (r = − 0.4) with a very high statistical significance (p b 0.01). Large territories from the western part of the CMR (e.g., most of Hungary, Slovakia) have significant positive Tmax tendency in April, while the regions outside the Carpathian Chain show an increasing pattern in February and March, with certain similarities for Tmin. A comparable seasonality of the trends was reported for Serbia by Božanić and Gasperič (2010), who reported an increase a bit higher than 0.4 °C/decade in the northern part of the country, including the area covered by the Carpathians. Analyzing the mean maximum temperature in Romania from 1960 to 1998, Tomozeiu et al. (2002) found significant and steep increasing trends for winter and summer, and a decreasing trend for autumn. Obviously, the first decade of the 21st century was warm enough to determine the shift of the autumn trend from negative to stable, as reflected in the CARPATCLIM results. Besides, Ioniţă et al. (2012) noted that warmer summers have become more frequent after 1990. Proxy information, such as increased timberline elevation and the replacement of coniferous forests with mixed ones, confirms the warming in the Southern Carpathians (Mihai et al., 2007). Based on “very reliable time series of daily data” from three stations between 115 and 2635 m, Lapin et al. (2005) remarked very significant increasing air temperatures in the high mountains of Slovakia in the April to August season after 1990. Concerning the future projections, based on long term series, Brázdil
Table 1 Meteorological variables, units, retrieval methodology, acronyms, and number of stations used. Meteorological variable
Unit
Methodology for retrieving the primary data (daily)
Acronym
Total number of stations and density (as km2/station)
Maximum air temperature Minimum air temperature Precipitation amount Average wind speed Maximum wind speed Cloud cover Sunshine duration Global radiation Relative humidity Surface air pressure Surface vapor pressure Snow depth
°C °C mm m/s m/s % hours MJ/m2 % hPa hPa cm
Standard daily measurements Standard daily measurements Standard daily measurements Computed from 3-hour measurements Computed from 3-hour measurements Computed from 3-hour measurements Standard daily measurements Computed from relative daily sunshine duration measurements and extraterrestrial radiation Standard daily measurements Standard daily measurements Computed from daily relative humidity and saturation vapor pressure Computed from daily grids of mean air temperature, precipitation and relative air humidity
Tmax Tmin PP VV Vmax CC SunD GR RH AP VP SD
542; 27.6 542; 27.6 1165; 19.6 510; 28.2 510; 28.2 498; 28.4 330; 36.3 333; 36.2 501; 28.6 405; 33.2 568; 27.4 N/A
88
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
et al. (2009) predict a warming trend in the Czech Republic, and Hirschi et al. (2011) found that a significant increase in the hot extremes is more likely in the Southeastern Europe, including the CMR. As for Romania, Busuioc et al. (2010) argued that under the A1B IPCC scenario the annual mean temperature will increase by 1.4 °C (±0.4 °C) for the period 2021–2050, and by 3.1 °C (±0.7 °C) for the period 2070–2100, relative to 1961–1990. The frequency and magnitude of temperature extremes will also increase.
periods for 2071–2100, compared to the 1961–1990 control period. The precipitation is very likely to decrease over the Romanian area of the CMR in summer (Busuioc et al., 2010). The monthly temperature and precipitation trend maps realized by the European Climate Assessment & Datasets (ECA&D; http://eca. knmi.nl/utils/mapserver/trend.php#bottom, accessed on February 2014) reveal for the CMR similar patterns as those retrieved by the CARPATCLIM project, although they report that for the period 1951– 2012, the homogenization procedure differs, and considerably fewer meteorological stations are considered.
4.2. Precipitation The Mann–Kendall statistics captured irrelevant patterns for the precipitation trends in the CMR, with randomly distributed spots of increasing or decreasing regime. One can remark some more spatial consistency only in September, with large increasing spots mainly inside the Carpathian Chain (Transylvanian Depression), and in October, with the same pattern, but outside the mountainous range (Fig. 3). Significant correlations were found between the September precipitation Theil–Sen slope and elevation; the correlation coefficients are relevant in the context of the study (r = 0.34; rho = 0.31). The analysis of 60 years of data (1949–2009) revealed comparable precipitation seasonality in Serbia, with the most visible increasing trend in autumn and less coherent variability in the other seasons (Božanić and Gasperič, 2010). A slight boost of the annual precipitation is mentioned for the mountains of Serbia, while the lower spots have become drier (Spasov et al., 2002). Kyselỳ (2009) found a cut-off between western and eastern parts of the Czech Republic in many precipitation characteristics, noted that the pattern of changes is gradually more inconsistent to the Eastern Europe. Brázdil et al. (2009) used data from 23 climatological stations, homogenized by means of Standard Normal Homogeneity Test (SNHT) (Alexandersson, 1986), and argued that lower precipitation amounts occurred in the Czech Republic. Domonkos (2003) emphasized a drier climate in Hungary, with 15–20% decrease in the annual precipitation total during the 20th century, while Hirschi et al. (2011) reached a similar conclusion for Southeastern Europe, based on the gridded E-Obs data set (Haylock et al., 2008), and highdensity station data from Austria, Czech Republic, Romania, and Bulgaria aggregated in the CECILIA project (www.cecilia-eu.org, accessed on February 2014). Tomozeiu et al. (2005) reported a decrease of the winter precipitation in Romania (1961–1996), partially – but firmly – confirmed by the CARPATCLIM results for February. As for the future precipitation in the CMR, Kundzewicz et al. (2006), and Christensen et al. (2007) assume decreasing amounts and longer dry
4.3. Wind Two wind variables were investigated in terms of their temporal variability and the possible territorial outline, namely the average wind speed WSave and the maximum wind value (WSmax). The results show a decrease in wind speed in all months, confirming that the terrestrial stilling is a globally spread phenomenon (McVicar et al., 2012). It is acknowledged that the wind speed is decreasing due to the increase in roughness of many terrains; however, one can remark that the VV decreasing signal is consistently spread over the entire CMR (Fig. 4), suggesting that larger mechanisms are more important than the local roughness. The WSave decreasing signal is consistently spread over the entire CMR (Fig. 4). A very similar pattern is valid for the WSmax (not shown). There are several months, i.e., December, January and August, when the Carpathian Chain establishes a fairly clear distinction between the ‘no trend’ western parts, and the ‘negative trend’ oriental regions. Previous findings show similar conclusions over Hungary (Tar et al., 2001), Czech Republic (Brázdil et al., 2009), and Romania (Vespremeanu-Stroe et al., 2012; Birsan et al., 2013a, 2013b), and it is likely that the decreasing trend of the average wind speed will maintain in the CMR through 2071–2011 (Christensen et al., 2007). Nevertheless, there are studies indicating that more intense wind events can be expected over the Central Europe in the next decades (Leckebusch et al., 2008). 4.4. Cloud cover, sunshine duration, and global radiation Generally, one can assume that cloudiness, sunshine duration and global radiation are reasonably correlated, as are the corresponding temporal trends. Except for some limited spots, the CC's variability emphasizes no relevant trend in most of the months, whereas the SD and GR are more or less increasing (Figs. 5–7). Excepting the Northwestern
Table 2 Summary of the trend results: percentages of significant trends at 10% level (two-tail test) for each parameter. Values below 5% are omitted. Variable
Up/Down
Tmax Tmin PP VV Vmax CC
Up Up Up Down Down Up Down Up Down Up Down Up Down Up Down Up Down
Significant trends as percentage of total grid cells (%) Dec
SunD GR RH AP VP SD
5.8 7.4
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
75.8 75.4
25.5
17.7 9.7
26.2
94.3 59.4
97.1 87.7
100 100
100 100
40.9 85.0
43.2 78.3 6.1
15.7 6.3
26.6
11.5
46.3 79.3
47.2 74.7
91.9 95.8
74.2 88.9
69.4 88.5
65.8 92.3
64.7 94.8
13.6 36.4
27.8
24.8
12.4 45.3
24.1
10.0
27.2
10.1
16.5
65.3
27.2
16.9 6.7 14.9
54.3
63.1
22.8 28.7
Oct
39.6 33.1 75.7 59.0
28.1 28.6 50.0 77.3 62.8
27.8
7.1
10.0
7.2
16.4
19.4 15.2
29.0 32.5
90.9 50.1
11.5 42.2
69.1
57.3
99.5
99.6
33.0
11.6
63.5 96.8
18.5
7.5 18.5 11.0
Nov
16.9 17.5
6.1 6.0
22.9
15.2
42.7
Sep
27.0
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
89
Fig. 2. Maximum temperature trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Fig. 3. Monthly precipitation trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend with slope N5%; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
90
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 4. Mean wind speed trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Fig. 5. Cloud cover trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
91
Fig. 6. Sunshine duration trends in the Carpathian Mountains Region (1961–2010).
and Northeastern Carpathians, in September and October, the cloud cover seems to follow an increasing trend over the area of interest (Fig. 5), and the global radiation is manifestly decreasing in the regions outside the Carpathian Chain (Fig. 7). Filipiak and Mietus (2009) confirm the light change signal in the cloud cover variability in Poland.
part of the year, with the highest values between January and March (r = 0.55; rho = 0.5). For the Slovakian part of the Carpathians, Lapin et al. (2005) remarked a decrease in the RH during the warm season (April to August) after 1990, while Brázdil et al. (2009) reported the decrease of the annual RH values over the Czech Republic.
4.6. Air pressure and water vapor pressure 4.5. Relative humidity The variability of the monthly relative humidity is unsystematic over the CMR, with spots of increasing or decreasing trends hardly grouped in consistent territorial patterns (Fig. 8). The positive trends apparently focus around the Carpathian Chain in December–January, and September–October, while negative trends virtually dominate the Pannonian and Transylvanian Depressions in February, and March. Moreover, the correlation coefficients between the Theil–Sen slope and elevations are quite important and highly significant (p b 0.01) in the most
December and April are the only 2 months with noticeable increasing trends in air pressure, with ‘no trend’ identified for the rest of the year (not shown). Brázdil et al. (2009) found no significant variations in the western part of the CMR for 1961–2005. Moreover, there are very little changes expected in the sea level pressure of the area until 2071–2100 (Christensen et al., 2007). Fig. 9 shows that the water vapor pressure has been either stable or increasing over the CMR throughout 1961–2010. The widest increase can be remarked in January and in May–August, probably as a result
92
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 7. Global radiation trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
of both large-scale processes and warming, and with little orographic interference. 4.7. Snow The snow cover trends in mountain regions of Europe are generally characterized by large regional and altitudinal variations (Lemke et al., 2007). In the CMR, snow depth values are decreasing over quite many areas, mainly in January–February (Fig. 10), but also in December (very few spots, mainly in the Northwestern Carpathians) and March. As regards the trends, variations between the mountainous chain and the lower territories can be observed, and they reflect into the high correlation coefficients between the Theil–Sen slopes and elevations (r from − 0.46 to − 0.54; rho from − 0.55 to − 0.67). Vojtek et al. (2003) reported a general decrease of the snow cover duration and solid precipitation in the Slovakian Carpathians along the period 1960–2000, but increasing trends above 2600 m. The drier and warmer
climate over 1950–2000 led to less snow in Serbia (Spasov et al., 2002), while in Romania a decreased frequency of the snowfalls defines the period 1961–2003 (Micu, 2009). The SwD decline resulted from the CARPATCLIM project is consistent with the findings of Birsan and Dumitrescu (2014) for Romania, very likely under the influence of the Atlantic large-scale circulation, but also due to the temperature increase, which led to an increase of the rain/snow ratio – which directly affects the hydrological regime during winter and spring (Birsan et al., 2012, 2013b). At the same time, the positive tendency of the snow cover mentioned by Falarz (2004) for the mountain regions in Poland has become indistinguishable while extending the analysis to 2010, very probably as a result of the warmer climate of the last decade. 5. Conclusions This study identifies the direction and magnitude of the decadal climate variability in different parts of the Carpathians, responding to one
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
93
Fig. 8. Relative humidity trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend with slope N5%. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
of the main scientific needs in the area (Kozak et al., 2011). The paper investigates the monthly climate trends in the Carpathian region in a unitary manner, covering the whole area, using high quality homogenous data sets, and tackling comprehensively the main climate variables. In brief, this study is based on the best and most complete available data sets in the Carpathian region, and it provides an overall perspective on the climate variability over the area in the last decades. The CARPATCLIM project (2010–2013) resulted in a flexible and robust database, very useful for climate researchers and stakeholders, addressing the quality control, border harmonization, homogenization, and gridding issues in a unitary manner, at feasible temporal (daily) and spatial resolutions (cca. 10 km), which is one of the primary exploitations of the results. We have overviewed the monthly trend of the main climatic variables over the Carpathian Mountain Region, throughout 1961–2010, as a background for more in-depth approaches, and we have provided very basic assumptions regarding the possible causes or relationships. This paper demonstrates that the climate of the CMR has an active
variability, very well emphasized at monthly scale for variables like air temperature, wind, sunshine duration, and relative humidity, and less distinct for others. The results complete and update previous investigations on climate variability in the area and emphasize the role of the permanent climate monitoring as a measure for preventing associate threats like changes in the heavy rainfall events, heat waves, or snow cover. The trends of each meteorological element were classified as significant ‘positive’, ‘negative’, or ‘relatively stable’ based on the Mann– Kendall statistics, at a 90% significance level, in each pixel of the CMR. The analysis has revealed the following situations: (1) ‘positive’ or ‘negative’ trend prevails over the largest area of the region; e.g., Tmax in May–August; WS in April; CC in October; and VP in July and August; (2) ‘positive’ or ‘negative’ trend is largely extended either over the inner or the outer parts of the Carpathian Chain, while the other parts have no definite trend; e.g., Tmax in February and
94
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
Fig. 9. Water vapor pressure trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: red – significant increasing trend and slope N5%; light red – significant increasing trend; white – non-significant trend. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
(3)
(4)
(5)
(6)
(7)
March; PP in October; WS in December and January; and RG in September and October; discrete emphasize of the Carpathian Chain comparing to the surrounding areas, either with significant trend or ‘no trend’; e.g., Tmax in January; and RH in January; apparently random trends in most situations; e.g., Tmax in December; Tmin in January and October; PP in March–August; WS in May; CC in January–February; and GR in May; no statistical trends over the CMR; e.g., Tmax in autumn; Tmin in November; PP in November and December; CC in most of the months; GR in December and January; decreasing and increasing trends do not occur in parallel over large territories in the same month; the very few such situations refer to rather small areas; e.g., SD in many months; RH in December, February, and April; except for the GR and RH, either decreasing or increasing trend characterizes one variable over the whole year; e.g., increasing
for Tmax, Tmin, PP, CC, SD, AP, and VP; decreasing for WSave, WSmax, and SwD; (8) due to their specific characteristics, some variables exhibit different tendencies, well separated between the higher areas and lower territories; it is the case of maximum temperature in January, and snow cover in March, for example. It has to be mentioned that a systematic comparison of the outputs reported in the specialized literature tackling the area and topic, must consider all the specific methodologies applied (e.g., homogeneity, missing data volume accepted, time interval, variables, and metadata). However, there is quite a qualitative agreement of this study's results with previous publications, even if there are many differences in the data sets used, period of reference and methodology, indicating that proper quality of the climatic databases and adequate expertise are in place in the CMR countries.
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
95
Fig. 10. Snow depth trends in the Carpathian Mountains Region (1961–2010). The colors on the maps indicate the following: white – non-significant trend; blue – significant decreasing trend; deep blue – significant decreasing trend with slope N5%. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.
Ultimately, this paper challenges a few directions for the climate change studies in the Carpathian region. There is a need to decipher the mechanisms controlling the variability of each element. Teleconnections, mainly the North Atlantic Oscillation, have often been linked to the variations in the atmospheric variables in the CMR (e.g., Tomozeiu et al., 2005; Brázdil et al., 2009), but other causes should certainly be considered. Based on the results presented in this paper, one can figure out that the interactions between the general atmospheric circulation and the major relief (Carpathian Mountains) run the climate variations in the CMR. In some cases, the general atmospheric circulation overcomes the orography (e.g., Tmax in May–August, Tmin in June–August, and water vapor pressure in July–August), while the role of the relief prevails in other situations (e.g., Tmax in February and March, precipitation in October). Finer temporal scales may be useful for various applications, i.e., extreme indices built upon daily data relevant for agriculture, drought or flood control; the CARPATCLIM database can eventually support such developments.
Acknowledgments We thank the two anonymous reviewers for their comments and suggestions which led to an overall improvement of the original manuscript. This study used data from the CARPATCLIM Database (Szalai et al 2013), available online at carpatclim-eu.org, © European Commission – Joint Research Center, 2013. We acknowledge the members of the CARPATCLIM consortium for their contribution in developing proper methodologies, and for their significant role in the data collection and computation of the gridded variables.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gloplacha.2014.04.005. References Alexandersson, H., 1986. A homogeneity test applied to precipitation data. J. Climatol. 6, 661–675. http://dx.doi.org/10.1002/joc.3370060607. Birsan, M.-V., Dumitrescu, A., 2014. Snow variability in Romania in connection to largescale atmospheric circulation. Int. J. Climatol. 34 (1), 134–144. http://dx.doi.org/10. 1002/joc.3671. Birsan, M.V., Zaharia, L., Chendes, V., Branescu, E., 2012. Recent trends in streamflow in Romania (1976–2005). Rom. Rep. Phys. 64 (1), 275–280. Birsan, M.V., Marin, L., Dumitrescu, A., 2013a. Seasonal trends in wind speed in Romania. Rom. Rep. Phys. 65 (4), 1479–1484. Birsan, M.V., Zaharia, L., Chendes, V., Branescu, E., 2013b. Seasonal trends in Romanian streamflow. Hydrol. Process. http://dx.doi.org/10.1002/hyp.9961. Björnsen Gurung, A., Bokwa, A., Chełmicki, W., Elbakidze, M., Hirschmugl, M., Hostert, P., Ibisch, P., Kozak, J., Kuemmerle, T., Matei, E., Ostapowicz, K., Pociask-Karteczka, J., Schmidt, L., van der Linden, S., Zebisch, M., 2009. Global change research in the Carpathian Mountain Region. Mt. Res. Dev. 28 (3), 282–288. http://dx.doi.org/10. 1659/mrd.1105. Božanić, D., Gasperič, M. (Eds.), 2010. Initial National Communication of the Republic of Serbia Under the United Nations Framework Convention on Climate Change. Ministry of Environment and Spatial Planning, Belgrade (151 pp.). Brázdil, R., Trnka, M., Dobrovolný, P., Chromá, K., Hlavinka, P., Žalud, Z., 2009. Variability of droughts in the Czech Republic, 1881–2006. Theor. Appl. Climatol. 97, 297–315. http://dx.doi.org/10.1007/s00704-008-0065-x. Busuioc, A., Caian, M., Cheval, S., Bojariu, R., Boroneant, C., Baciu, M., Dumitrescu, A., 2010. Variability and Climate Change in Romania. Pro Universitaria, Bucharest (226 pp., in Romanian). Cheval, S., Dumitrescu, A., Petrişor, A.I., 2011. The July Surface Temperature Lapse in the Romanian Carpathians. Carpath. J. Earth Environ. 6, 189–198. Christensen, J.H., Hewitson, B., Busuioc, A., Chen, A., Gao, X., Held, I., Jones, R., Kolli, R.K., Kwon, W.T., Laprise, R., Magaña Rueda, V., Mearns, L., Menéndez, C.G., Räisänen, J., Rinke, A., Sarr, A., Whetton, P., 2007. Regional climate projections. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.
96
S. Cheval et al. / Global and Planetary Change 118 (2014) 85–96
), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge (United Kingdom and New York, NY, USA). Domonkos, P., 2003. Recent precipitation trends in Hungary in the context of larger scale climatic changes. Nat. Hazards 29, 255–271. Falarz, M., 2004. Variability and trends in the duration and depth of snow cover in Poland in the 20th century. Int. J. Climatol. 24 (13), 1713–1727. http://dx.doi.org/10.1002/ joc.1505. Filipiak, J., Mietus, M., 2009. Spatial and temporal variability of cloudiness in Poland, 1971–2000. Int. J. Climatol. 29, 1294–1311. http://dx.doi.org/10.1002/joc.1777. Haylock, M.R., et al., 2008. A European daily high-resolution gridded dataset of surface temperature and precipitation for 1950–2006. J. Geophys. Res. 113, D20119. Helsel, D.R., Hirsch, R.M., 1992. Statistical Methods in Water Resources. Elsevier. Hirschi, M., Seneviratne, S.I., Alexandrov, V., Boberg, F., Boroneant, C., Christensen, O.B., Formayer, H., Orlowsky, B., Stepanek, P., 2011. Observational evidence for soilmoisture impact on hot extremes in southeastern Europe. Nat. Geosci. 4, 17–21. http://dx.doi.org/10.1038/NGEO1032. Ioniţă, M., Rîmbu, N., Chelcea, S., Pătruţ, S., 2012. Multidecadal variability of summer temperature over Romania and its relation with Atlantic Multidecadal Oscillation. Theor. Appl. Climatol. http://dx.doi.org/10.1007/s00704-012-0786-8. Kendall, M.G., 1975. Rank Correlation Methods. Charles Griffin, London. Kozak, J.., Björnsen Gurung, A., Ostapowicz, K. (Eds.), 2011. Research Agenda for the Carpathians: 2010–2015. Kraków (http://www.forumcarpaticum.org/FC-main/ Download/Research_agenda_for_the_Carpathians.pdf). Kundzewicz, Z.W., Radziejewski, M., Pinskwar, I., 2006. Precipitation extremes in the changing climate of Europe. Clim. Res. 31, 51–58. Kyselỳ, J., 2009. Trends in heavy precipitation in the Czech Republic over 1961–2005. Int. J. Climatol. 29, 1745–1758. http://dx.doi.org/10.1002/joc.1784. Lakatos, M., Szentimrey, T., Bihari, Z., Szalai, S., 2013. Creation of a homogenized climate database for the Carpathian region by applying the MASH procedure and the preliminary analysis of the data. Időjárás 117 (1), 143–158. Lapin, M., Štastny, P., Chmelík, M., 2005. Detection of climate change in the Slovak Mountains. Croat. Meteorol. J. 40 (40), 101–104. Leckebusch, G.C., Weimer, A., Pinto, J.G., Reyers, M., Speth, P., 2008. Extreme wind storms over Europe in present and future climate: a cluster analysis approach. Meteorol. Z. 17 (1), 067–082. http://dx.doi.org/10.1127/0941-2948/2008/0266. Lemke, P., Ren, J., Alley, R.B., Allison, I., Carrasco, J., Flato, G., Fujii, Y., Kaser, G., Mote, P., Thomas, R.H., Zhang, T., 2007. Observations: changes in snow, ice and frozen ground. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge (United Kingdom and New York, NY). Mann, H.B., 1945. Nonparametric tests against trend. Econometrica 13, 245–259. McVicar, T.R., Roderick, M.L., Donohue, R.J., Li, L.T., Van Niel, T.G., Thomas, A., Grieser, J., Jhajharia, D., Himri, Y., Mahowald, N.M., Mescherskaya, An.V., Kruger, A.C., Rehman, S., Dinpashoh, Y., 2012. Global review and synthesis of trends in observed terrestrial near-surface wind speeds: implications for evaporation. J. Hydrol. 416-417, 182–205. http://dx.doi.org/10.1016/j.jhydrol.2011.10.024. Micu, D., 2009. Snow pack in the Romanian Carpathians under changing climatic conditions. Meteorol. Atmos. Phys. 105 (1–2), 1–16. http://dx.doi.org/10.1007/s00703009-0035-6. Mihai, B., Savulescu, I., Sandric, I., 2007. Change detection analysis (1986–2002) of vegetation cover in Romania. A study of alpine, subalpine, and forest landscapes in the Iezer Mountains, Southern Carpathians. Mt. Res. Dev. 27 (3), 250–258. http://dx. doi.org/10.1659/mred.0645.
Ruffini, F.V., Ptaček, P. (Eds.), 2008. Atlas of the Carpathian Macroregion. EURAC Research Bolzano & Department of Geography, Palacky University of Olomouc (62 pp.). Ruffini, F.V., Streifeneder, T., Eiselt, B., 2006. Implementing an international mountain convention. An approach for the delimitation of the Carpathian Convention areaEuropean Academy, Bolzano/Bozen (119 pp.). Salas, J.D., 1993. Analysis and modeling of hydrologic time series. Chapter 19 in In: Maidment, D.R. (Ed.), Handbook of Hydrology, McGraw Hill. Spasov, P., Spasov, D., Petrović, P., 2002. Changes in drought occurrences in Serbia. Proceedings of International Conference on Drought Mitigation and Prevention of Land Desertification'. Bled, Slovenia, 21-25 April 2002 (CD-ROM). International Commission on Irrigation and Drainage (ICID), Ljubljana, Slovenia. Spinoni, J., Antofie, T., Barbosa, P., Bihari, Z., Lakatos, M., Szalai, S., Szentimrey, T., Vogt, J., 2013. An overview of drought events in the Carpathian Region in 1961–2010. Adv. Sci. Res. 10, 21–32. http://dx.doi.org/10.5194/asr-10-21-2013. Szalai, S., 2012. CARPATCLIM: A Project for Climate Atlas with Data Harmonization. Balwois. http://balwois.com/2012/USB/papers/355.pdf (accessed: February 2014). Szentimrey, T., 1999. Multiple Analysis of Series for Homogenization (MASH). Proceedings of the 2nd Seminar for Homogenization of Surface Climatological Data. WMO, Budapest, Hungary, pp. 27–46 (WCDMP-No. 41). Szentimrey, T., 2008. Development of MASH homogenization procedure for daily data. Proceedings of the Fifth Seminar for Homogenization and Quality Control in Climatological Databases, Budapest, Hungary, 2006, WCDMP-No. 71, pp. 123–130. Szentimrey, T., 2011. Manual of Homogenization Software MASHv3.03, Hungarian Meteorological Service. p. 64. Szentimrey, T., Bihari, Z., 2007. Mathematical Background of the Spatial Interpolation Methods and the Software MISH (Meteorological Interpolation based on Surface Homogenized Data Basis). Proceedings from the Conference on Spatial Interpolation in Climatology and Meteorology, Budapest, Hungary, 2004, COST Action, 719. COST Office, pp. 17–27. Tar, K., Makra, L., Horváth, S., Kircsi, A., 2001. Temporal change of some statistical characteristics of wind speed over the Great Hungarian Plain. Theor. Appl. Climatol. 69, 69–79. http://dx.doi.org/10.1007/s007040170035. Tomozeiu, R., Busuioc, A., Stefan, S., 2002. Changes in seasonal mean of maximum air temperature in Romania and their connection with large-scale circulation. Int. J. Climatol. 22 (10), 1181–1196. http://dx.doi.org/10.1002/joc.785. Tomozeiu, R., Stefan, S., Busuioc, A., 2005. Winter precipitation variability and large-scale circulation patterns in Romania. Theor. Appl. Climatol. 81 (3-4), 193–201. http://dx. doi.org/10.1007/s00704-004-0082-3. UNEP, 2007. Carpathian Environment Outlook. United Nations Environment Programme (UNEP), Geneva (236 pp.). Venema, V.K.C., Mestre, O., Aguilar, E., Auer, I., Guijarro, J.A., Domonkos, P., Vertacnik, G., Szentimrey, T., Stepanek, P., Zahradnicek, P., Viarre, J., Muller-Westermeier, G., Lakatos, M., Williams, C.N., Menne, M., Lindau, R., Rasol, D., Rustemeier, E., Kolokythas, K., Marinova, T., Andresen, L., Acquaotta, F., Fratianni, S., Cheval, S., Klancar, M., Brunetti, M., Gruber, C., Prohom Duran, M., Likso, T., Esteban, P., Brandsma, T., 2012. Benchmarking homogenization algorithms for monthly data. Clim. Past 8, 89–115. http://dx.doi.org/10.5194/cp-8-89-2012. Vespremeanu-Stroe, A., Cheval, S., Tătui, D., 2012. The wind regime of Romania—characteristics, trends and North Atlantic Oscillation influences. Forum Geogr. XI (2), 118–126. http://dx.doi.org/10.5775/fg.2067-4635.2012.003.d. Villarini, G., 2011. Analyses of annual and seasonal maximum daily rainfall accumulations for Ukraine, Moldova, and Romania. Int. J. Climatol. 32 (14), 2213–2226. http://dx.doi. org/10.1002/joc.3394. Vojtek, M., Fasko, P., Stastny, P., 2003. Some selected snow climate trends in Slovakia with respect to altitude. Acta Meteorol. Univ. Comenianae 32, 17–27.