Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Future changes in extreme precipitation in central Serbia

Future changes in extreme precipitation in central Serbia J. Hydrol. Hydromech., 69, 2021, 2, 196–208 ©2021. This is an open access article distributed DOI: 10.2478/johh-2021-0006 under the Creative Commons Attribution ISSN 1338-4333 NonCommercial-NoDerivatives 4.0 License 1* 1 2 1 Ranka Erić , Ratko Kadović , Vladimir Đurđević , Vesna Đukić University in Belgrade, Faculty of Forestry, Kneza Višeslava 1, 11030 Belgrade, Serbia. University of Belgrade, Faculty of Physics, Institute of Meteorology, Dobračina 16, 11000 Belgrade, Serbia. Corresponding author. Tel.: + 381 11 3053957. E-mail: ranka.eric@sfb.bg.ac.rs Abstract: This paper presents the results of a study focused on the projected changes in extreme precipitation during the 21st century in Central Serbia. The changes are investigated on the basis of historical and modelled data sets of daily precipitation. The historical observation data were recorded at 18 synoptic weather stations in Central Serbia and modelled data were extracted from the regional climate model EBU-POM (Eta Belgrade University-Princeton Ocean Model) under the A1B scenario. The average number of days in a year with precipition ≥ 20, 30, 40 and 50 mm (R20, R30, R40 and R50), the share of daily precipitation above the 20, 30, 40 and 50 mm (P20, P30, P40, P50) in the total annual precipitation and the monthly distribution of these heavy daily precipitation are used as indices of changes in extreme precipitation. These indices, for the three periods 2011–2040, 2041–2070 and 2071–2100, are determined and compared with those obtained for the historical reference period 1961–1990. The results have shown that the main changes in extreme precipitation in Central Serbia will be in their spatial distribution, and the uncertainty of the occurrence of extreme events will decrease. In the future the increase will be more pronounced than the decrease of these indices. We strongly emphasize the benefit of this paper for both the prevention of natural disasters in the study area and for the improvement of the regional climate model. Keywords: Future extreme precipitation; Climate changes; Regional climate model; Central Serbia. INTRODUCTION the twenty-first century is expected to have more frequent and more intense extreme precipitation (Jiang et al., 2012). One of The whole world has experienced different natural the study in Hungary shows that the annual sum is not likely to catastrophic events caused by climate changes. Extreme change significantly in the region, but the seasonal precipitation weather, fuelled by climate change, struck every corner of the is projected to change significantly during the 21st century globe in 2019 (Kramer and Ware, 2019). Hirabayashi et al. (Pongracz et al., 2011). (2008) found that the frequency of floods in many regions of Schneider et al. (2013) evaluated the impact of climate the world has increased. Riverbank erosion is also a serious change on river flow regimes in Europe and found that on the hazard that directly or indirectly causes the suffering of European scale, climate change can be expected to significantly millions of people. Aktar (2013) shows that climate change will modify flow regimes, especially in the Mediterranean and bore- play a significant role in riverbank erosion. All of the disasters al climate zones. For Balkan areas, society relevant strong caused economic damage and also took away a huge number of impacts of such extreme precipitation changes could be ex- humans lives. Because of that, many worldwide studies of pected in particular concerning flood-related damages (Tram- climate changes and their impact on different aspects of the blay and Somot, 2018). Therefore, there are different studies environment, have been performed so far, and their number is about climate changes and their environmental impact in Serbia still growing today (Aktar, 2013; Cavicchia et al., 2018; and in the neighboring countries. They are related to extreme Cooper, 2019; DeGateano and Castellano, 2017; Fotso-Nguemo temperatures, precipitation, soil temperatures, soil moisture, et al., 2019; Huang, Wang et al., 2018; Soltani et al., 2016; Tan droughts and wet periods (Bocheva et al., 2010; Gocić and et al., 2017). Among them, those related to temperature and Trajković, 2013, 2014; Harpa et al., 2019; Janković et al., 2019; precipitation extremes play an important role. The relationship Kržič et al., 2011; Mihailović et al., 2016; Pongracz et al., between extreme precipitation and the corresponding 2011; Unkašević and Tošić, 2011; Vuković et al., 2018). Kržič temperature is of great importance for predicting precipitation et al. (2011) have shown an overall increase in the surface air extremes in the future, given the rapid increase of global temperature of about 2 °C and a decrease in seasonal precipita- warming (Yang et al., 2020). The most important conclusion of tion sums of about 13 mm in Serbia. Unkašević and Tošić the studies conducted were the increases in magnitude, (2011) found that the mean annual precipitation on the wettest frequency, and probability of the occurrence of extreme day during the 20th century across Serbia increased by nearly temperature and extreme precipitation events (Alexander et al., 9%. Gocić and Trajković (2013) analysed changes in precipita- 2006; Bocheva et al., 2010; Harpa et al., 2019; Huang, Wang et tion using Mann–Kendall and Sen's slope estimator statistical al., 2018; Janković et al., 2019; Jiang et al., 2012; Kjellström tests during the period 1980–2010. These analyses are of great and Ruosteenoja, 2007; Li et al., 2019; Tramblay and Somot, importance for water resources planning and for the improve- 2018; Vuković et al., 2018). Kjellström and Ruosteenoja (2007) ment of water resources management in the study area. investigated simulated changes in the precipitation over the Janković et al. (2019) have shown significant decrease of heat- Baltic Sea at the end of the 21st century and the precipitation is ing degree days (HDD) and increase of cooling degree days projected to increase in the Baltic Sea area, especially during (CDD) in the 21st century over Serbia. This study should help winter. For extreme precipitation indices in China, the end of the policy of energy management and planning in Serbia and in 196 Future changes in extreme precipitation in central Serbia the region. Consequently, any future analysis should be helpful Serbia has an area of 55,965 km , hilly and mountainous con- for understanding and decreasing of negative climate change figuration intersected with local river basins (Figure 1). The impacts on natural and anthropogenic systems. climate of Central Serbia is moderate continental with some The territory of Serbia was the study area for several previ- specific characteristics, caused by the geographic location, ously conducted studies on historical and future changes in relief, terrain exposure, etc. The mean annual temperature for precipitation indices (Gocić and Trajković, 2013; Kržič et al., areas with an altitude of up to 300 m is 10.9 °C, and for areas 2011; Unkašević and Tošić, 2011; Vuković et al., 2018). The with an altitude of 300 m to 500 m around 10.0 °C. In moun- mentioned papers mostly describe changes in average monthly, tainous areas with an altitude of over 1000 m, the average an- seasonal or annual precipitation, which have been so far. Cli- nual temperatures are around 6.0 °C, and at an altitude of over mate change can affect the intensity and frequency of precipita- 1500 m around 3.0 °C. Annual precipitation generally rises tion and cause an increase in extreme precipitation. On the with altitude. In lower regions, it ranges in the interval from other side, increases in extreme precipitation may not always 540 to 820 mm. Areas with altitudes of over 1000 m receive on lead to an increase in total precipitation over a season or over average 700 to 1000 mm, and some mountain tops in south- the year – just that precipitation is occurring in more intense western Serbia up to 1500 mm. The major part of Central Ser- events. However, extreme precipitation analyses are very im- bia has the continental precipitation regime, with a peak in the portant due to the potential impacts of extreme precipitation as earlier summer period, except for southwest, which receives the soil erosion, an increase in flood risk due to heavy rain, etc. highest precipitation in autumn. May-June is the rainiest month, According to that, this paper presents the results of a study with an average of 12 to 13% of total annual amount. February focused on the projected changes in extreme precipitation in and October have the lowest precipitation (Republic Hydrome- Central Serbia during three periods in the 21st century: 2021– teorological Service of Serbia, 2020). 2040, 2041–2070, and 2071–2100. There are many contribu- tions of these results. The most important are the possibility of Data predicting future flood events and reducing the risk of their occurrence in Central Serbia. The knowledge of the climate Both historical and modelled data were used in this study. change projections in extreme precipitation is an important The historical data are daily precipitation time series from 18 decision support tool for civil protection and the prevention meteorological stations over a 30-year period (1961–1990). major of natural disasters. More details about the data, methods This period is the reference period for this study. Technical and and results can be found in the next sections. quality controls for these measurements were made by the National Meteorological Service of Serbia. The spatial distribu- DATA AND METHODS tion of the weather stations considered is presented in Figure 1 Study area and their geographical coordinates are listed in Table 1. Except for the direct observation data from stations The Republic of Serbia is located on the Balkan Peninsula (MSTAT), there are also E-OBS daily precipitation data between latitudes 41°–47°N and longitudes 18°–23°E. Central (Cornes et al., 2018) for all 18 stations over the reference period Fig. 1. Location of Serbia in the Europe, altitude of the terrain in Serbia (m a.s.l.) and spatial distribution of the meteorological stations. Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Table 1. Geographical descriptions of the synoptic stations used in dataset that can be explained by the model. It ranges from 0.0 the study. (poor model) to 1.0 (perfect model). Eq. (3) is used to calculate the root mean squared error (RMSE). This metric records in Station name Longitude Latitude Elevation real units the level of overall agreement between the observed (E) (N) (m a.s.l.) and modelled datasets. It is a non-negative metric that has no 1. Belgrade 20°28′ 44°48′ 132 upper bound and for a perfect model the result would be zero 2. Ćuprija 21°22′ 43°56′ (Dawson et al., 2007). 21°58′ 44°08′ 1,027 3. Crni Vrh 22°45′ 43°01′ 450 4. Dimitrovgrad N   () RR−−(R R ) 20°42′ 43°43′ 215    5. Kraljevo obs,i obs mod,i mod i =1   20°56′ 44°02′ 185 6. Kragujevac R = (1) NN  20°48′ 43°17′ 1,711 22 7. Kopaonik   () RR−−(R R )  obs,m i obs od,i mod 21°21′ 43°34′ 166 8. Kruševac     ii == 11 19°14′ 44°43′ 121 9. Loznica 22°33′ 44°14′ 42 10. Negotin () RR−−(R R )  obs,i obs mod,i mod 21°54′ 43°20′ 204 11. Niš i =1 r = (2) 20°02′ 43°50′ 310 12. Požega NN 20°57′ 44°22′ 121 () RR−−(R R ) 13. S. Palanka  obs,i obs mod,i mod ii == 11 19°55′ 44°17′ 176 14. Valjevo 21°31′ 44°45′ 80 15. Veliko Gradište 21°55′ 42°33′ 432 2 16. Vranje () RR − obs,m i od,i 22°17′ 43°53′ 144 17. Zaječar i =1 RMSE = (3) 19°43′ 43°44′ 1,028 18. Zlatibor N where R is indice of the observed data, R is indice of the obs,i mod,i (1961–1990). The E-OBS dataset is a gridded product provid- modelled data, obs is the average value of the observed data ing daily precipitation data obtained through interpolating sta- indices, is the average value of the modelled data indices, mod tion observations on a regular 10 km resolution grid. N is the total number of observations. The many future changes analyses in extreme precipitation Within the model validation, the extreme precipitation indi- were based on the simulations of regional climate models. Most ces of the observed and modelled data are determined and of the regional climate model validations show that they are compared. The threshold for heavy precipitation events in Ser- performing reasonably well in reproducing the spatial patterns bia is in the range of 36.6–52.5 mm (Jovanović et al., 2018). of observed precipitation (DeGateano and Castellano, 2017; Based on this, the indices which are used within this study are Fotso-Nguemo et al., 2019; Harpa et al., 2019; Janković et al., determined. The indices are the average number of days in a 2019; Jiang et al., 2012; Kjellström and Ruosteenoja, 2007; Li year with precipitation amounts of ≥ 20, 30, 40 and 50 mm et al., 2019; Mihailović et al., 2016). Therefore, they provide a (R20, R30, R40, R50, respectively). In order to determine fu- reliable basis for the forecasting of future precipitation and ture changes in extreme precipitation, two more indices were other climate indices. So, we employed the output of the EBU- developed. They are: the share of daily precipitation ≥ 20, 30, POM coupled regional climate model (RCM) for the 1961– 40 and 50 mm (P20, P30, P40, P50) within the total annual 2100 period according to the SRES-A1B scenario (MA1B). precipitation, and the monthly distribution of these heavy daily The horizontal resolution of the regional model is 25 km precipitation. These three indices were determined for the fu- (Đurđević and Rajković, 2008; Đurđević, 2010). A detailed ture modelled data, and also for the observed data. The changes description of the A1B scenario can be found in the IPCC Spe- are detected by comparing these two groups of indices. cial Report (The Special Report on Emissions Scenarios (SRES), Nakićenović and Swart, 2000). RESULTS AND DISCUSSION Validation of the model Method The results of the model validation are shown in the follow- At the beginning of this analysis, model results (MA1B) ing figures and graphs. were validated against the observed data from 18 synoptic Figures 2 and 3 show that E-OBS is in better accordance stations (MSTAT and E-OBS) over the reference period (1961– with the MA1B data than MSTAT. That was probably because 1990). The coefficient of determination (R ), the Pearson the E-OBS and MA1B data have resolutions, and these data are correlation coefficient (r) (Eq. (2)) and root mean square error derived based on the measured data. On the other hand, the (RMSE) (Eq. (3)) were adopted for the model evaluation. The MSTAT data are measured data from one location (meteorolog- Pearson correlation coefficient (Eq. (2)) is used to measure the ical station) and have more precise values. The values of r strength of a linear association between modelled and observed indicate that there are medium-strong to very-strong data corre- data. The value r = 1 means a perfect positive correlation and lations between the MSTAT and the E-OBS and MA1B data. the value r = –1 means a perfect negative correlation. When the The values of the coefficient of determination (most are above Pearson's coefficient is used as a decision-making tool, the 0.5) indicate that a significant part of data variance is explained following statistical inference is common: (0.00, 0.40) weak by the model. RMSE score better results for E-OBS data than correlation; (0.40, 0.75) moderate correlation; (0.75, 0.85) is a for MSTAT data. The model validation shows that the regional good correlation, and (0.85, 1.00) is an excellent correlation. climate model reproduces the spatial variability very well, with The coefficient of determination R² (Eq. (1)) describes the a slightly lower magnitude accuracy of these indices (R20, R30, proportion of the total statistical variance in the observed R40, and R50) for the points at the meteorological stations. 198 Future changes in extreme precipitation in central Serbia Fig. 2. Average number of days in a year with precipitation amounts ≥ 20, 30, 40 and 50 mm for 18 synoptic stations in Central Serbia over the reference period 1961–1990; MSTAT – observed values on meteorological stations, E-OBS – E-OBS gridded climatology from European Climate Assessment & Dataset (ECA&D) project, MA1B – modelled values from the EBU-POM model projection for A1B scenario. Fig. 3. Modelled vs. observed data: Comparation of the average number of days in a year with precipitiation amounts above 20, 30, 40 and 50 mm, root mean squared error, the coefficient of correlation and the coefficient of determination between MA1B and MSTAT data, MA1B and E-OBS data. Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić The correlation weakens with an increasing amount of precipi- decrease (R20 and R30) until the end of the century, while tation. The strongest correlation was found for R20 and the others indicate a significant increase in frequency or intensity weakest for R50. According to these results, the E-OBS data (R40 and R50). Similar results were obtained in similar anal- have a better performance in relation to the MSTAT, so they yses performed in the surrounding area and in other parts of the are used for further analysis of the future changes in extreme world (Jiang et al., 2012; Pongracz et al., 2011; Tramblay and precipitation. The spatial distributions of the R20, R30, R40 Somot, 2018). Detailed results of this study are presented in the and R50 during the reference period are shown in Figure 4. The following sections (all values <1 characterize the decrease, if historical data show that the west and northeast parts of Central they are equal to 1 then they indicate no change and those that Serbia have more days with heavy precipitation than southern are >1 indicate an increase). and southeastern parts. The maximum values of R20, R30, and R40 are recorded in the west. However, the maximum of R50 Changes detected in the R20, R30, R40 and R50 (very heavy precipitation) is found in the northeastern part of the observed area. Figure 5 shows that the spatial distribution of daily precipita- tion ≥ 20 mm in Central Serbia will change during the 21st Future changes in extreme precipitation indices century. During the reference period, most days with precipita- tion ≥ 20 mm were in the western part of the study area. In the The future evolution of extreme precipitation indices in future there will be a rotation, i.e. it will increase and decrease Central Serbia revealed different results depending on the indi- in areas where R20 was low and high, respectively. The ces. Some of them indicate a constant increase or even a small increase will be more noticeable in the south (at the Kopaonik Fig. 4. Spatial distribution of R20, R30, R40 and R50 during the reference period 1961–1990 (E-OBS data). 200 Future changes in extreme precipitation in central Serbia Fig. 5. Projected changes (times „x“) in the average number of days in a year with precipitation amounts ≥ 20 mm (R20) compared to the historical data (E-OBS R20; 1961–1990) for 18 selected locations in Central Serbia through the 21st century (period 2011–2040; period 2041–2070 and period 2071–2100). and Vranje stations) during the first two periods (2011–2040 the lowest number of days with precipitation ≥ 30 mm in the and 2041–2070). The last period (2071–2100) is characterized past and vice versa. The average change intensity indicates that by an increase in R20. It covers the area from south to east, this parameter increases in the whole area and this change where the Zaječar station stands out (during the reference varies over time. Namely, the highest intensity of change period Zaječar had the lowest R20). The largest increase of should be at the end of the century, and the lowest one is in the 80% in R20 is expected in the period 2011–2040. The decrease middle of the century. The maximum increase is about 110% of R20 is noticeable in most of the area during the first two (2.1x), and the highest decrease is about 40%. The maximum periods, and the maximum is 20%. The average intensity of increase is noticed at Kopaonik, Loznica, Zaječar and Niš, and R20 change for the whole analyzed area has a variable trend. the maximum decrease of R30 is observed for Dimitrovgrad. First, there is an increase of this parameter followed by its Future changes in R40 do not follow a certain pattern. Dur- decrease through the middle of the century, and at the end of ing the three future periods, the increase in R40 covers more the century there is another increase of R20. and more parts of the analysed area (Figure 7). The locations of The change of R30 (Figure 6) is very similar to the change the most intense changes are changing. Vranje stood out at the of R20. An increase of R30 occurs in the part of the area with beginning of the century, Loznica and Belgrade in the middle Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 6. Projected changes (times „x“) in the average number of days in a year with precipitation amounts ≥ 30 mm (R30) compared to the historical data (E-OBS R30; 1961–1990) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070 and period 2071–2100). of the century, and at the end of the century the biggest increase These changes are in line with previous researches. Tram- is expected for Kopaonik (even 5 times „x“). Compared to the blay and Somot (2018) showed that in the Mediterranean, fu- reference period, a decrease is also observed for Požega and it ture changes in the spatial distribution of extreme precipitation reached up to 80%. are expected, as well as a decrease and increase in their occur- The change intensity increases as the precipitation amount rence. The end of the twenty-first century is expected to have a increases. The most intensive increase of R50 is expected at the more frequent and more intense extreme precipitation in Cen- end of the analysed period, up to 9x (Figure 8). The southern tral Serbia. Such results were also obtained in another study for part of Central Serbia is the most critical (Vranje and the relevant study area (Jiang et al., 2012). Kopaonik), while at the beginning of the 21st century the most critical points were in the west (Loznica and Valjevo). On the Changes detected in the monthly distribution of R20, R30, R40 other hand, the decrease is more represented than the increase and R50 of R50 at the beginning of the 21st century. However, the number of stations with a daily precipitation amount ≥ 50 mm The presented results (Figure 9) indicate that future changes increases over time. in the monthly distribution of these four indicators (R20, R30, 202 Future changes in extreme precipitation in central Serbia Fig. 7. Projected changes (times „x“) in the average number of days with precipitation amounts ≥ 40 mm (R40) compared to the historical data (E-OBS R40) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070; period 2071–2100). R40, R50) are very similar. During the reference period, the months of the year. In the period from 2041–2070, the majority highest values of R20 and R30 were in June, while for R40 and of days with precipitation higher than 20, 30, 40 and 50 mm are R50 the situation was slightly different. This confirms the fact expected in July, August and October. The results show that June has been one of the rainiest months in Serbia so far (Figure 9) that October will be the month with the largest (Republic Hydrometeorological Service of Serbia, 2020). number of days with extreme precipitation (R20, R30, R40 and Extreme precipitation ≥ 40 mm is most often observed in R50) in the last period of the 21st century. About 70% of the August, then in July and June. Precipitation ≥ 50 mm most total number of extreme events in the year will occur in often occurs in October, June and August. In the period October. This indicates that the uncertainty of extreme 1961–1990, the uncertainty of the event occurrence increases precipitation occurrence will decrease during the 21st century. with the precipitation amount increase. Namely, the probability This change is also typical in the area of neighboring Hungary. of occurrence of R40 and R50 is almost equal over several There, summer precipitation is projected to decrease months, which is not the case with R20 and R30. In the significantly during the 21st century, autumn and winter near future, extreme precipitation is mostly expected in July precipitation amounts are projected to increase (Pongracz et al., (around 40–55%), and the rest is distributed during the other 2011). Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 8. Projected changes (times „x“) in the average number of days with precipitation amounts ≥ 50 mm (R50) compared to the historical data (E-OBS R50) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070; period 2071–2100). Changes detected in the share of precipitation ≥ 20, 30, 40 and increase or decrease in the share for each station (zero indicates 50 mm within the annual precipitation a 100% decrease). It can be noticed that the stations Kopaonik, Niš, Zaječar, and Loznica are the most pronounced in terms of During the reference period, the average shares of daily pre- an increase, and a decrease is noticed at the stations Dimitrov- cipitation in the total annual precipitation were about 17%, grad, Požega, and Smederevska Palanka. 6.6%, 2.5% and 1.1% for daily precipitation ≥ 20 mm, ≥ 30 Since uncertainty assessment is an important part of the mm, ≥ 40 mm, and ≥ 50 mm, respectively (Figure 10). Accord- analysis of the future climate projections, and considering that ing to the results of the modelled data analysis, the average in this study the results of only one model are used, it is worth share of this precipitation in the annual precipitation amount is underlining that results from the EBU-POM model, for the expected to increase in the future. The average share of P20 future projections of annual and seasonal precipitation change will be about 20%, of P30 about 9%, of P40 about 4%, and of are within, and close to the median of the multi-model ensem- P50 about 1.8% of the total annual precipitation. The average ble results (MEP, 2017) from the ENSEMBLES project (Van increase in P20, P30, and P40 is around 1.2x, 1.3x, and 1.55x, der Linden and Mitchell, 2009). Consequently, presented re- respectively. The share of P50 decreases at the beginning of the sults can be considered as a plausible realization of the changes 21st century, and then from 2041–2100 it is expected to in- in future precipitation regime, that is representative in terms to crease on average up to 2.45x. Table 2 shows the individual the other regional climate models results for the same scenario. 204 Future changes in extreme precipitation in central Serbia Fig. 9. Months with the largest number of days with precipitation amounts ≥20, ≥ 30, ≥ 40, ≥ 50 mm for Central Serbia during the reference period and during the 21st century. Table 2. Changes of P20, P30, P40 and P50 during the three periods of the 21st century in relation to the period 1961–1990, expressed by their ratio. INDICES P20 P30 P40 P50 STATIONS Beograd 0.92 1.00 1.12 0.82 1.04 1.43 1.07 1.62 1.51 0.46 0.48 1.01 Ćuprija 1.00 1.09 1.28 0.94 1.09 1.46 1.38 1.76 2.57 0.38 1.30 4.04 Crni Vrh 1.11 1.08 1.26 1.08 1.22 1.65 1.60 1.38 2.48 1.14 1.14 2.39 Dimitrovgrad 0.98 1.10 1.26 0.68 0.94 1.15 0.51 1.60 1.62 0.00 2.80 2.06 Kraljevo 1.13 1.15 1.25 1.00 1.20 1.54 0.71 1.21 1.96 0.17 1.03 1.64 Kragujevac 1.10 1.05 1.19 0.93 1.07 1.49 1.06 0.91 1.77 0.74 0.91 3.54 1.71 1.66 1.77 1.75 1.79 2.58 1.29 1.35 5.41 0.00 3.12 5.90 Kopaonik 1.29 1.40 1.53 1.24 1.69 1.73 1.93 1.66 2.97 1.12 1.80 4.41 Kruševac 1.29 1.13 1.27 1.62 1.48 1.58 1.69 1.97 2.21 2.34 2.40 1.98 Loznica 1.12 1.05 1.45 1.20 0.93 1.70 1.33 1.27 2.42 1.12 1.00 2.57 Negotin Niš 1.05 1.24 1.43 1.34 1.76 2.39 0.89 2.00 3.54 1.01 5.13 10.19 Požega 0.95 0.97 1.26 0.74 1.15 1.56 0.20 1.92 2.26 0.00 1.33 4.09 S. Palanka 1.20 1.15 1.26 1.14 1.23 1.60 0.79 0.72 1.74 0.00 0.68 1.79 Valjevo 1.03 1.00 1.05 0.99 1.05 1.28 0.77 1.28 1.13 1.19 3.18 4.68 V. Gradište 1.10 1.06 1.21 1.18 1.11 1.56 0.71 1.06 1.62 0.56 1.32 0.76 Vranje 1.44 1.50 1.65 1.41 1.38 1.99 2.24 1.50 2.80 0.87 4.55 5.47 Zaječar 1.12 1.14 1.82 1.11 1.18 2.58 0.97 1.38 2.81 0.00 2.36 3.29 Zlatibor 1.13 1.09 1.28 1.30 1.15 1.65 0.78 1.22 1.78 0.27 1.83 1.63 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 10. The share of daily precipitation ≥ 20, 30, 40 and 50 mm in the total annual precipitation in Central Serbia (%); EOBS – E-OBS gridded climatology from European Climate Assessment & Dataset (ECA&D) project, MA1B – modelled values from the EBU-POM model projection for A1B scenario. CONCLUSION precipitation amount increase the changes are more intense. The future decrease of extreme precipitation ranges from This study analyzed future changes in extreme precipitation 10–100%. in Central Serbia using regional climate simulation data Changes in the monthly distribution of R20, R30, R40 and obtained from the EBU-POM model under the SRES-A1B R50 are very similar for all indices. The time of occurrence of scenario. This analysis and other mentioned analyses for the extreme precipitation in Central Serbia is gradually shifting area of Serbia and beyond, show that the climate changes towards the second part of a year. In the period 1961–1990, the during the 21st century have significant impact on extreme most frequent occurrences of extreme precipitation were during precipitation. Therefore, the investigations of climate changes June. Also, the occurrence of these events was more uncertain and their consequences are very important. with the increase in the precipitation amount (they occur during Within this analysis, the number of days with precipitation ≥ several months in the year). In the period 2011–2040, extreme 20, 30, 40 and 50 mm (R20, R30, R40 and R50), their monthly precipitation is expected to be the most frequent in July. In the distribution, and these precipitation shares in the annual middle of the 21st century, they will mostly occur in August precipitation amount are derived for 18 synoptic weather and October, and at the end of the 21st century, October is stations in Central Serbia. Results from the 2011–2040, 2041– expected to be the month with the largest number of extreme 2070 and 2071–2100 periods were compared to results from the precipitation events. In addition to that, in comparison with reference period 1961–1990. historical data, the uncertainty of the occurrence of extreme Consequently, changes are expected in the spatial events will decrease (Figure 9). distribution of R20, R30, R40 and R50. An increase of these In relation to the historical data, the average shares of these indices is expected in the territory, where their minimum extreme precipitation events in the total annual precipitation occurred during the reference period and vice versa. An will be increasing in the future. The increase is 1.2–2.4x, except increase in extreme precipitation is the most expected in the for P50 which is expected to decrease at the beginning of the south of Central Serbia, where Kopaonik, Niš and Vranje stand 21st century. The most noticeable stations are Kopaonik, Niš, out, as well as Loznica (west) and Zaječar (east). The increase Vranje, Zaječar and Loznica. Considering that these stations of these indices is more pronounced compared to their decrease. stood out in terms of an increase in R20, R30, R40 and R50, it The maximum increase was observed for R50, which is was to be expected that they would be also stand out for other as much as 9x. In addition, it is noticed that with the parameters as well. 206 Future changes in extreme precipitation in central Serbia The summary of these results indicates that extreme precipi- extreme precipitation intensity-durationfrequency curves for tation in Central Serbia shifts spatially towards the south of the climate adaptation planning in New York State. Climate territory, and its time of occurrence is autumn (in the past it Services, 5, 23–35. used to be summer). According to that, it is important to note Đurđević, V., 2010. Simulacija klime i klimatskih promena u that the climate and soil conditions are different, and this can jugoistočnoj Evropi korišćenjem regionalnog klimatskog significantly affect the processes of the hydrological cycle. modela. PhD Thesis. University of Belgrade, Belgrade, Ser- Future work could be focused on the combining of these results bia. (In Serbian.) with the analyses of the other hydrological cycle components. Đurđević, V., Rajković, B., 2008.Verification of a coupled In this way, future changes in other hydrological cycle compo- atmosphere-ocean model using satellite observations over nents, and also their joint effects will be determined. In accord- the Adriatic Sea. Annales Geophysicae, 26, 1935–1954. ance with them, the capacities for risk reduction and disaster Fotso-Nguemo, T., Diallo, I., Diakhate, M., Vondou, D., risk management will be strengthened, as well as the integration Mbaye, M., Haensler, A., Gaye, A., Tchawoua, C., 2019. of climate change adaptation practices in the analyzed area. The Projected changes in the seasonal cycle of extreme rainfall application of these analyses in many engineering sciences events from CORDEX simulations over Central Africa. (civil engineering, agriculture, meteorology, etc.) would im- Climatic Change, 155, 339–357. prove the practice and progress of these sciences. Research Gocić, M., Trajković, S., 2013.Spatiotemporal characteristics of efforts such as this one are important for improving the regional drought in Serbia. Journal of Hydrology, 510, 110–123. climate model and for reducing the uncertainties in regional Gocić, M., Trajković, S., 2014. Spatio-temporal patterns of climate modeling. precipitation in Serbia. Theoretical and Applied Climatolo- gy, 117, 419–431. Acknowledgements. The study is supported by the Ministry of Harpa, G.V., Croitoru, A.E., Đurđević, V., Horvath, C., Education, Science and Technological Development, Republic 2019.Future changes in five extreme precipitation indices in of Serbia, No. 43007. We acknowledge the E-OBS dataset from the lowlands of Romania. International Journal of Climatol- the EU-FP6 project UERRA (https://www.uerra.eu) and the ogy, 39, 5720–5740. Copernicus Climate Change Service, and the data providers in Hirabayashi, Y., Kanae, S., Emori, S., Oki, T. and Kimoto, M., the ECA&D project (https://www.ecad.eu). 2008.Global projections of changing risks of floods and droughts in a changing climate. Hydrol. Sci. J., 53, 4, 754– REFERENCES 772. Huang, G., Wang, X., Guo, J., Li, Y., Yang, L., 2018. Future Aktar, N., 2013. Impact of climate change on riverbank erosion. changes in precipitation extremes over China projected by a International Journal of Sciences: Basic and Applied Re- regional climate model ensemble. Atmospheric Environ- search (IJSBAR), 7, 1, 36–42. ment, 188, 142–156. Alexander, L.V., Zhang, X., Peterson, T.C., Caesar, J. Gleason, Janković, A., Podraščanin, Z., Đurđević, V., 2019. Future cli- B., Klein Tank, A.M.G., Haylock, M., Collins, D., Trewin, mate change impacts on residential heating and cooling de- B., Rahimzadeh, F., Tagipour, A., Rupa Kumar, K., Re- gree days in Serbia. Időjárás, Quarterly Journal of the Hun- vadekar, J., Griffiths, G., Vincent, L., Stephenson, D.B., garian Meteorological Service, 123, 351–370. Burn, J., Aguilar, E., Brunet, M., Taylor, M., New, M., Zhai, Jiang, Z., Song, J., Li, L., Chen, W., Wang, Z., Wang, J., 2012. P., Rusticucci, M., Vazquez-Aguirre, J.L., 2006. Global ob- Extreme climate events in China: IPCC-AR4 modelevalua- served changes in daily climate extremes of temperature and tion and projection. Climatic Change, 110, 385–401. precipitation. Journal of Geophysical Research, 111, Jovanović, S., Anđelković, G., Manojlović, S., Samardžić, I., D05109. https://doi.org/10.1029/2005JD006290 Živković, Lj., Šabić, D., Gatarić, D., Džinović, M., 2018. Bocheva, L., Gospodinov, I., Simeonov, P. Marinova, T., Extreme precipitation events in Serbia: Defining the thresh- 2010.Climatological analysis of the synoptic situations caus- old criteria for emergency preparedness. Atmosphere, 9, 5, ing torrential precipitation events in Bulgaria during the pe- 188. riod 1961–2007. In: Global Environmental Change: Chal- Kjellström, E., Ruosteenoja, K., 2007. Present-day and future lenges to Science and Society in Southeastern Europe, pp. precipitation in the Baltic Sea region as simulated in a suite 97–108. of regional climate models. Climatic Change, 81, 281–291. Cavicchia, L., Scoccimarro, E., Gualdi, S., Marson, P., Ahren, Kramer, K., Ware, J., 2019. Counting the cost: 2019: a year of B., Berthou, S., Conte, D., Aquila, A., Drobinski, P., climate breakdown. Christian Aid, London, United King- Đurđević, V., Dubois, C., Gallardo, C., Li, L., Oddo, P., dom. Sanna, A., Torma, C., 2018. Mediterranean extreme precipi- Kržič, A., Tošić, I., Đurđević, V., Veljović, K., Rajković, B., tation: a multi‑model assessment. Climate Dynamics, 51, 2011. Changes in climate indices for Serbia according to the 901–913. SRES-A1B and SRES-A2 scenarios. Climate Research, 49, Cooper, R., 2019. Projection of future precipitation extremes 73–86. across the Bangkok Metropolitan Region. Heliyon, 5, Li, L., Yao, N., Liu, D.L., Song, S., Lin, H., Chen, X., Li, Y., E01678. 2019. Historical and future projected frequency of extreme Cornes, R., Van der Schrier, G., Van den Besselaar, E.J.M., precipitation indicators using the optimized cumulative dis- Jones, P.D., 2018. An ensemble version of the E-OBS tem- tribution functions in China. Journal of Hydrology, 579, Ar- perature and precipitation datasets. J. Geophys. Res. Atmos., ticle Number: 124170. 123. DOI: 10.1029/2017JD028200 MEP, 2017. Second National Communication of the Republic Dawson, C.W., Abrahart, R.J., See, L.M., 2007. HydroTest: of Serbia under the United Nations Framework Convention Aweb-based toolbox of evaluation metrics for the standard- on Climate Change. Ministry of Environmental Protection of the Republic of Serbia, Belgrade, Serbia, 162 p. ised assessment of hydrological forecasts. Environmental http://www.klimatskepromene.rs/wp- Modelling & Software, 22, 1034–1052. content/uploads/2017/09/SNC_eng.pdf DeGaetano, A., Castellano, C., 2017. Future projections of Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Mihailović, D., Drešković, N., Arsenić, I., Ćirić, V., Đurđević, tation extreme events over Iran. Theoretical and Applied V., Mimić, G., Pap, I., Balaž, I., 2016. Impact of climate Climatology, 126, 775–795 change on soil thermal and moisture regimes in Serbia: An Tan, X., Gan, T., Shao, D., 2017. Effects of persistence and analysis with data from regional climate simulations under large-scale climate anomalies on trends and change points in SRES-A1B. Science of the Total Environment, 571, 398–409. extreme precipitation of Canada. Journal of Hydrology, 550, Nakicenovic, N., Swart, R. (Eds.), 2000. Special Report on 453–465. Emissions Scenarios. A Special Report of Working Group Tramblay, Y., Somot, S., 2018. Future evolution of extreme III of the Intergovernmental Panel on Climate Change. precipitation in the Mediterranean. Climatic Change, 151, Cambridge University Press, Cambridge, United Kingdom 289–302. and New York, NY, USA. Unkašević, M., Tošić, I., 2011. A statistical analysis of the Pongracz, R., Bartholy, J., Miklos, E., 2011. Analysis of pro- daily precipitation over Serbia: trends and indices. Theoreti- jected climate change for Hungary using ensembles simula- cal and Applied Climatology, 106, 69–78. tions. Applied Ecology and Environmental Research, 9, 4, Van der Linden P., Mitchell, J.F.B. (eds.), 2009. ENSEMBLES: 387–398. Climate change and its impacts: Summary of research and re- Republic Hydrometeorological Service of Serbia, 2020. Pre- sults from the ENSEMBLES project. Met Office Hadley cipitation regime in Serbia. Retrieved May 3, 2020. Centre, FitzRoy Road, Exeter EX1 3PB, UK. 160p. http://www.hidmet.gov.rs/eng/meteorologija/klimatologija_ Vuković, A., Vujadinović, M., Rendulić, S., Đurđević, V., padav_rezim.php Ruml, M., Babić, V., Popović, D., 2018. Global warming Schneider, C., Laize, C.L.R., Acreman, M.C., Florke, M., 2013. impact on climate change in Serbia for the period 1961- How will climate change modify river flow regimes in Eu- 2100.Thermal Science, 22, 2267–2280. rope? Hydrol. Earth Syst. Sci., 17, 325–339. Yang, Z., Gao, X., Guo, M., Zhu, Q., Xu, Z., Gao., K., 2020. Soltani, M., Laux, P., Kunstmann, H., Stan, K., Sohrabi, M.M., Temperature dependence of extreme precipitation over Molanejad, M., Sabziparvar, A.A., SaadatAbadi, A., mainland China. Journal of Hydrology, 583, 1245. Ranjbar, F., Rousta, I., Zawar-Reza, P., Khoshakhlagh, F., Soltanzadeh, I., Babu, C.A., Azizi G.H., Martin, M.V., 2016. Received 16 October 2020 Assessment of climate variations in temperature and precipi- Accepted 3 February 2021 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Hydrology and Hydromechanics de Gruyter

Future changes in extreme precipitation in central Serbia

Download PDF
13 pages

Loading next page...
 
/lp/de-gruyter/future-changes-in-extreme-precipitation-in-central-serbia-Noyc2bDTPu

References (36)

Publisher
de Gruyter
Copyright
© 2021 Ranka Erić et al., published by Sciendo
ISSN
0042-790X
eISSN
1338-4333
DOI
10.2478/johh-2021-0006
Publisher site
See Article on Publisher Site

Abstract

J. Hydrol. Hydromech., 69, 2021, 2, 196–208 ©2021. This is an open access article distributed DOI: 10.2478/johh-2021-0006 under the Creative Commons Attribution ISSN 1338-4333 NonCommercial-NoDerivatives 4.0 License 1* 1 2 1 Ranka Erić , Ratko Kadović , Vladimir Đurđević , Vesna Đukić University in Belgrade, Faculty of Forestry, Kneza Višeslava 1, 11030 Belgrade, Serbia. University of Belgrade, Faculty of Physics, Institute of Meteorology, Dobračina 16, 11000 Belgrade, Serbia. Corresponding author. Tel.: + 381 11 3053957. E-mail: ranka.eric@sfb.bg.ac.rs Abstract: This paper presents the results of a study focused on the projected changes in extreme precipitation during the 21st century in Central Serbia. The changes are investigated on the basis of historical and modelled data sets of daily precipitation. The historical observation data were recorded at 18 synoptic weather stations in Central Serbia and modelled data were extracted from the regional climate model EBU-POM (Eta Belgrade University-Princeton Ocean Model) under the A1B scenario. The average number of days in a year with precipition ≥ 20, 30, 40 and 50 mm (R20, R30, R40 and R50), the share of daily precipitation above the 20, 30, 40 and 50 mm (P20, P30, P40, P50) in the total annual precipitation and the monthly distribution of these heavy daily precipitation are used as indices of changes in extreme precipitation. These indices, for the three periods 2011–2040, 2041–2070 and 2071–2100, are determined and compared with those obtained for the historical reference period 1961–1990. The results have shown that the main changes in extreme precipitation in Central Serbia will be in their spatial distribution, and the uncertainty of the occurrence of extreme events will decrease. In the future the increase will be more pronounced than the decrease of these indices. We strongly emphasize the benefit of this paper for both the prevention of natural disasters in the study area and for the improvement of the regional climate model. Keywords: Future extreme precipitation; Climate changes; Regional climate model; Central Serbia. INTRODUCTION the twenty-first century is expected to have more frequent and more intense extreme precipitation (Jiang et al., 2012). One of The whole world has experienced different natural the study in Hungary shows that the annual sum is not likely to catastrophic events caused by climate changes. Extreme change significantly in the region, but the seasonal precipitation weather, fuelled by climate change, struck every corner of the is projected to change significantly during the 21st century globe in 2019 (Kramer and Ware, 2019). Hirabayashi et al. (Pongracz et al., 2011). (2008) found that the frequency of floods in many regions of Schneider et al. (2013) evaluated the impact of climate the world has increased. Riverbank erosion is also a serious change on river flow regimes in Europe and found that on the hazard that directly or indirectly causes the suffering of European scale, climate change can be expected to significantly millions of people. Aktar (2013) shows that climate change will modify flow regimes, especially in the Mediterranean and bore- play a significant role in riverbank erosion. All of the disasters al climate zones. For Balkan areas, society relevant strong caused economic damage and also took away a huge number of impacts of such extreme precipitation changes could be ex- humans lives. Because of that, many worldwide studies of pected in particular concerning flood-related damages (Tram- climate changes and their impact on different aspects of the blay and Somot, 2018). Therefore, there are different studies environment, have been performed so far, and their number is about climate changes and their environmental impact in Serbia still growing today (Aktar, 2013; Cavicchia et al., 2018; and in the neighboring countries. They are related to extreme Cooper, 2019; DeGateano and Castellano, 2017; Fotso-Nguemo temperatures, precipitation, soil temperatures, soil moisture, et al., 2019; Huang, Wang et al., 2018; Soltani et al., 2016; Tan droughts and wet periods (Bocheva et al., 2010; Gocić and et al., 2017). Among them, those related to temperature and Trajković, 2013, 2014; Harpa et al., 2019; Janković et al., 2019; precipitation extremes play an important role. The relationship Kržič et al., 2011; Mihailović et al., 2016; Pongracz et al., between extreme precipitation and the corresponding 2011; Unkašević and Tošić, 2011; Vuković et al., 2018). Kržič temperature is of great importance for predicting precipitation et al. (2011) have shown an overall increase in the surface air extremes in the future, given the rapid increase of global temperature of about 2 °C and a decrease in seasonal precipita- warming (Yang et al., 2020). The most important conclusion of tion sums of about 13 mm in Serbia. Unkašević and Tošić the studies conducted were the increases in magnitude, (2011) found that the mean annual precipitation on the wettest frequency, and probability of the occurrence of extreme day during the 20th century across Serbia increased by nearly temperature and extreme precipitation events (Alexander et al., 9%. Gocić and Trajković (2013) analysed changes in precipita- 2006; Bocheva et al., 2010; Harpa et al., 2019; Huang, Wang et tion using Mann–Kendall and Sen's slope estimator statistical al., 2018; Janković et al., 2019; Jiang et al., 2012; Kjellström tests during the period 1980–2010. These analyses are of great and Ruosteenoja, 2007; Li et al., 2019; Tramblay and Somot, importance for water resources planning and for the improve- 2018; Vuković et al., 2018). Kjellström and Ruosteenoja (2007) ment of water resources management in the study area. investigated simulated changes in the precipitation over the Janković et al. (2019) have shown significant decrease of heat- Baltic Sea at the end of the 21st century and the precipitation is ing degree days (HDD) and increase of cooling degree days projected to increase in the Baltic Sea area, especially during (CDD) in the 21st century over Serbia. This study should help winter. For extreme precipitation indices in China, the end of the policy of energy management and planning in Serbia and in 196 Future changes in extreme precipitation in central Serbia the region. Consequently, any future analysis should be helpful Serbia has an area of 55,965 km , hilly and mountainous con- for understanding and decreasing of negative climate change figuration intersected with local river basins (Figure 1). The impacts on natural and anthropogenic systems. climate of Central Serbia is moderate continental with some The territory of Serbia was the study area for several previ- specific characteristics, caused by the geographic location, ously conducted studies on historical and future changes in relief, terrain exposure, etc. The mean annual temperature for precipitation indices (Gocić and Trajković, 2013; Kržič et al., areas with an altitude of up to 300 m is 10.9 °C, and for areas 2011; Unkašević and Tošić, 2011; Vuković et al., 2018). The with an altitude of 300 m to 500 m around 10.0 °C. In moun- mentioned papers mostly describe changes in average monthly, tainous areas with an altitude of over 1000 m, the average an- seasonal or annual precipitation, which have been so far. Cli- nual temperatures are around 6.0 °C, and at an altitude of over mate change can affect the intensity and frequency of precipita- 1500 m around 3.0 °C. Annual precipitation generally rises tion and cause an increase in extreme precipitation. On the with altitude. In lower regions, it ranges in the interval from other side, increases in extreme precipitation may not always 540 to 820 mm. Areas with altitudes of over 1000 m receive on lead to an increase in total precipitation over a season or over average 700 to 1000 mm, and some mountain tops in south- the year – just that precipitation is occurring in more intense western Serbia up to 1500 mm. The major part of Central Ser- events. However, extreme precipitation analyses are very im- bia has the continental precipitation regime, with a peak in the portant due to the potential impacts of extreme precipitation as earlier summer period, except for southwest, which receives the soil erosion, an increase in flood risk due to heavy rain, etc. highest precipitation in autumn. May-June is the rainiest month, According to that, this paper presents the results of a study with an average of 12 to 13% of total annual amount. February focused on the projected changes in extreme precipitation in and October have the lowest precipitation (Republic Hydrome- Central Serbia during three periods in the 21st century: 2021– teorological Service of Serbia, 2020). 2040, 2041–2070, and 2071–2100. There are many contribu- tions of these results. The most important are the possibility of Data predicting future flood events and reducing the risk of their occurrence in Central Serbia. The knowledge of the climate Both historical and modelled data were used in this study. change projections in extreme precipitation is an important The historical data are daily precipitation time series from 18 decision support tool for civil protection and the prevention meteorological stations over a 30-year period (1961–1990). major of natural disasters. More details about the data, methods This period is the reference period for this study. Technical and and results can be found in the next sections. quality controls for these measurements were made by the National Meteorological Service of Serbia. The spatial distribu- DATA AND METHODS tion of the weather stations considered is presented in Figure 1 Study area and their geographical coordinates are listed in Table 1. Except for the direct observation data from stations The Republic of Serbia is located on the Balkan Peninsula (MSTAT), there are also E-OBS daily precipitation data between latitudes 41°–47°N and longitudes 18°–23°E. Central (Cornes et al., 2018) for all 18 stations over the reference period Fig. 1. Location of Serbia in the Europe, altitude of the terrain in Serbia (m a.s.l.) and spatial distribution of the meteorological stations. Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Table 1. Geographical descriptions of the synoptic stations used in dataset that can be explained by the model. It ranges from 0.0 the study. (poor model) to 1.0 (perfect model). Eq. (3) is used to calculate the root mean squared error (RMSE). This metric records in Station name Longitude Latitude Elevation real units the level of overall agreement between the observed (E) (N) (m a.s.l.) and modelled datasets. It is a non-negative metric that has no 1. Belgrade 20°28′ 44°48′ 132 upper bound and for a perfect model the result would be zero 2. Ćuprija 21°22′ 43°56′ (Dawson et al., 2007). 21°58′ 44°08′ 1,027 3. Crni Vrh 22°45′ 43°01′ 450 4. Dimitrovgrad N   () RR−−(R R ) 20°42′ 43°43′ 215    5. Kraljevo obs,i obs mod,i mod i =1   20°56′ 44°02′ 185 6. Kragujevac R = (1) NN  20°48′ 43°17′ 1,711 22 7. Kopaonik   () RR−−(R R )  obs,m i obs od,i mod 21°21′ 43°34′ 166 8. Kruševac     ii == 11 19°14′ 44°43′ 121 9. Loznica 22°33′ 44°14′ 42 10. Negotin () RR−−(R R )  obs,i obs mod,i mod 21°54′ 43°20′ 204 11. Niš i =1 r = (2) 20°02′ 43°50′ 310 12. Požega NN 20°57′ 44°22′ 121 () RR−−(R R ) 13. S. Palanka  obs,i obs mod,i mod ii == 11 19°55′ 44°17′ 176 14. Valjevo 21°31′ 44°45′ 80 15. Veliko Gradište 21°55′ 42°33′ 432 2 16. Vranje () RR − obs,m i od,i 22°17′ 43°53′ 144 17. Zaječar i =1 RMSE = (3) 19°43′ 43°44′ 1,028 18. Zlatibor N where R is indice of the observed data, R is indice of the obs,i mod,i (1961–1990). The E-OBS dataset is a gridded product provid- modelled data, obs is the average value of the observed data ing daily precipitation data obtained through interpolating sta- indices, is the average value of the modelled data indices, mod tion observations on a regular 10 km resolution grid. N is the total number of observations. The many future changes analyses in extreme precipitation Within the model validation, the extreme precipitation indi- were based on the simulations of regional climate models. Most ces of the observed and modelled data are determined and of the regional climate model validations show that they are compared. The threshold for heavy precipitation events in Ser- performing reasonably well in reproducing the spatial patterns bia is in the range of 36.6–52.5 mm (Jovanović et al., 2018). of observed precipitation (DeGateano and Castellano, 2017; Based on this, the indices which are used within this study are Fotso-Nguemo et al., 2019; Harpa et al., 2019; Janković et al., determined. The indices are the average number of days in a 2019; Jiang et al., 2012; Kjellström and Ruosteenoja, 2007; Li year with precipitation amounts of ≥ 20, 30, 40 and 50 mm et al., 2019; Mihailović et al., 2016). Therefore, they provide a (R20, R30, R40, R50, respectively). In order to determine fu- reliable basis for the forecasting of future precipitation and ture changes in extreme precipitation, two more indices were other climate indices. So, we employed the output of the EBU- developed. They are: the share of daily precipitation ≥ 20, 30, POM coupled regional climate model (RCM) for the 1961– 40 and 50 mm (P20, P30, P40, P50) within the total annual 2100 period according to the SRES-A1B scenario (MA1B). precipitation, and the monthly distribution of these heavy daily The horizontal resolution of the regional model is 25 km precipitation. These three indices were determined for the fu- (Đurđević and Rajković, 2008; Đurđević, 2010). A detailed ture modelled data, and also for the observed data. The changes description of the A1B scenario can be found in the IPCC Spe- are detected by comparing these two groups of indices. cial Report (The Special Report on Emissions Scenarios (SRES), Nakićenović and Swart, 2000). RESULTS AND DISCUSSION Validation of the model Method The results of the model validation are shown in the follow- At the beginning of this analysis, model results (MA1B) ing figures and graphs. were validated against the observed data from 18 synoptic Figures 2 and 3 show that E-OBS is in better accordance stations (MSTAT and E-OBS) over the reference period (1961– with the MA1B data than MSTAT. That was probably because 1990). The coefficient of determination (R ), the Pearson the E-OBS and MA1B data have resolutions, and these data are correlation coefficient (r) (Eq. (2)) and root mean square error derived based on the measured data. On the other hand, the (RMSE) (Eq. (3)) were adopted for the model evaluation. The MSTAT data are measured data from one location (meteorolog- Pearson correlation coefficient (Eq. (2)) is used to measure the ical station) and have more precise values. The values of r strength of a linear association between modelled and observed indicate that there are medium-strong to very-strong data corre- data. The value r = 1 means a perfect positive correlation and lations between the MSTAT and the E-OBS and MA1B data. the value r = –1 means a perfect negative correlation. When the The values of the coefficient of determination (most are above Pearson's coefficient is used as a decision-making tool, the 0.5) indicate that a significant part of data variance is explained following statistical inference is common: (0.00, 0.40) weak by the model. RMSE score better results for E-OBS data than correlation; (0.40, 0.75) moderate correlation; (0.75, 0.85) is a for MSTAT data. The model validation shows that the regional good correlation, and (0.85, 1.00) is an excellent correlation. climate model reproduces the spatial variability very well, with The coefficient of determination R² (Eq. (1)) describes the a slightly lower magnitude accuracy of these indices (R20, R30, proportion of the total statistical variance in the observed R40, and R50) for the points at the meteorological stations. 198 Future changes in extreme precipitation in central Serbia Fig. 2. Average number of days in a year with precipitation amounts ≥ 20, 30, 40 and 50 mm for 18 synoptic stations in Central Serbia over the reference period 1961–1990; MSTAT – observed values on meteorological stations, E-OBS – E-OBS gridded climatology from European Climate Assessment & Dataset (ECA&D) project, MA1B – modelled values from the EBU-POM model projection for A1B scenario. Fig. 3. Modelled vs. observed data: Comparation of the average number of days in a year with precipitiation amounts above 20, 30, 40 and 50 mm, root mean squared error, the coefficient of correlation and the coefficient of determination between MA1B and MSTAT data, MA1B and E-OBS data. Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić The correlation weakens with an increasing amount of precipi- decrease (R20 and R30) until the end of the century, while tation. The strongest correlation was found for R20 and the others indicate a significant increase in frequency or intensity weakest for R50. According to these results, the E-OBS data (R40 and R50). Similar results were obtained in similar anal- have a better performance in relation to the MSTAT, so they yses performed in the surrounding area and in other parts of the are used for further analysis of the future changes in extreme world (Jiang et al., 2012; Pongracz et al., 2011; Tramblay and precipitation. The spatial distributions of the R20, R30, R40 Somot, 2018). Detailed results of this study are presented in the and R50 during the reference period are shown in Figure 4. The following sections (all values <1 characterize the decrease, if historical data show that the west and northeast parts of Central they are equal to 1 then they indicate no change and those that Serbia have more days with heavy precipitation than southern are >1 indicate an increase). and southeastern parts. The maximum values of R20, R30, and R40 are recorded in the west. However, the maximum of R50 Changes detected in the R20, R30, R40 and R50 (very heavy precipitation) is found in the northeastern part of the observed area. Figure 5 shows that the spatial distribution of daily precipita- tion ≥ 20 mm in Central Serbia will change during the 21st Future changes in extreme precipitation indices century. During the reference period, most days with precipita- tion ≥ 20 mm were in the western part of the study area. In the The future evolution of extreme precipitation indices in future there will be a rotation, i.e. it will increase and decrease Central Serbia revealed different results depending on the indi- in areas where R20 was low and high, respectively. The ces. Some of them indicate a constant increase or even a small increase will be more noticeable in the south (at the Kopaonik Fig. 4. Spatial distribution of R20, R30, R40 and R50 during the reference period 1961–1990 (E-OBS data). 200 Future changes in extreme precipitation in central Serbia Fig. 5. Projected changes (times „x“) in the average number of days in a year with precipitation amounts ≥ 20 mm (R20) compared to the historical data (E-OBS R20; 1961–1990) for 18 selected locations in Central Serbia through the 21st century (period 2011–2040; period 2041–2070 and period 2071–2100). and Vranje stations) during the first two periods (2011–2040 the lowest number of days with precipitation ≥ 30 mm in the and 2041–2070). The last period (2071–2100) is characterized past and vice versa. The average change intensity indicates that by an increase in R20. It covers the area from south to east, this parameter increases in the whole area and this change where the Zaječar station stands out (during the reference varies over time. Namely, the highest intensity of change period Zaječar had the lowest R20). The largest increase of should be at the end of the century, and the lowest one is in the 80% in R20 is expected in the period 2011–2040. The decrease middle of the century. The maximum increase is about 110% of R20 is noticeable in most of the area during the first two (2.1x), and the highest decrease is about 40%. The maximum periods, and the maximum is 20%. The average intensity of increase is noticed at Kopaonik, Loznica, Zaječar and Niš, and R20 change for the whole analyzed area has a variable trend. the maximum decrease of R30 is observed for Dimitrovgrad. First, there is an increase of this parameter followed by its Future changes in R40 do not follow a certain pattern. Dur- decrease through the middle of the century, and at the end of ing the three future periods, the increase in R40 covers more the century there is another increase of R20. and more parts of the analysed area (Figure 7). The locations of The change of R30 (Figure 6) is very similar to the change the most intense changes are changing. Vranje stood out at the of R20. An increase of R30 occurs in the part of the area with beginning of the century, Loznica and Belgrade in the middle Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 6. Projected changes (times „x“) in the average number of days in a year with precipitation amounts ≥ 30 mm (R30) compared to the historical data (E-OBS R30; 1961–1990) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070 and period 2071–2100). of the century, and at the end of the century the biggest increase These changes are in line with previous researches. Tram- is expected for Kopaonik (even 5 times „x“). Compared to the blay and Somot (2018) showed that in the Mediterranean, fu- reference period, a decrease is also observed for Požega and it ture changes in the spatial distribution of extreme precipitation reached up to 80%. are expected, as well as a decrease and increase in their occur- The change intensity increases as the precipitation amount rence. The end of the twenty-first century is expected to have a increases. The most intensive increase of R50 is expected at the more frequent and more intense extreme precipitation in Cen- end of the analysed period, up to 9x (Figure 8). The southern tral Serbia. Such results were also obtained in another study for part of Central Serbia is the most critical (Vranje and the relevant study area (Jiang et al., 2012). Kopaonik), while at the beginning of the 21st century the most critical points were in the west (Loznica and Valjevo). On the Changes detected in the monthly distribution of R20, R30, R40 other hand, the decrease is more represented than the increase and R50 of R50 at the beginning of the 21st century. However, the number of stations with a daily precipitation amount ≥ 50 mm The presented results (Figure 9) indicate that future changes increases over time. in the monthly distribution of these four indicators (R20, R30, 202 Future changes in extreme precipitation in central Serbia Fig. 7. Projected changes (times „x“) in the average number of days with precipitation amounts ≥ 40 mm (R40) compared to the historical data (E-OBS R40) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070; period 2071–2100). R40, R50) are very similar. During the reference period, the months of the year. In the period from 2041–2070, the majority highest values of R20 and R30 were in June, while for R40 and of days with precipitation higher than 20, 30, 40 and 50 mm are R50 the situation was slightly different. This confirms the fact expected in July, August and October. The results show that June has been one of the rainiest months in Serbia so far (Figure 9) that October will be the month with the largest (Republic Hydrometeorological Service of Serbia, 2020). number of days with extreme precipitation (R20, R30, R40 and Extreme precipitation ≥ 40 mm is most often observed in R50) in the last period of the 21st century. About 70% of the August, then in July and June. Precipitation ≥ 50 mm most total number of extreme events in the year will occur in often occurs in October, June and August. In the period October. This indicates that the uncertainty of extreme 1961–1990, the uncertainty of the event occurrence increases precipitation occurrence will decrease during the 21st century. with the precipitation amount increase. Namely, the probability This change is also typical in the area of neighboring Hungary. of occurrence of R40 and R50 is almost equal over several There, summer precipitation is projected to decrease months, which is not the case with R20 and R30. In the significantly during the 21st century, autumn and winter near future, extreme precipitation is mostly expected in July precipitation amounts are projected to increase (Pongracz et al., (around 40–55%), and the rest is distributed during the other 2011). Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 8. Projected changes (times „x“) in the average number of days with precipitation amounts ≥ 50 mm (R50) compared to the historical data (E-OBS R50) for 18 selected locations in Central Serbia during the 21st century (period 2011–2040; period 2041–2070; period 2071–2100). Changes detected in the share of precipitation ≥ 20, 30, 40 and increase or decrease in the share for each station (zero indicates 50 mm within the annual precipitation a 100% decrease). It can be noticed that the stations Kopaonik, Niš, Zaječar, and Loznica are the most pronounced in terms of During the reference period, the average shares of daily pre- an increase, and a decrease is noticed at the stations Dimitrov- cipitation in the total annual precipitation were about 17%, grad, Požega, and Smederevska Palanka. 6.6%, 2.5% and 1.1% for daily precipitation ≥ 20 mm, ≥ 30 Since uncertainty assessment is an important part of the mm, ≥ 40 mm, and ≥ 50 mm, respectively (Figure 10). Accord- analysis of the future climate projections, and considering that ing to the results of the modelled data analysis, the average in this study the results of only one model are used, it is worth share of this precipitation in the annual precipitation amount is underlining that results from the EBU-POM model, for the expected to increase in the future. The average share of P20 future projections of annual and seasonal precipitation change will be about 20%, of P30 about 9%, of P40 about 4%, and of are within, and close to the median of the multi-model ensem- P50 about 1.8% of the total annual precipitation. The average ble results (MEP, 2017) from the ENSEMBLES project (Van increase in P20, P30, and P40 is around 1.2x, 1.3x, and 1.55x, der Linden and Mitchell, 2009). Consequently, presented re- respectively. The share of P50 decreases at the beginning of the sults can be considered as a plausible realization of the changes 21st century, and then from 2041–2100 it is expected to in- in future precipitation regime, that is representative in terms to crease on average up to 2.45x. Table 2 shows the individual the other regional climate models results for the same scenario. 204 Future changes in extreme precipitation in central Serbia Fig. 9. Months with the largest number of days with precipitation amounts ≥20, ≥ 30, ≥ 40, ≥ 50 mm for Central Serbia during the reference period and during the 21st century. Table 2. Changes of P20, P30, P40 and P50 during the three periods of the 21st century in relation to the period 1961–1990, expressed by their ratio. INDICES P20 P30 P40 P50 STATIONS Beograd 0.92 1.00 1.12 0.82 1.04 1.43 1.07 1.62 1.51 0.46 0.48 1.01 Ćuprija 1.00 1.09 1.28 0.94 1.09 1.46 1.38 1.76 2.57 0.38 1.30 4.04 Crni Vrh 1.11 1.08 1.26 1.08 1.22 1.65 1.60 1.38 2.48 1.14 1.14 2.39 Dimitrovgrad 0.98 1.10 1.26 0.68 0.94 1.15 0.51 1.60 1.62 0.00 2.80 2.06 Kraljevo 1.13 1.15 1.25 1.00 1.20 1.54 0.71 1.21 1.96 0.17 1.03 1.64 Kragujevac 1.10 1.05 1.19 0.93 1.07 1.49 1.06 0.91 1.77 0.74 0.91 3.54 1.71 1.66 1.77 1.75 1.79 2.58 1.29 1.35 5.41 0.00 3.12 5.90 Kopaonik 1.29 1.40 1.53 1.24 1.69 1.73 1.93 1.66 2.97 1.12 1.80 4.41 Kruševac 1.29 1.13 1.27 1.62 1.48 1.58 1.69 1.97 2.21 2.34 2.40 1.98 Loznica 1.12 1.05 1.45 1.20 0.93 1.70 1.33 1.27 2.42 1.12 1.00 2.57 Negotin Niš 1.05 1.24 1.43 1.34 1.76 2.39 0.89 2.00 3.54 1.01 5.13 10.19 Požega 0.95 0.97 1.26 0.74 1.15 1.56 0.20 1.92 2.26 0.00 1.33 4.09 S. Palanka 1.20 1.15 1.26 1.14 1.23 1.60 0.79 0.72 1.74 0.00 0.68 1.79 Valjevo 1.03 1.00 1.05 0.99 1.05 1.28 0.77 1.28 1.13 1.19 3.18 4.68 V. Gradište 1.10 1.06 1.21 1.18 1.11 1.56 0.71 1.06 1.62 0.56 1.32 0.76 Vranje 1.44 1.50 1.65 1.41 1.38 1.99 2.24 1.50 2.80 0.87 4.55 5.47 Zaječar 1.12 1.14 1.82 1.11 1.18 2.58 0.97 1.38 2.81 0.00 2.36 3.29 Zlatibor 1.13 1.09 1.28 1.30 1.15 1.65 0.78 1.22 1.78 0.27 1.83 1.63 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 2011–2040 2041–2070 2071–2100 Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Fig. 10. The share of daily precipitation ≥ 20, 30, 40 and 50 mm in the total annual precipitation in Central Serbia (%); EOBS – E-OBS gridded climatology from European Climate Assessment & Dataset (ECA&D) project, MA1B – modelled values from the EBU-POM model projection for A1B scenario. CONCLUSION precipitation amount increase the changes are more intense. The future decrease of extreme precipitation ranges from This study analyzed future changes in extreme precipitation 10–100%. in Central Serbia using regional climate simulation data Changes in the monthly distribution of R20, R30, R40 and obtained from the EBU-POM model under the SRES-A1B R50 are very similar for all indices. The time of occurrence of scenario. This analysis and other mentioned analyses for the extreme precipitation in Central Serbia is gradually shifting area of Serbia and beyond, show that the climate changes towards the second part of a year. In the period 1961–1990, the during the 21st century have significant impact on extreme most frequent occurrences of extreme precipitation were during precipitation. Therefore, the investigations of climate changes June. Also, the occurrence of these events was more uncertain and their consequences are very important. with the increase in the precipitation amount (they occur during Within this analysis, the number of days with precipitation ≥ several months in the year). In the period 2011–2040, extreme 20, 30, 40 and 50 mm (R20, R30, R40 and R50), their monthly precipitation is expected to be the most frequent in July. In the distribution, and these precipitation shares in the annual middle of the 21st century, they will mostly occur in August precipitation amount are derived for 18 synoptic weather and October, and at the end of the 21st century, October is stations in Central Serbia. Results from the 2011–2040, 2041– expected to be the month with the largest number of extreme 2070 and 2071–2100 periods were compared to results from the precipitation events. In addition to that, in comparison with reference period 1961–1990. historical data, the uncertainty of the occurrence of extreme Consequently, changes are expected in the spatial events will decrease (Figure 9). distribution of R20, R30, R40 and R50. An increase of these In relation to the historical data, the average shares of these indices is expected in the territory, where their minimum extreme precipitation events in the total annual precipitation occurred during the reference period and vice versa. An will be increasing in the future. The increase is 1.2–2.4x, except increase in extreme precipitation is the most expected in the for P50 which is expected to decrease at the beginning of the south of Central Serbia, where Kopaonik, Niš and Vranje stand 21st century. The most noticeable stations are Kopaonik, Niš, out, as well as Loznica (west) and Zaječar (east). The increase Vranje, Zaječar and Loznica. Considering that these stations of these indices is more pronounced compared to their decrease. stood out in terms of an increase in R20, R30, R40 and R50, it The maximum increase was observed for R50, which is was to be expected that they would be also stand out for other as much as 9x. In addition, it is noticed that with the parameters as well. 206 Future changes in extreme precipitation in central Serbia The summary of these results indicates that extreme precipi- extreme precipitation intensity-durationfrequency curves for tation in Central Serbia shifts spatially towards the south of the climate adaptation planning in New York State. Climate territory, and its time of occurrence is autumn (in the past it Services, 5, 23–35. used to be summer). According to that, it is important to note Đurđević, V., 2010. Simulacija klime i klimatskih promena u that the climate and soil conditions are different, and this can jugoistočnoj Evropi korišćenjem regionalnog klimatskog significantly affect the processes of the hydrological cycle. modela. PhD Thesis. University of Belgrade, Belgrade, Ser- Future work could be focused on the combining of these results bia. (In Serbian.) with the analyses of the other hydrological cycle components. Đurđević, V., Rajković, B., 2008.Verification of a coupled In this way, future changes in other hydrological cycle compo- atmosphere-ocean model using satellite observations over nents, and also their joint effects will be determined. In accord- the Adriatic Sea. Annales Geophysicae, 26, 1935–1954. ance with them, the capacities for risk reduction and disaster Fotso-Nguemo, T., Diallo, I., Diakhate, M., Vondou, D., risk management will be strengthened, as well as the integration Mbaye, M., Haensler, A., Gaye, A., Tchawoua, C., 2019. of climate change adaptation practices in the analyzed area. The Projected changes in the seasonal cycle of extreme rainfall application of these analyses in many engineering sciences events from CORDEX simulations over Central Africa. (civil engineering, agriculture, meteorology, etc.) would im- Climatic Change, 155, 339–357. prove the practice and progress of these sciences. Research Gocić, M., Trajković, S., 2013.Spatiotemporal characteristics of efforts such as this one are important for improving the regional drought in Serbia. Journal of Hydrology, 510, 110–123. climate model and for reducing the uncertainties in regional Gocić, M., Trajković, S., 2014. Spatio-temporal patterns of climate modeling. precipitation in Serbia. Theoretical and Applied Climatolo- gy, 117, 419–431. Acknowledgements. The study is supported by the Ministry of Harpa, G.V., Croitoru, A.E., Đurđević, V., Horvath, C., Education, Science and Technological Development, Republic 2019.Future changes in five extreme precipitation indices in of Serbia, No. 43007. We acknowledge the E-OBS dataset from the lowlands of Romania. International Journal of Climatol- the EU-FP6 project UERRA (https://www.uerra.eu) and the ogy, 39, 5720–5740. Copernicus Climate Change Service, and the data providers in Hirabayashi, Y., Kanae, S., Emori, S., Oki, T. and Kimoto, M., the ECA&D project (https://www.ecad.eu). 2008.Global projections of changing risks of floods and droughts in a changing climate. Hydrol. Sci. J., 53, 4, 754– REFERENCES 772. Huang, G., Wang, X., Guo, J., Li, Y., Yang, L., 2018. Future Aktar, N., 2013. Impact of climate change on riverbank erosion. changes in precipitation extremes over China projected by a International Journal of Sciences: Basic and Applied Re- regional climate model ensemble. Atmospheric Environ- search (IJSBAR), 7, 1, 36–42. ment, 188, 142–156. Alexander, L.V., Zhang, X., Peterson, T.C., Caesar, J. Gleason, Janković, A., Podraščanin, Z., Đurđević, V., 2019. Future cli- B., Klein Tank, A.M.G., Haylock, M., Collins, D., Trewin, mate change impacts on residential heating and cooling de- B., Rahimzadeh, F., Tagipour, A., Rupa Kumar, K., Re- gree days in Serbia. Időjárás, Quarterly Journal of the Hun- vadekar, J., Griffiths, G., Vincent, L., Stephenson, D.B., garian Meteorological Service, 123, 351–370. Burn, J., Aguilar, E., Brunet, M., Taylor, M., New, M., Zhai, Jiang, Z., Song, J., Li, L., Chen, W., Wang, Z., Wang, J., 2012. P., Rusticucci, M., Vazquez-Aguirre, J.L., 2006. Global ob- Extreme climate events in China: IPCC-AR4 modelevalua- served changes in daily climate extremes of temperature and tion and projection. Climatic Change, 110, 385–401. precipitation. Journal of Geophysical Research, 111, Jovanović, S., Anđelković, G., Manojlović, S., Samardžić, I., D05109. https://doi.org/10.1029/2005JD006290 Živković, Lj., Šabić, D., Gatarić, D., Džinović, M., 2018. Bocheva, L., Gospodinov, I., Simeonov, P. Marinova, T., Extreme precipitation events in Serbia: Defining the thresh- 2010.Climatological analysis of the synoptic situations caus- old criteria for emergency preparedness. Atmosphere, 9, 5, ing torrential precipitation events in Bulgaria during the pe- 188. riod 1961–2007. In: Global Environmental Change: Chal- Kjellström, E., Ruosteenoja, K., 2007. Present-day and future lenges to Science and Society in Southeastern Europe, pp. precipitation in the Baltic Sea region as simulated in a suite 97–108. of regional climate models. Climatic Change, 81, 281–291. Cavicchia, L., Scoccimarro, E., Gualdi, S., Marson, P., Ahren, Kramer, K., Ware, J., 2019. Counting the cost: 2019: a year of B., Berthou, S., Conte, D., Aquila, A., Drobinski, P., climate breakdown. Christian Aid, London, United King- Đurđević, V., Dubois, C., Gallardo, C., Li, L., Oddo, P., dom. Sanna, A., Torma, C., 2018. Mediterranean extreme precipi- Kržič, A., Tošić, I., Đurđević, V., Veljović, K., Rajković, B., tation: a multi‑model assessment. Climate Dynamics, 51, 2011. Changes in climate indices for Serbia according to the 901–913. SRES-A1B and SRES-A2 scenarios. Climate Research, 49, Cooper, R., 2019. Projection of future precipitation extremes 73–86. across the Bangkok Metropolitan Region. Heliyon, 5, Li, L., Yao, N., Liu, D.L., Song, S., Lin, H., Chen, X., Li, Y., E01678. 2019. Historical and future projected frequency of extreme Cornes, R., Van der Schrier, G., Van den Besselaar, E.J.M., precipitation indicators using the optimized cumulative dis- Jones, P.D., 2018. An ensemble version of the E-OBS tem- tribution functions in China. Journal of Hydrology, 579, Ar- perature and precipitation datasets. J. Geophys. Res. Atmos., ticle Number: 124170. 123. DOI: 10.1029/2017JD028200 MEP, 2017. Second National Communication of the Republic Dawson, C.W., Abrahart, R.J., See, L.M., 2007. HydroTest: of Serbia under the United Nations Framework Convention Aweb-based toolbox of evaluation metrics for the standard- on Climate Change. Ministry of Environmental Protection of the Republic of Serbia, Belgrade, Serbia, 162 p. ised assessment of hydrological forecasts. Environmental http://www.klimatskepromene.rs/wp- Modelling & Software, 22, 1034–1052. content/uploads/2017/09/SNC_eng.pdf DeGaetano, A., Castellano, C., 2017. Future projections of Ranka Erić, Ratko Kadović, Vladimir Đurđević, Vesna Đukić Mihailović, D., Drešković, N., Arsenić, I., Ćirić, V., Đurđević, tation extreme events over Iran. Theoretical and Applied V., Mimić, G., Pap, I., Balaž, I., 2016. Impact of climate Climatology, 126, 775–795 change on soil thermal and moisture regimes in Serbia: An Tan, X., Gan, T., Shao, D., 2017. Effects of persistence and analysis with data from regional climate simulations under large-scale climate anomalies on trends and change points in SRES-A1B. Science of the Total Environment, 571, 398–409. extreme precipitation of Canada. Journal of Hydrology, 550, Nakicenovic, N., Swart, R. (Eds.), 2000. Special Report on 453–465. Emissions Scenarios. A Special Report of Working Group Tramblay, Y., Somot, S., 2018. Future evolution of extreme III of the Intergovernmental Panel on Climate Change. precipitation in the Mediterranean. Climatic Change, 151, Cambridge University Press, Cambridge, United Kingdom 289–302. and New York, NY, USA. Unkašević, M., Tošić, I., 2011. A statistical analysis of the Pongracz, R., Bartholy, J., Miklos, E., 2011. Analysis of pro- daily precipitation over Serbia: trends and indices. Theoreti- jected climate change for Hungary using ensembles simula- cal and Applied Climatology, 106, 69–78. tions. Applied Ecology and Environmental Research, 9, 4, Van der Linden P., Mitchell, J.F.B. (eds.), 2009. ENSEMBLES: 387–398. Climate change and its impacts: Summary of research and re- Republic Hydrometeorological Service of Serbia, 2020. Pre- sults from the ENSEMBLES project. Met Office Hadley cipitation regime in Serbia. Retrieved May 3, 2020. Centre, FitzRoy Road, Exeter EX1 3PB, UK. 160p. http://www.hidmet.gov.rs/eng/meteorologija/klimatologija_ Vuković, A., Vujadinović, M., Rendulić, S., Đurđević, V., padav_rezim.php Ruml, M., Babić, V., Popović, D., 2018. Global warming Schneider, C., Laize, C.L.R., Acreman, M.C., Florke, M., 2013. impact on climate change in Serbia for the period 1961- How will climate change modify river flow regimes in Eu- 2100.Thermal Science, 22, 2267–2280. rope? Hydrol. Earth Syst. Sci., 17, 325–339. Yang, Z., Gao, X., Guo, M., Zhu, Q., Xu, Z., Gao., K., 2020. Soltani, M., Laux, P., Kunstmann, H., Stan, K., Sohrabi, M.M., Temperature dependence of extreme precipitation over Molanejad, M., Sabziparvar, A.A., SaadatAbadi, A., mainland China. Journal of Hydrology, 583, 1245. Ranjbar, F., Rousta, I., Zawar-Reza, P., Khoshakhlagh, F., Soltanzadeh, I., Babu, C.A., Azizi G.H., Martin, M.V., 2016. Received 16 October 2020 Assessment of climate variations in temperature and precipi- Accepted 3 February 2021

Journal

Journal of Hydrology and Hydromechanics – de Gruyter

Published: Jun 1, 2021

Keywords: Future extreme precipitation; Climate changes; Regional climate model; Central Serbia

There are no references for this article.