Drought Vulnerability Index (DVI) in Attica Region, Greece Using Climate Models Upon the Climate Scenarios A1B and A2

Drought Vulnerability Index (DVI) in Attica Region, Greece using climate models upon the climate scenarios A1B and A2

ORIENTGATE Technical report

June 2014

The present work is part of the Greek Pilot Study 4 “Effects of climate change on wetland ecosystems of Attica, Greece” which was carried out under the Thematic Centre “Drought, Water and Coasts” of ORIENTGATE project, which co-financed through the South East Europe Transnational Cooperation Programme.

Working Team

Greek Biotope/Wetland Centre (ΕΚΒΥ): Papadimos Dimitris, Katsavouni Sotiria

Republic Hydrometeorological Service of Serbia, RHMSS: Aleksandra Krzic, Vladimir Djurdjevic

Reference:

Greek Biotope/Wetland Centre (EKBY) and Republic Hydrometeorological Service of Serbia (RHMSS). 2014. Drought Vulnerability Index (DVI) in Attica Region, Greece using climate models upon the climate scenarios A1B and A2. ORIENTGATE Technical Report. Thessaloniki.

C O N T E N T S

p. List of Tables List of Figures List of Maps

INTRODUCTION 1

PART Α

CLIMATE VARIATION IN THE PERIOD 1970-2010 2

1. METHODOLOGICAL APPROACH 3

1.1. Gathering-Processing-Completion of meteorological data 3 1.2. Meteorological data in the region of Attica 4

2. DROUGHT INDEX SPI (Standardized

Precipitation Index) 19

2.1. Calculation of SPI index 20

3. DROUGHT VULNERABILITY INDEX (DVI) 24

i PART Β

ESTIMATION OF CLIMATE VARIATION IN THE PERIOD 2010-2100 27

4. RAINFALL TIME-SERIES OF CLIMATE

CHANGE SCENARIOS Α1Β & Α2 28

4.1. Climate scenario Α1Β 28

4.2. Climate scenario Α2 34

5. CALCULATION OF SPI INDEX 46

6. CALCULATION OF DROUGHT

VULNERABILITY INDEX DVI 49

REFERENCES 52

ANNEX Ι ANNEX ΙΙ

L I S T O F T A B LE S

p. Table 1. Features of meteorological stations in the study area 4 Table 2. Linear correlation coefficients(r) among meteorological stations for monthly rainfall in the region of Attica 7 Table 3. Drought classification based on the index SPI and corresponding occurrence probabilities 20 Table 4. Meteorological stations and corresponding time intervals studied in the context of “Drought - Water shortage Management Plan” and project OrientGate 21 Table 5. Classification of parameters for calculating the DVI into classes and Score 24

L I S T O F F I G U R E S

p. Figure 1. Period of data availability of monthly rainfall in the study area 5

iii Figure 2. Meteorological stations in the region of Attica 6 Figure 3. Annual rainfall and linear trend in the Vyronas station 8 Figure 4. Annual rainfall and linear trend in the Elefsina station 9 Figure 5. Annual rainfall and linear trend in the Elliniko station 9 Figure 6. Annual rainfall and linear trend in the Marathonas station 10 Figure 7. Annual rainfall and linear trend in the Markopoulo station 10 Figure 8. Annual rainfall and linear trend in the Peristeri station 11 Figure 9. Annual rainfall and linear trend in the Tatoi station 11 Figure 10. Annual rainfall and linear trend in the Nea Filadelfeia station 12 Figure 11. Annual rainfall and linear trend in the Chalandri station 12 Figure 12. Annual rainfall and linear trend in the Piraeus station 13 Figure 13. Average monthly rainfall in Vyronas station for the period 1970-2010 14 Figure 14. Average monthly rainfall in Elefsina station for the period 1970-2010 14 Figure 15. Average monthly rainfall in Elliniko station for the period 1970-2010 15 Figure 16. Average monthly rainfall in Marathonas station for the period 1970-2010 15 Figure 17. Average monthly rainfall in Markopoulo station for the period 1970-2010 16 Figure 18. Average monthly rainfall in Peristeri station for the period 1970-2010 16 Figure 19. Average monthly rainfall in Tatoi station for the period 1970- 2010 17 Figure 20. Average monthly rainfall in Nea Filadelfeia station for the period 1970-2010 17 Figure 21. Average monthly rainfall in Chalandri station for the period 1970-2010 18 Figure 22. Average monthly rainfall in Piraeus station for the period 1970-2010 18 Figure 23. Drought events, intensity and duration in the meteorological stations of Attica for the historical period (1970 - 2010) 23

iv Figure 24. Annual rainfall and linear trend in the Vyronas station for the scenario A1B 28 Figure 25. Annual rainfall and linear trend in the Elefsina station for the scenario A1B 29 Figure 26. Annual rainfall and linear trend in the Elliniko station for the scenario A1B 29 Figure 27. Annual rainfall and linear trend in the Marathonas station for the scenario A1B 30 Figure 28. Annual rainfall and linear trend in the Markopoulo station for the scenario A1B 30 Figure 29. Annual rainfall and linear trend in the Peristeri station for the scenario A1B 31 Figure 30. Annual rainfall and linear trend in the Tatoi station for the scenario A1B 31 Figure 31. Annual rainfall and linear trend in the Nea Filadelfeia station for the scenario A1B 32 Figure 32. Annual rainfall and linear trend in the Chalandri station for the scenario A1B 32 Figure 33. Annual rainfall and linear trend in the Piraeus station for the scenario A1B 33 Figure 34. Annual rainfall and linear trend in the Vyronas station for the scenario A2 34 Figure 35. Annual rainfall and linear trend in the Elfesina station for the scenario A2 35 Figure 36. Annual rainfall and linear trend in the Elliniko station for the scenario A2 35 Figure 37. Annual rainfall and linear trend in the Marathonas station for the scenario A2 36 Figure 38. Annual rainfall and linear trend in the Markopoulo station for the scenario A2 36 Figure 39. Annual rainfall and linear trend in the Peristeri station for the scenario A2 37

v Figure 40. Annual rainfall and linear trend in the Tatoi station for the scenario A2 37 Figure 41. Annual rainfall and linear trend in the Nea Filadelfeia station for the scenario A2 38 Figure 42. Annual rainfall and linear trend in the Chalandri station for the scenario A2 38 Figure 43. Annual rainfall and linear trend in the Piraeus station for the scenario A2 39 Figure 44. Average monthly rainfall in Vyronas station for the climate scenarios A1B & A2 40 Figure 45. Average monthly rainfall in Elefsina station for the climate scenarios A1B & A2 40 Figure 46. Average monthly rainfall in Elliniko station for the climate scenarios A1B & A2 41 Figure 47. Average monthly rainfall in Marathonas station for the climate scenarios A1B & A2 41 Figure 48. Average monthly rainfall in Markopoulo station for the climate scenarios A1B & A2 42 Figure 49. Average monthly rainfall in Peristeri station for the climate scenarios A1B & A2 42 Figure 50. Average monthly rainfall in Tatoi station for the climate scenarios A1B & A2 43 Figure 51. Average monthly rainfall in Nea Filadelfeia station for the climate scenarios A1B & A2 43 Figure 52. Average monthly rainfall in Chalandri station for the climate scenarios A1B & A2 44 Figure 53. Average monthly rainfall in Piraeus station for the climate scenarios A1B & A2 44 Figure 54. Annual rainfall for the historical period (1970-2010) and for the climate scenarios Α1Β & Α2 (2010-2100) 45 Figure 55. Drought events, intensity and duration in the meteorological stations of Attica for the climate scenario A1B (2010 – 2100) 47

vi Figure 56. Drought events, intensity and duration in the meteorological stations of Attica for the climate scenario A2 (2010 – 2100) 48 Figure 57. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenarios A1B & A2 (2010- 2100) 49

L I S T O F M A P S

σελ. Map 1. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the historical period 1970 – 2010 26 Map 2. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenario A1B (2010-2100) 50 Map 3. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenario A2 (2010-2100) 51

vii INTRODUCTION

A special study entitled "Drought and Water shortage Management Plan" was commissioned within the preparation of the River Basin Management Plan of Attica Water District (Ministry of Environment 2013). Its main purpose was to: a) quantify the effects of drought and water shortage in the water district of Attica, b) to examine possible methodologies for predicting future incidents and c) to propose remedies for various levels of risk. This essay aims to strengthen the effectiveness of the above "Drought and Water shortage Management Plan" of the Attica Water District, by enhancing the forecasting process of future drought incidents. Specifically: a) Time series of monthly rainfall were created, in each meteorological station of the study area, for the next 100 years (2010-2100) by using climate models upon the climate scenarios A1B and A2. b) The same drought index (SPI-12) and drought vulnerability index (DVI), as well as the same characterization limits of these indices, were used in order the results to be compatible with those in the "Drought and Water shortage Management Plan".

Following the above procedure, the first step was to collect rainfall data of the meteorological stations located within Attica Prefecture and Water District and then to process and complete the rainfall data of each station. In order to ensure the drought index calculation accuracy, it was attempted that the duration of the historical time series data is above 30 years (1970-2010). The next step was to calculate SPI-12 and DVI indices, for the historical time series data, following the same methodology as the "Drought and Water shortage Management Plan". In this context there was a cross check with the results of the "Drought and Water shortage Management Plan". Finally, the SPI-12 and DVI indices were calculated based on time series of monthly rainfall data for future years and the future incidents of drought, their duration and intensity were identified.

PART A

CLIMATE VARIATION IN THE PERIOD 1970-2010

2 1. METHODOLOGICAL APPROACH

1.1. GATHERING - PROCESSING - COMPLETION OF METEOROLOGICAL DATA

The main criteria for the selection of the monitoring stations were: a) the available information covering the period (1970-2010), and (b) the geographical distribution, covering evenly as much of the study area as possible. Sources of raw data were monthly log sheets of the meteorological observations of the Ministry of Environment, Energy and Climate Change (MEECC.) and the National Meteorological Service (NMS). The primary piece of information posted in monthly steps for rainfall and temperature was obtained using Excel and Hydrognomon software (www.hydroscope.gr). The primary recordings of meteorological parameters were the following processing stages: • Filling of gaps (based on monthly values), which are usually due to instrument damage or observers omissions. • Time projection of monthly values in cases where the stations operated in shorter periods than those of interest. Simple linear regression (Koutsoyiannis and Xanthopoulos 1999, Papamichail 2001, Cleaver 2006) was used for data projection. According to this method, the projected value Yi is estimated from the corresponding value Xi (independent) of an adjacent reference station for the period of lack of data by station Y, based on the linear relationship:

Yi = aXi +b where Yi: the value to complete Xi: the value of the reference station, i: the relevant month, a and b: parameters are estimated so as to minimize the squared error of the estimate. The degree of suitability of the method is determined by the coefficient of linear correlation (r) between the values of two stations whose values range in the interval (-1, 1). To be statistically significant, a correlation coefficient r must be in absolute value greater than the critical value: r > 0.7. In this way, the correlation of monthly rainfall and temperature between stations was estimated.

3 1.2. METEOROLOGICAL DATA IN THE REGION OF ATTICA Table 1 below presents the meteorological stations, the altitude, the managing body of the station and the meteorological parameters. In the study area of Attica, the available meteorological stations that were taken into account are given in Figure 1.

Table 1: Features of meteorological stations in the study area

α/α Name station Altitude (m) Managing Meteorological Body parameters 1 Tatoi 237,9 NMS rainfall, temperature 2 Elefsina 30 NMS rainfall, temperature 3 Elliniko 10 NMS rainfall, temperature 4 Marathonas - NMS rainfall, temperature 5 Nea Filadelfeia 136 NMS rainfall, temperature 6 Piraeus 7 NMS rainfall, temperature 7 Paiania 152 NMS rainfall, temperature 8 Megara - NMS rainfall, temperature 9 Spata 67,6 NMS rainfall, temperature 10 Markopoulo 83,6 MEECC rainfall, temperature 11 Peristeri 75,4 MEECC rainfall 12 Vyronas 226,4 MEECC rainfall, temperature 13 Chalandri 189,3 MEECC rainfall 14 Piraeus - MEECC rainfall 15 3rd Cemetery 67,2 MEECC rainfall Nikaias-Egaleo 16 Agios Ierotheos - MEECC rainfall

4 VYRONAS AIGALEO MARKOPOULO PERISTERI CHALANDRI PIRAEUS AG.IEROTHEOS ELEFSINA ELLINIKO MARATHONAS MEGARA NEA FILADELFEIA PAIANIA PIRAEUS SPATA TATOI Monthly Rainfall

17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 07-Jan-52 10-Feb-53 17-Mar-54 21-Apr-55 25-May-56 29-Jun-57 03-Aug-58 07-Sep-59 11-Oct-60 15-Nov-61 20-Dec-62 24-Jan-64 27-Feb-65 03-Apr-66 08-May-67 11-Jun-68 16-Jul-69 20-Aug-70 24-Sep-71 28-Oct-72 02-Dec-73 06-Jan-75 10-Feb-76 16-Mar-77 20-Apr-78 25-May-79 28-Jun-80 02-Aug-81 06-Sep-82 11-Oct-83 14-Nov-84 19-Dec-85 23-Jan-87 27-Feb-88 02-Apr-89 07-May-90 11-Jun-91 15-Jul-92 19-Aug-93 23-Sep-94 28-Oct-95 01-Dec-96 05-Jan-98 09-Feb-99 15-Mar-00 19-Apr-01 24-May-02 28-Jun-03 01-Aug-04 05-Sep-05 10-Oct-06 14-Nov-07 18-Dec-08 22-Jan-10 26-Feb-11 01-Apr-12 06-May-13

Figure 1: Period of data availability of monthly rainfall in the study area

From the above figure it is understood that in all stations presented several shortcomings either in individual months or into broader intervals covering different periods. In 6 of these stations, the observation period was too short and therefore chosen not to use their data (stations: Megara, Spata, Agios Ierotheos, Aigaleo, Paiania, Piraeus (MEECC). For the remaining 10 stations, to obtain complete time series of rainfall data, the available data were supplemented and projected to cover the period January 1970 – December 2010.

Figure 2: Meteorological stations in the region of Attica

The completion and projection of the data took place on a monthly scale. Initially estimated the correlation of monthly rainfall all stations between them. Correlation coefficients obtained from this analysis are given in table 2. The completion and projection of the data of each station was based on the data of an adjacent reference station with the greatest correlation.

6 Table 2: Linear correlation coefficients(r) among meteorological stations for monthly rainfall in the region of Attica

Correlation coefficient (r) Meteorologi Elefsi Ellini Marat Markop Perist Tatoi Nea Chala Piraeus cal stations na ko honas oulo eri Filadelfei ndri a Vyronas 0,758 0,865 0,854 0,814 0,778 0,716 0,786 0,820 0,763 Elefsina 0,853 0,866 0,786 0,801 0,750 0,867 0,819 0,777 Elliniko 0,884 0,907 0,825 0,793 0,898 0,898 0,856 Marathonas 0,852 0,832 0,790 0,858 0,828 0,881 Markopoulo 0,785 0,788 0,843 0,882 0,783 Peristeri 0,754 0,878 0,846 0,760 Tatoi 0,831 0,843 0,672 Nea 0,907 0,809 Filadelfeia Chalandri 0,811

• Gap filling of station Markopoulo Data of the Elliniko station were used in order to supplement monthly rainfall values of Markopoulo station, due to their strong positive correlation. • Gap filling of station Elliniko Data of the Markopoulo and Vyrona station were used in order to supplement monthly rainfall values of Elliniko station, due to their strong positive correlation. • Gap filling of station Vyronas Data of the Elliniko station were used in order to supplement monthly rainfall values of Vyronas station, due to their strong positive correlation. • Gap filling of station Chalandri Data of the Nea Filadelfeia and Elliniko station were used in order to supplement monthly rainfall values of Chalandri station, due to their strong positive correlation. • Gap filling of station Nea Filadelfeia Data of the Chalandri and Elliniko station were used in order to supplement monthly rainfall values of Nea Filadelfeia station, due to their strong positive correlation. • Gap filling of station Elefsina Data of the Nea Filadelfeia station were used in order to supplement monthly rainfall values of Elefsina station, due to their strong positive correlation.

7 • Filling station Piraeus Data of the Elliniko station were used in order to supplement monthly rainfall values of Piraeus station, due to their strong positive correlation. • Filling station Tatoi Data of the Chalandri station were used in order to supplement monthly rainfall values of Tatoi station, due to their strong positive correlation. • Filling station Marathonas Data of the Elliniko station were used in order to supplement monthly rainfall values of Marathona station, due to their strong positive correlation. • Filling station Peristeri Data of the Nea Filadelfeia station were used in order to supplement monthly rainfall values of Peristeri station, due to their strong positive correlation.

In Figures 3-12 is shown graphically the variation of the annual rainfall of stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus and the linear trend respectively.

Annual rainfall Linear trend

900

800

700

600

500

400

300 200

Annual Annual rainfall (mm) Vyrona 100

2 0 970 74 976 978 980 82 84 986 88 90 92 94 96 98 00 0 04 06 1 1 1972 19 1 1 1 19 19 1 19 19 19 19 19 19 20 20 20 20 2008 20 Years

Figure 3: Annual rainfall and linear trend in the Vyronas station

8 Annual rainfall Linear trend

800

700

600

500

400

300

200

Annual Annual rainfall (mm) Elefsina 100

70 76 78 84 86 92 94 00 02 06 08 10 19 1972 1974 19 19 1980 1982 19 19 1988 1990 19 19 1996 1998 20 20 2004 20 20 20 Years

Figure 4: Annual rainfall and linear trend in the Elefsina station

Annual rainfall Linear trend

800

700

600

500

400

300

200

100

Annual Elliniko rainfall (mm) 0

0 994 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 199 1992 1 1996 1998 2000 2002 2004 2006 2008 2010 Years

Figure 5: Annual rainfall and linear trend in the Elliniko station

Annual rainfall Linear trend

800

700

600

500

400

300

200

100 Annual Annual rainfall (mm) Marathona

6 2 970 972 974 976 978 982 984 98 990 992 994 996 998 000 00 008 010 1 1 1 1 1 1980 1 1 1 1988 1 1 1 1 1 2 2 2004 2006 2 2 Years

Figure 6: Annual rainfall and linear trend in the Marathonas station

Annual rainfall Linear trend 800

700

600

500

400

300

200

100 Annual rainfall (mm) Markopoulo

2 4 8 0 2 4 8 0 92 4 6 98 0 2 4 8 0 97 97 98 00 00 1970 1 197 1976 1 198 1 198 1986 198 199 19 199 199 19 200 2 200 2006 2 201 Years

Figure 7: Annual rainfall and linear trend in the Markopoulo station

10 Annual rainfall Linear trend

1000 900 800 700 600 500 400 300 200 Annual rainfall (mm) Peristeri 100 0

0 4 8 0 4 6 8 0 4 6 8 0 2 4 6 8 0 8 9 0 0 197 1972 197 1976 197 198 1982 198 198 19 199 1992 199 199 19 200 200 20 200 20 201 Years

Figure 8: Annual rainfall and linear trend in the Peristeri station

Annual rainfall Linear trend

900

800

700

600

500

400

300

200

100 Annual rainfall (mm) Tatoi

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Years

Figure 9: Annual rainfall and linear trend in the Tatoi station

11 Annual rainfall Linear trend

800

700

600

500

400

300

200

100

0 Annual Annual rainfall (mm) Nea Filadelfeia

1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Years

Figure 10: Annual rainfall and linear trend in the Nea Filadelfeia station

Annual rainfall Linear trend

800

700

600

500

400

300

200

100 Annual rainfall (mm) Chalandri 0

0 4 8 2 6 90 94 98 02 06 10 197 1972 197 1976 197 1980 198 1984 198 1988 19 1992 19 1996 19 2000 20 2004 20 2008 20 Years

Figure 11: Annual rainfall and linear trend in the Chalandri station

12 Annual rainfall Linear trend 800

700

600

500

400

300

200

100 Annual (mm) Piraeusrainfall 0

2 8 4 0 8 4 0 70 76 82 88 94 96 02 08 19 197 1974 19 197 1980 19 198 1986 19 199 1992 19 19 199 2000 20 200 2006 20 201 Years

Figure 12: Annual rainfall and linear trend in the Piraeus station

The average annual rainfall of stations Vyrona, Elefsina, Elliniko, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus is 407,0 mm, 357,8 mm, 371,8 mm, 453,9 mm, 387,1 mm, 447,0 mm, 410,5 mm, 436,9 mm, 337,9 mm respectively, after the completion of missing values for the period January 1970 to December 2010, while the Marathon station amounted to 433,0 mm for the period January 1988 to December 2010. In figures 13 to 22 the monthly average values of rainfall for stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus respectively after their completion and for the above-mentioned period.

13 Vyronas

80,0

70,0

60,0

50,0

40,0

Rainfall (mm) Rainfall 30,0

20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 13: Average monthly rainfall in Vyronas station for the period 1970-2010

Elefsina

70,0

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall

20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 14: Average monthly rainfall in Elefsina station for the period 1970-2010

14 Elliniko

70,0

60,0

50,0

40,0

30,0 Rainfall(mm) 20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 15: Average monthly rainfall in Elliniko station for the period 1970-2010

Marathonas

80,0

70,0

60,0

50,0

40,0

Rainfall (mm) Rainfall 30,0

20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 16: Average monthly rainfall in Marathonas station for the period 1970-2010

15 Markopoulo

90,0 80,0 70,0 60,0 50,0 40,0 Rainfall (mm) Rainfall 30,0 20,0 10,0

0,0 JFMAMJJASOND

Months

Figure 17: Average monthly rainfall in Markopoulo station for the period 1970-2010

Peristeri

70,0

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall

20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 18: Average monthly rainfall in Peristeri station for the period 1970-2010

16 Tatoi

90,0 80,0 70,0 60,0 50,0 40,0 Rainfall (mm) Rainfall 30,0 20,0 10,0 0,0 JFMAMJJASOND

Months

Figure 19: Average monthly rainfall in Tatoi station for the period 1970-2010

Nea Filadelfeia

80,0

70,0 60,0

50,0

40,0

Rainfall(mm) 30,0 20,0

10,0 0,0 JFMAMJJASOND

Months

Figure 20: Average monthly rainfall in Nea Filadelfeia station for the period 1970-2010

17 Chalandri

80,0

70,0

60,0

50,0

40,0

Rainfall (mm) Rainfall 30,0

20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 21: Average monthly rainfall in Chalandri station for the period 1970-2010

Piraeus

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND

Months

Figure 22: Average monthly rainfall in Piraeus station for the period 1970-2010

18 From the above figures, it appears that the maximum rainfall occurs during the month of December for all of the meteorological stations except the Marathon station where the maximum is observed in the month of November. Regarding the minimum of rainfall observed during the month of June for the stations at Vironas, Elliniko, Peristeri, Tatoi, during the month of July for the stations at Markopoulo and Pireus and the month of August for stations at Elefsina, Marathona, Nea Filadelfeia and Chalandri.

2. DROUGHT INDEX SPI (Standardized Precipitation Index)

The drought index SPI was developed by McKee et al (1993) and it is based on rainfall observations for a given period. The longer this period, the more reliable are the results. According to McKee et al. (1993) the period must be at least 30 years. The calculation of the SPI is obtained by subtracting the average price of the period investigated from the rainfall, and dividing the result by the standard deviation. However, because the rainfall has no normal distribution there is a setting that allows SPI to have normal distribution. Therefore, the average value of the SPI for a time period and for a specific region is 0 and the standard deviation is 1. The normalization of the SPI index has the advantage that wetter and drier portions can be represented in the same way. The nature of the SPI allows the identification of a rare incident of drought or an extremely wet episode that can occur in any area and at any time, provided that there are sufficient data. The SPI can be calculated for different time scales. The McKee et al (1993) calculated the SPI for 3-, 6-, 12-, 24- and 48- month intervals. The small time scales are sufficient to assess and characterize the agricultural and the meteorological drought while for the hydrological drought larger scales are required. It has been shown that the 3- or 6- months SPI is related to soil moisture (Sims et al., 2002; Ji and Peters, 2003), which is crucial for vegetation and agriculture and thus the agricultural as well as the meteorological drought is controlled (McKee et al. 1993; Hayes et al. 1999), while the 12- months SPI is related to water resources (reservoirs, rivers, groundwater) and therefore the hydrological drought is controlled (Szalai et al. 2000; Hayes et al. 1999). The SPI is a dimensionless index; where positive values indicate rainfall higher than 50 % of the observations and respectively, negative values indicate rainfall lower than 50 % of the observations. This normalization of SPI has the advantage that wetter and drier incidents can be represented in the same way. The characterization of drought incidents, based on the SPI grading scale given by McKee et al. (1993), are given in Table 3 (Apostolou 2010). The table also contains the corresponding probabilities of each category’s occurrence. The criteria for an incident of drought at any time scale were also set. An episode of drought begins when the SPI index gets a negative value, continues with negative values and becomes intense when the SPI index gets a negative value less than or equal to -1.5. The episode ends when the SPI index gets a positive value. So each drought incident has a duration, which is determined by a start, an end and intensity for each month that the incident continues.

19 Table 3. Drought classification based on the index SPI and corresponding occurrence probabilities

SPI Value Category Probability % SPI ≥ + 2.00 Extremely wet 2.3 1.50 ≤ SPI ≤ 1.99 Severely wet 4.4 1.00 ≤ SPI ≤ 1.49 Moderately wet 9.2 0.00 ≤ SPI ≤ 0.99 Near normal 34.1 - 0.99 ≤ SPI ≤ 0.00 Near normal 34.1 - 1.49 ≤ SPI ≤ - 1.00 Moderately dry 9.2 - 1.99 ≤ SPI ≤ - 1.50 Severely dry 4.4 SPI ≤ - 2.00 Extremely dry 2.3 Source: Apostolou (2010)

2.1. CALCULATION OF SPI INDEX In the “Drought and Water Shortage Management Plan” drawn up for the Basin Management Plan of Attica, SPI indicator was used for 12 months period (SPI12). For the calculation of SPI12 monthly rainfall data were used, collected from 11 meteorological stations of the specific region through the period from 1980 to 2010. The stations and the time periods studied are presented at Table 4. In the context of OrientGate project, the same stations were used except the Asteroskopeio and the Marathon dam’s stations. Additionally, the duration of the studied time-series of the monthly rainfall values of each station, was chosen to be grater than 30 years (1970 – 2010), to ensure the reliability of the results (Table 4) regarding the calculation of the drought indicator.

20 Table 4: Meteorological stations and corresponding time intervals studied in the context of “Drought - Water shortage Management Plan” and project OrientGate a/a Meteorological Drought – Water shortage OrientGate station Management Plan Period from to from to 1 Asteroskopio 1980 – 81 2008 – 09 2 Vyronas 1980 – 81 2009 – 10 1970 – 71 2009 – 10 3 Elefsina 1980 – 81 2000 – 01 1970 – 71 2009 – 10 4 Elliniko 1980 – 81 2005 – 06 1970 – 71 2009 – 10 5 Marathonas 1980 – 81 2000 – 01 1988-1989 2009-2010 6 Markopoulo 1980 – 81 2008 – 09 1970 – 71 2009 – 10 7 Peristeri 1980 – 81 2009 – 10 1970 – 71 2009 – 10 8 Tatoi 1980 – 81 2000 – 01 1970 – 71 2009 – 10 9 Filadelfeia 1980 – 81 2005 – 06 1970 – 71 2009 – 10 10 Dam of Marathona 1980 – 81 2000 – 01 11 Chalandri 1980 – 81 2000 – 01 1970 – 71 2009 – 10 12 Piraeus 1970 – 71 2009 – 10

For the detailed calculation of SPI-12 index, it was followed the methodology developed by McKee et al. (1993) as described in «Drought-Water shortage Management Plan» (MEECC 2013). In accordance with the above, the SPI index is calculated from the relationship:

 c + tc + tc 2  z = SPI −= t − 0 1 2   2 3  for 0 < H(x) < 0.5  1+ 1td + 2td + 3td 

 c + tc + tc 2  z = SPI += t − 0 1 2  for 0.5 < H(x) < 1.0  2 3   1+ 1td + 2td + 3td 

21 where:

 1  =   Η t ln 2  for 0 < (x) < 0.5  (H (x)) 

 1  =   Η t ln 2  for 0.5 < (x) < 1.0  0.1( − H (x))  and c0 = 2.515517 c1 = 0.802853 c2 = 0.010328 d1 = 1.432788 d2 = 0.189269 d3 = 0.001308

H(x) = q + (1 – q)G(x) and

x 1 ) G x)( = t α −1e−t dt Γ α) ∫ )( 0 where q is the possibility of zero-rainfall. If m is the amount of the zero-rainfall occasions during the time-series, then q can be calculated from the equation q=m/n. Other approaches for handling the zero-rainfall occasions is by replacing them with a very small quantity (i.e. 1 mm) (Loukas and Vasiliades 2004) These relations were used through Excel software and SPI-12 index values were calculated for all the stations. In Annex I, SPI-12 values are graphically presented for the time period between 1970 and 2010. In Figure 23 the drought events, intensity and duration are given, as calculated for each station for the historical period (1970 – 2010).

Vyronas Tatoi Piraeus Peristeri Markopoulo Marathonas Nea Filadelfeia Elliniko Elefsina Chalandri

30 Drought Magnitude 20

0 Jul-75 Jul-01 Jan-70 Jun-71 Oct-72 Jan-81 Oct-83 Jun-86 Jan-07 Oct-09 Sep-94 Jan-96 Jun-97 Oct-98 Apr-78 Apr-04 Feb-85 Dec-91 Feb-11 Nov-76 Aug-79 Nov-87 Aug-90 Nov-02 Aug-05 Mar-74 Mar-89 Mar-00 May-82 May-08 May-93

Duration (months)

Figure 23: Drought events, intensity and duration in the meteorological stations of Attica for the historical period (1970 - 2010)

23 3. DROUGHT VULNERABILITY INDEX (DVI)

Within the “Drought and Water Shortage Management Plan” project, for the further valuation of the drought vulnerability in the regions subjected in the Basin Management Plan of Attica and the determination of the high risk zones, the Drought Vulnerability Index (DVI) was developed and used (MEECC 2013). According to the writers: “… this indicator aims to the overlaying of all the drought characteristics (duration, intensity, occurrence frequency), as they are analyzed and calculated with the SPI-12 indicator”. In this current study, indicator DVI and its calculation method have been used, as well as the characterization boundaries of the drought vulnerability. According to the methodology, described in the “Drought and Water Shortage Management Plan” (MEECC 2013), DVI indicator is calculated with the following equation: DVIi = 0.25 * Score (drought events) + 0.25 * Score (drought events with duration > 24 months) + 0.25 * Score (DM_max) + 0.25 * Score (Duration_max) where: DVIi: Drought Vulnerability Index for station i ° drought events: Number of drought events observed in the investigation period ° drought events with duration > 24 months: Number of drought events with duration longer than 24 months observed in the investigation period ° DM_max: Maximum size of drought event in the investigation period ° Duraton_max: Maximum duration of drought events in the investigation period

The above parameters were calculated for each station (i) based on the index SPI-12, and on the classification in 4 classes where the score receives values from 1-4 (1: low Vulnerability, 4: high Vulnerability) (Table 5).

Table 5. Classification of parameters for calculating the DVI into classes and Score Drought Drought events Maximum size Maximum Score incidents with of drought duration of duration>24 drought event months 0-10 0-4 0 – 30 0 – 20 1 11-20 5-6 31 – 40 21 – 30 2 21-30 7-8 41 – 50 31 – 40 3 >31 >9 > = 51 >=41 4

24 Based on the SPI index: ° As starting time of a drought event is set the moment (month in the current analysis) when SPI-12 takes a negative value, and reaches subsequently the -1 value for the following months, without getting any positive value in the meantime. ° As ending time of the event, the month is set when the first positive value is observed. ° In the case that the SPI-12 Index is negatively valued, for several consequent months without reaching -1 value (negative values lower than -1), then this event is not considered as drought event, but simply a dryer period than the average. ° The size of a drought event DM (Drought Magnitude), is defined as the absolute value of the sum of the time-series SPI-12, during the months that the drought event lasted. The DVI index was calculated for each meteorological station, for the historic period 1970 – 2010. The optical imaging is shown in the Maps 1.

Map 1: Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the historical period 1970 – 2010

PART B

ESTIMATION OF CLIMATE VARIATION IN THE PERIOD 2010- 2100

4. RAINFALL TIME-SERIES OF CLIMATE CHANGE SCENARIOS Α1Β & Α2 The forecasting climate model (Kržič et al. 2011) was applied to the climate change scenarios Α1Β and Α2 in the 10 meteorological stations of the study area for the period 2010-2100. The characteristics of scenarios Α1Β and Α2 are summarised as follows: Scenario Α1Β: Rapid economic growth. Particularly strong consumption of energy, but also spread of new and efficient technologies. Use of both fossil fuels and alternative energy sources. Small changes in land uses. Rapid rise in global population by 2050 and gradual decline thereafter. Large increase in CO2 concentration in the atmosphere, to 720 ppm by 2100. Scenario Α2: Moderate increase in global average per capita income. Particularly strong energy consumption. Rapid rise in global population. Slow and fragmented technological change, and modest to major changes in land uses. Rapid rise in CO2 concentration in the atmosphere, to 850 ppm by 2100.

4.1. Climate scenario Α1Β Figures 24 to 33 show the variation and the linear trend of annual rainfall in the stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus for the period 2010-2100.

Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200

Annual Annual rainfall (mm) Vyrona 100

18 22 26 30 4 8 2 6 66 70 74 78 2 6 0 2010 2014 20 20 20 20 203 203 204 204 2050 2054 2058 2062 20 20 20 20 208 208 209 2094 2098 Years

Figure 24. Annual rainfall and linear trend in the Vyronas station for the scenario A1B

28 Annual rainfall-Senario A1B Linear trend 700

600

500

400

300

200

100 Annual Elefsina rainfall (mm)

2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2058 2062 2066 2070 2074 2078 2082 2086 2090 2094 2098 Years

Figure 25. Annual rainfall and linear trend in the Elefsina station for the scenario A1B

Annual rainfall-Senario A1B Linear trend

800

700

600

500

400

300

200 Annual Annual rainfall (mm) Elliniko 100

0 8 2 6 0 4 8 2 6 0 4 8 2 6 4 8 2 6 0 4 8 201 2014 201 202 202 203 203 203 204 204 205 205 205 206 206 2070 207 207 208 208 209 209 209 Years

Figure 26. Annual rainfall and linear trend in the Elliniko station for the scenario A1B

29 Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200

Annual Marathona (mm) rainfall 100

2 2 10 14 18 66 70 74 78 20 20 20 202 2026 2030 2034 2038 2042 2046 2050 2054 2058 2062 20 20 20 20 208 2086 2090 2094 2098 Years

Figure 27. Annual rainfall and linear trend in the Marathonas station for the scenario A1B

Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200

Annual rainfall (mm) Markopoulo 100

8 2 5 6 0 0 2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2 2 2066 2070 2074 2078 2082 2086 2090 2094 2098 Years

Figure 28. Annual rainfall and linear trend in the Markopoulos station for the scenario A1B

30 Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Peristeri 100

8 6 4 6 4 0 8 6 14 22 30 42 50 58 66 74 82 94 2010 20 201 20 202 20 203 2038 20 204 20 205 20 2062 20 207 20 207 20 208 2090 20 2098 Years

Figure 29. Annual rainfall and linear trend in the Peristeri station for the scenario A1B

Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Tatoi 100

8 2 5 6 0 0 2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2 2 2066 2070 2074 2078 2082 2086 2090 2094 2098 Years

Figure 30. Annual rainfall and linear trend in the Tatoi station for the scenario A1B

31 Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200

100 Annual Annual rainfall (mm) Nea Filadelfeia

6 4 2 26 34 42 50 58 018 2010 2014 2 2022 20 2030 20 2038 20 2046 20 2054 20 2062 206 2070 207 2078 208 2086 2090 2094 2098 Years

Figure 31. Annual rainfall and linear trend in the Nea Filadelfeia station for the scenario A1B

Annual rainfall-Senario A1B Linear trend

800

700

600

500

400

300

200 Annual Annual rainfall (mm) Chalandri 100

0 4 8 2 6 0 8 2 6 0 4 8 2 6 0 4 8 2 6 0 4 2 6 201 201 201 202 20 203 2034 203 204 204 205 205 205 20 206 207 207 207 208 208 209 209 2098 Years

Figure 32. Annual rainfall and linear trend in the Chalandri station for the scenario A1B

32 Annual rainfall-Senario A1B Linear trend 800

700

600

500

400

300

200

Annual Piraeusrainfall (mm) 100

2 6 6 0 2 2 8 9 0 0 0 0 2010 2014 2018 2 2 2030 2034 2038 2042 2046 2050 2054 2058 2062 2066 2070 2074 2078 2082 2 2 2094 2098

Years

Figure 33. Annual rainfall and linear trend in the Piraeus station for the scenario A1B

The mean annual rainfall in the stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Pyraeus is 298,5 mm, 293,8 mm, 289,0 mm, 349,9 mm, 345,1 mm, 280,1 mm, 351,9 mm, 332,4 mm, 362,8 mm, 250,9 mm, respectively, for the period January 2010 to December 2100 for the climate scenario A1B.

33 4.2. Climate scenario Α2 Figures 34 to 43 show the variation and the linear trend of annual rainfall in the stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus for the period 2010-2100.

Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200

Annual rainfall Vyrona(mm) 100

0 4 8 2 6 0 4 8 2 6 0 4 8 2 6 0 4 8 2 6 0 4 8 201 201 201 202 202 203 203 203 204 204 205 205 205 206 206 207 207 207 208 208 209 209 209 Years

Figure 34. Annual rainfall and linear trend in the Vyronas station for the scenario A2

34 Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Elefsina 100

38 42 46 50 54 58 62 6 70 4 8 2 6 4 06 07 07 08 08 09 2010 2014 2018 2022 2026 2030 2034 20 20 20 20 20 20 20 2 20 2 2 2 2 2090 2 2098 Years

Figure 35. Annual rainfall and linear trend in the Elfesina station for the scenario A2

Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Elliniko rainfall (mm)

100

18 22 58 78 038 054 094 2010 2014 20 20 2026 2030 2034 2 2042 2046 2050 2 20 2062 2066 2070 2074 20 2082 2086 2090 2 2098 Years

Figure 36. Annual rainfall and linear trend in the Elliniko station for the scenario A2

35 Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Marathonarainfall (mm) 100

6 58 78 038 094 2010 2014 2018 2022 202 2030 2034 2 2042 2046 2050 2054 20 2062 2066 2070 2074 20 2082 2086 2090 2 2098 Years

Figure 37. Annual rainfall and linear trend in the Marathonas station for the scenario A2

Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200

Annual Annual rainfall (mm) Markopoulo 100

2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2058 2062 2066 2070 2074 2078 2082 2086 2090 2094 2098

Years Figure 38. Annual rainfall and linear trend in the Markopoulo station for the scenario A2

36 Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Peristeri

100

2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2058 2062 2066 2070 2074 2078 2082 2086 2090 2094 2098

Years Figure 39. Annual rainfall and linear trend in the Peristeri station for the scenario A2

Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Tatoi

100

4 010 014 018 022 026 030 03 038 042 046 050 054 058 062 066 070 078 082 086 090 094 098 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2074 2 2 2 2 2 2 Years

Figure 40. Annual rainfall and linear trend in the Tatoi station for the scenario A2

37 Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200

Annual Annual rainfall (mm) Nea Filadelfeia 100

8 2 5 6 0 0 2010 2014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2 2 2066 2070 2074 2078 2082 2086 2090 2094 2098 Years

Figure 41. Annual rainfall and linear trend in the Nea Filadelfeia station for the scenario A2

Annual rainfall-Senario A2 Linear trend 800

700

600

500

400

300

200 Annual Annual rainfall (mm) Chalandri

100

10 14 18 22 26 30 42 46 50 54 58 62 66 70 74 78 94 98 20 20 20 20 20 20 2034 2038 20 20 20 20 20 20 20 20 20 20 2082 2086 2090 20 20 Years

Figure 42. Annual rainfall and linear trend in the Chalandri station for the scenario A2

38 Annual rainfall-Senario A2 Linear trend

800

700

600

500

400

300

200 Annual Annual rainfall (mm) Piraeus

100

10 14 18 26 30 34 38 42 46 54 58 62 66 70 74 78 86 90 94 98 20 20 20 2022 20 20 20 20 20 20 2050 20 20 20 20 20 20 20 2082 20 20 20 20

Years

Figure 43. Annual rainfall and linear trend in the Piraeus station for the scenario A2

The mean annual rainfall in the stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Pyraeus is 288,0 mm, 290,4 mm, 280,9 mm, 350,6 mm, 334,3 mm, 277,7 mm, 352,5 mm, 331,8 mm, 368,2 mm, 250,3 mm, respectively, for the period January 2010 to December 2100 for the climate scenario A2.

Figures 44 to 53 show the mean monthly rainfall in the stations Vyronas, Elefsina, Elliniko, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Pyraeus for the scenarios Α1Β and Α2.

39 Vyronas Senario A1B (2010-2100) 60,0 Senario A2 (2010-2100)

50,0

40,0

30,0

Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 44. Average monthly rainfall in Vyronas station for the climate scenarios A1B & A2

Elefsina

60,0 Senario A1B (2010-2100) Senario A2 (2010-2100)

50,0

40,0

30,0

Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 45. Average monthly rainfall in Elefsina station for the climate scenarios A1B & A2

40 Elliniko

60,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 50,0

40,0

30,0

Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 46. Average monthly rainfall in Elliniko station for the climate scenarios A1B & A2

Marathonas

80,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 70,0

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall

20,0

10,0

0,0 JFMAMJJASOND Months

Figure 47. Average monthly rainfall in Marathonas station for the climate scenarios A1B & A2

41 Markopoulo

80,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 70,0

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall

20,0

10,0

0,0 JFMAMJJASOND Months

Figure 48. Average monthly rainfall in Markopoulo station for the climate scenarios A1B & A2

Peristeri

60,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 50,0

40,0

30,0

Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 49. Average monthly rainfall in Peristeri station for the climate scenarios A1B & A2

42 Tatoi

80,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 70,0

60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall

20,0

10,0

0,0 JFMAMJJASOND Months

Figure 50. Average monthly rainfall in Tatoi station for the climate scenarios A1B & A2

Nea Filadelfeia

70,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 51. Average monthly rainfall in Nea Filadelfeia station for the climate scenarios A1B & A2

43 Chalandri

70,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 60,0

50,0

40,0

30,0 Rainfall (mm) Rainfall 20,0

10,0

0,0 JFMAMJJASOND Months

Figure 52. Average monthly rainfall in Chalandri station for the climate scenarios A1B & A2

Piraeus

50,0 Senario A1B (2010-2100) Senario A2 (2010-2100) 45,0 40,0 35,0 30,0 25,0 20,0 Rainfall (mm) Rainfall 15,0 10,0 5,0 0,0 JFMAMJJASOND Months

Figure 53. Average monthly rainfall in Piraeus station for the climate scenarios A1B & A2

44 Based on Figures 44-53, the maximum rainfall appears in December in all meteorological stations for both climate scenarios. The minimum rainfall appears for the scenario A1B in August in stations Elefsina, Marathonas, Markopoulo, Peristeri, Tatoi, Nea Filadelfeia, Chalandri, Piraeus and in September in stations Vyronas, Elliniko, and for the scenario A2 in August in stations Marathonas, Tatoi, in September in stations Vyronas, Elliniko, Nea Filadelfeia and in July in stations Elefsina, Markopoulo, Peristeri, Chalandri, Piraeus.

Figure 54 shows the average annual rainfall in all stations during the historical period 1970-2010 and the period 2010-2100 (based on climate scenarios A1B and A2); in all stations, the annual rainfall is decreased in scenarios A1B and A2 compared to the historical period.

470 1970-2010 450 Senario Α1Β (2010-2100) 430 Senario Α2 (2010-2100) 410 390 370 350 330 310

Annual rainfall Annual (mm) 290 270 250 230

i a ri i d onas iko n r in poulo ister Tatoi elfe la o er d a Vy Elefsina Ell P a Piraeus ark il Ch Marathonas M N.F Stations

Figure 54. Annual rainfall for the historical period (1970-2010) and for the climate scenarios Α1Β & Α2 (2010-2100)

45 5. CALCULATION OF SPI INDEX The SPI-12 index in the period 2010-2100 was calculated based on the methodology presented in Chapter 2. In Annex I is shown graphically the SPI-12 index in the period 2010-2100 for the climate scenarios A1B and A2. In Annex II is given in tabular form the duration and the intensity of drought events in each station in the period 2010-2100 for the climate scenarios A1B and A2. Figures 55 and 56 show the drought events, their intensity and duration in the period 2010-2100 for the climate scenarios A1B and A2, respectively. Based on Figure 55, in which the drought events for the climate scenario A1B are shown, it can be concluded: • Up to 2031 are expected three to four drought events of low intensity and short duration; the last event would be more intensive in the stations of Nea Filadelfeia, Elefsina and Elliniko. • From 2033 to 2064, the frequency of drought events is expected to increase. However, the intensity of drought events would not be high, except for two events in 2040 and 2054. The regions in the vicinity of stations Elefsina, Peristeri, Tatoi, Chalandri is expected to appear the higher intensity. • From 2064 to 2092, the frequency and the intensity of drought events are considerably increased in all stations. Based on Figure 56, in which the drought events for the climate scenario A2 are shown, it can be concluded: • Up to 2031 are expected a few drought events of low intensity and short duration; the last event would be more intensive in the stations of Nea Filadelfeia, Elefsina and Elliniko. • From 2041 to 2060, the frequency of drought events is expected to increase; their intensity, however, would be low. • From 2073 to 2099, the frequency and the intensity of drought events are considerably increased in all stations.

Vyronas Tatoi Piraeus Peristeri Markopoulo Marathonas Nea Filadelfeia Elliniko Elefsina Chalandri 60

DroughtMagnitude 20

0 Jul-21 Jul-44 Jul-67 Jul-90 Jan-10 Sep-11 Jun-26 Oct-29 Jun-31 Jan-33 Sep-34 Jun-49 Oct-52 Jan-56 Sep-57 Jun-72 Oct-75 Jan-79 Sep-80 Jun-95 Oct-98 Apr-18 Apr-41 Apr-64 Apr-87 Feb-28 Dec-14 Feb-51 Feb-74 Dec-37 Feb-97 Dec-19 Dec-60 Dec-83 Aug-16 Nov-24 Nov-47 Aug-39 Nov-42 Nov-70 Aug-62 Nov-65 Nov-93 Aug-85 Nov-88 Mar-23 Mar-46 Mar-69 Mar-92 May-13 May-36 May-54 May-59 May-77 May-82 May-00

Duration (months)

Figure 55. Drought events, intensity and duration in the meteorological stations of Attica for the climate scenario A1B (2010 – 2100)

Vyronas Tatoi Piraeus Peristeri Markopoulo Marathonas Nea Filadelfeia Elliniko Elefsina Chalandri

15 DroughtMagnitude

0 Jul-21 Jul-44 Jul-67 Jul-90 Jan-10 Sep-11 Jun-26 Oct-29 Jun-31 Jan-33 Sep-34 Jun-49 Oct-52 Jan-56 Sep-57 Jun-72 Oct-75 Jan-79 Sep-80 Jun-95 Oct-98 Apr-18 Apr-41 Apr-64 Apr-87 Feb-28 Feb-51 Dec-14 Feb-74 Feb-97 Dec-37 Dec-19 Dec-60 Dec-83 Aug-16 Nov-24 Nov-47 Aug-39 Nov-42 Nov-70 Aug-62 Nov-65 Nov-93 Aug-85 Nov-88 Mar-23 Mar-46 Mar-69 Mar-92 May-13 May-36 May-54 May-59 May-77 May-82 May-00

Duration (months)

Figure 56. Drought events, intensity and duration in the meteorological stations of Attica for the climate scenario A2 (2010 – 2100)

48 6. CALCULATION OF DROUGHT VULNERABILITY INDEX (DVI)

The DVI index is calculated in each meteorological station for the two climate change scenarios A1B and A2 (Figure 57). The DVI value is greater in scenario A1B than A2 in all stations except for Marathonas and Chalandri. In Maps 2 and 3 is shown the spatial display of DVI index for the scenarios A1B and A2, respectively.

3 Senario A1B Senario A2 2,5

1,5 DVI

0,5

0 Vyronas Elefsina Elliniko Marathonas Markopoulo Peristeri Tatoi Ν.Filadelfeia Chalandri Piraeus Stations

Figure 57. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenarios A1B & A2 (2010-2100)

Map 2. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenario A1B (2010-2100)

Map 3. Drought Vulnerability Index (DVI) in the meteorological stations of Attica for the climate scenario A2 (2010-2100)

51 REFERENCES Agnew, C. T. 2000. Using the SPI to Identify Drought. Drought Network News, Vol. 12, No. 1: 6-12, Winter 1999–Spring 2000. Apostolou, Κ. 2010. Drought assessment and forecasting using indices in Thessaly river basin district. Thesis. School of Civil Engineering, National Technical University of Athens. Hayes, M. J. 1999. Drought Indices. NDMC – Drought Happens, Drought Indices. http://enso.unl.edu/ndmc/enigma/indices.htm/, updated 12 October 1999. Ji, L. and A. Peters. 2003. Assessing vegetation response to drought in the northern Great Plains using vegetation and drought indices. Remote Sensing of Environment, 87: pp. 85–89. Kržič, A. I. Tošić, V. Djurdjević, K. Veljović, B. Rajković. 2011. Changes in climate indices for Serbia according to the SRES-A1B and SRES-A2 scenarios. Clim. Res. 49:73-86. Loukas, A. and L. Vasiliades. 2004. Probabilistic analysis of drought spatiotemporal characteristics in Thessaly region, Greece. Natural Hazards and Earth System Sciences, 4: 719–731. McKee, T. B., N. J. Doesken, and J. Kleist. 1993. The relationship of drought frequency and duration to time scales. In Proceedings of the 8th Conference on Applied Climatology,Anaheim, California, 17-23 January. Ministry of Environment, Energy and Climate Change. 2013. River basin management plan of Attica water district (GR06). Partnership NAMA Consulting Engineers– Gamma4 EPE–Ν. Sideris, SPEED Development consulting SA, F. Pergantis, Ath. Νtaskas, G. Gianelis, N. Christou, Anna Mpitsakaki–Tsoukia, Εf. Chatzikostas. Sims, A. P., D. S. Niyogi, and S. Raman. 2002. Adopting drought indices for estimating soil moisture: A North Carolina case study. Geophys. Res. Lett., 29, 8, 1183. Szalai, S. and C. Szinell. 2000. Comparison of two drought indices for drought monitoring in Hungary — a case study. In Drought and Drought Mitigation in Europe, Vogt JV, Somma F (eds). Kluwer: Dordrecht: 161–166.