Assesment of the Drought Pattern Change in Çamldere Basin Using SPI Index

Prévia do material em texto

BALWOIS 2010 - Ohrid, Republic of Macedonia - 25, 29 May 2010 
 
1
Assesment of the Drought Pattern Change in Çamlıdere Basin Using 
SPI Index 
 
Fatih KESKIN1, A.Ünal ŞORMAN2 
1State Hydraulic Works, Ankara, Turkey, e-mail : fatihk@dsi.gov.tr 
2Middle East Technical University, Ankara, Turkey, e-mail : sorman@metu.edu.tr 
 
Abstract 
Drought and floods which are the major issues of today’s world have several inverse impacts on economy, 
society and ecology. The severity of the drought is related with main usage of water and can be 
expressed in different indexes. These indexes could give different information for the drought analysis and 
one of them is the standardized precipitation index (SPI). 
 
In this study, the analysis of severe and longer duration of drought events has been carried out by 
applying the SPI. The severity and change pattern of drought has been analyzed by SPI for the Çamlıdere 
dam basin where the basin supplies most of the domestic water to the capital city of Turkey where the 
total supplied water is about 800.000 m3/day. The meteorological data for the period of 1926-2008 is used 
and the period is divided into two periods as 1926-1966 and 1967-2008. The analysis of these periods 
have been carried out by the investigation of short term and long term periods with different cumulative 
months as 1,3,6,9,12,24. Additionally, the change in drought pattern is analyzed by examining the change 
pattern for different return period of drought events. 
 
The results showed that there is %20 increase in the second period of 24 month SPI and % 35 increase 
for 12 month SPI, where as 10% increase in short term SPI is examined. Also there is no difference for 
the 6 and 9 months SPI. The difference between two periods showed that there is a change in the pattern 
of droughts in Çamlıdere dam basin. The results also prove the change of drought pattern in the basin 
with also supporting that mostly the change is on river flow and groundwater with in the basin. 
 
Keywords: Turkey, Drought, Çamlıdere, SPI, pattern change 
 
Introduction 
Droughts are one of the world’s most severe and collectively affective natural disasters that cause an 
average $6-8 billion in global damages yearly (Wilhite 2000). Agricultural crop damage is expected to be 
same order of magnitude in Turkey because of the 2007 severe drought events took place in the country. 
The usable water sources consist of soil moisture, groundwater, snow pack, runoff and reservoir storage. 
Any drought is directly related with the one or more of these five sources of supply. The time difference 
from the precipitation occurs to water becomes available in each useable form is extremely high. Water 
uses also have characteristic time scales. Consequently, the impacts of a water deficit are a complex 
function of water source and water use. The time scale over which precipitation deficits accumulate 
becomes extremely important and functionally separates different types of drought. Agricultural (soil 
moisture) droughts, for example, typically have a much shorter time scale than hydrologic (groundwater, 
runoff and reservoir storage) droughts (Mckee et al, 1993). 
 
There are several studies not directly related to drought but focused on analyzing the conditions and 
trends of hydrological variables (Van Belle & Hughes, 1984; Zhang et al., 2000, 2001). There are also 
studies (Kadıoğlu, 1997; Turkeş et al., 2002; Karabörk, 2007) examining the temperature trends of 
Turkey, in which warming and cooling trends in some parts of Turkey were obtained by the precipitation 
station records. Some other studies (Turkeş, 1996; Kalayci & Kahya, 2006; Karabörk, 2007) focused on 
the precipitation and runoff records in Turkey and detected significant trends in nearly one-third and half of 
the stations. Ünal & Karaca (2003) made a study for clustering the climate zones in Turkey and found six 
different zones named from A to F (Figure 1). Among these, Zone D represens the Central Anatolia of 
Turkey, within which our study area is located. Drought can be examined in three different types which 
are referring to a water deficit in a hydrological cycle with a connection that a drought in one stage of the 
cycle can also lead to a drought in other stages. It starts with a less than normal amount of precipitation 
BALWOIS 2010 - Ohrid, Republic of Macedonia - 25, 29 May 2010 
 
2
which is called a meteorological drought. After that soil moisture drought and hydrological drought might 
develop. An agricultural drought is characterized by low soil moisture content, leading to insufficient water 
supply to cultivated plants. The term hydrological drought is applied to less than normal amounts of water 
in the different types of water bodies, and represented by low water levels in streams, reservoirs, and 
lakes, as well as groundwater aquifers. 
 
 
Fig. 1 Climate zones of Turkey, Unal & Karaca (2003) 
 
Several different indices have been developed to define types of droughts. These indices have been 
presented in different studies (Heim, 2002; Keyantash & Dracup, 2002) where most of them (Standardized 
Precipitation Index (SPI), Standardized Runoff Index (SRI), Surface Humidity Index (SHI)) are separate 
indices and define one of the previously defined drought types. Among the indices, SPI is the most used 
one for defining the meteorological drought by using the precipitation data. In this study, SPI is used for 
the analysis of drought pattern change in the study area. 
 
Data and Methodology 
The study area is the Çamlidere Dam basin that supplies most of the domestic water to Ankara city in 
Turkey, where the total water supply is about 800 000 m3/day. The location of the basin and the Ankara 
city is shown in Figure 2. The study area lies in the zone “D” of zones that are defined by Unal & Karaca 
(2003). 
 
Legend
Rivers
City Center
Camlıdere Basin
Dams
#* Meteorological Stations
& Snow Locations
 
BALWOIS 2010 - Ohrid, Republic of Macedonia - 25, 29 May 2010 
 
3
Fig. 2 The basin and the stations 
 
Monthly precipitation data recorded by the State Meteorological Service (DMI) were used in the study... 
Locations of the stations in Ankara, Esenboğa and Kizilcahamam are shown in Figure 2. These three 
stations are either inside or close to the settlement areas so that they can not represent the study area 
alone. Herewith, precipitation data from Guvenc basin that was recorded by Ministry od Agriculture and 
Rural Affairs (MARA) is also included in the study. Since high correlation coefficient (0.80) between 
Esenboga and Ankara stations were obtained, data of the Ankara station was omitted to prevent 
redundancy. The temperature data was obtained from DMI for the study period (1988-2008). 
 
SPI index 
SPI is a common index used to detect the change in drought. Firstly, a time series of the precipitation 
value of interest is generated. Then, a frequency distribution is selected and a statistical fit to the data is 
determined. The cumulative distribution is formed from the fitted frequency distribution. The percentile for 
the particular time series element of interest, usually the latest one, is selected from the cumulative 
distribution. For "ties" (multiple instances of the same value), the upper value is used (probability of non-
exceedance). For any other theoretical probability distribution, the analogous point on its associated 
cumulative frequency distribution can be determined. Here, the normal distribution is used, with mean 
zero and one standard deviation, and value in standardized units of a given percentile can readily be 
determined. For the normal distribution, these are exactly the same units of standard deviations. The 
Standardized Precipitation Index can be considered as the number of standard deviations that the 
precipitation value of interest would be away from the mean for an equivalent normal distribution, and 
adequate choice of fitted theoretical distribution forthe actual data. In effect, the method consists of a 
transformation of one frequency distribution to another frequency distribution, in this case the widely used 
normal (Gaussian) distribution. The SPI is calculated by deducting the precipitation difference from the 
mean for a time period and dividing this value by the standard deviation of the whole recorded 
precipitation. The precipitation data is assumed as normally distributed (Mckee et all., 1993). The 
precipitation data can also be expressed with Gamma or Log Normal 3 distributions (Shukla & Wood, 
2008). Mckee et al. (1993) suggested a methodology to calculate the SPI with selecting a j months set of 
aggregated precipitation data to determine a set of time scales, where j is 3, 6, 12, 24, 48 months. The 
dataset reorganized in a moving sense in that a new series is formed from the previous j months. These 
datasets are fitted to gamma function to get the probability of precipitation for the recorded data. The 
probability of a precipitation value is determined with an estimate of the inverse normal for calculating the 
precipitation for a normally distributed probability density with a mean of zero and standard deviation of 
unity (Mckee et al. 1993). This value is called as SPI for the particular precipitation data value. The 
advantage of the SPI is that after normalization, both wet and dry periods can be expressed with SPI. 
 
Mckee et al. (1993) defined drought event for time period j where SPI is continuously negative and SPI 
reaches -1.0 or less. The defined categories of SPI for the drought severity definition are given in Table 1. 
 
 
Table 1 : Drought categorization values (Mckee et al. 1993) 
SPI Values Drought Category 
0 to -0.99 mild drought 
-1.00 to -1.49 moderate drought 
-1.50 to -1.99 severe drought 
Figure 7. The 
resulting graph shows that there is no difference for the 6 and 9 month SPI whereas about %20 
differences are observed in 24 month SPI. An increase of % 35 in 12 month SPI also represents that the 
drought pattern is changing within the basin. The same results can also been seen in Figure 8 for 100 
year period return period. But additionally there is an increasing in drought pattern for the 6 and 9 month 
SPI values. 
BALWOIS 2010 - Ohrid, Republic of Macedonia - 25, 29 May 2010 
 
7
500 Year Return Period SPI
0
0.5
1
1.5
2
2.5
3
3.5
1 month 3 month 6 month 9 month 12 month 24 month
Month
SP
I
500 (1926-1966) 500 (1967-2008)
 
Figure 7 : Different SPI values for 500 year return period 
100 Year Return Period SPI
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
1 month 3 month 6 month 9 month 12 month 24 month
Month
SP
I
100 (1926-1966) 100 (1967-2008)
 
Figure 8 : Different SPI values for 100 year return period 
 
 
Conclusion 
In this study, meteorological drought analysis has been carried out with monthly precipitation data for the 
period of 1926-2008. The period was divided into two separate periods as 1926-1966 and 1967-2008. 
After the calculation of the meteorological drought index, SPI, the change analysis was done by 
comparing the SPI values for different month(s) period SPI. The drought pattern change for the periods 
was examined and the difference between two periods was identified. Normally 1-3 month(s) SPI values 
can show the effect on agricultural applications. 6-12 months SPI can show the drought on reservoirs’ 
levels and especially on river discharges if there is snow depletion and melting at the upstream of the 
river. 12-24 months SPI can generally show the effect on groundwater table and groundwater recharge. 
 
The results showed that there is %20 increase in the second period of 24 month SPI and % 35 increase 
for 12 month SPI, whereas 10% increase in short term SPI is examined. Moreover, there is no difference 
BALWOIS 2010 - Ohrid, Republic of Macedonia - 25, 29 May 2010 
 
8
for the 6 and 9 months SPI. The difference between two periods showed that there is a change in the 
pattern of droughts in Çamlıdere dam basin. The results also proves the change of drought pattern in the 
basin with also supporting that mostly the change is on river flow and groundwater with in the basin. 
 
References 
Heim, R. R. Jr. (2002). A review of the twentieth-century drought indices used in United States. Bull. Am. 
Met. Soc. 83, 1149-1165. 
 
Kadıoğlu, M. (1997). Trends in Surface air Temperature data over Turkey. Int. J. Climatol. 17, 511-520. 
 
Kalayci, S. & Kahya, E. (2006). Assessment of Runoff variability modes in Turkey:1964-1994., J. Hydrol. 
324, 163-177 
 
Karabörk, M. C. (2007). Trends in drought Patterns of Turkey. J. Environ. Sci. 6, 42-52. 
 
Keyantash, J. A. & Dracup, J. A. (2002). The Quantification of Drought:An Evaluation of Drought İndexs. 
Bull. Am. Met. Soc. 83, 1167-1180. 
 
Mckee, T. B., Doesken, N.J., & Kleist, J. (1993). The Relationship of Drought Frequency & Duration to 
Time Scales. Eighth Conference on Applied Climatology, 17-22 January, Anaheim: California. 
 
Shukla, S. & Wood, A. W. (2008). Use of Standardized runoff index for characterizing hydrologic drought. 
Geophys. Res. Lett. 35, article in proof. 
 
Türkeş, M. (1996). Spatial and Temporal Analysis of Abbual Rainfall Variations in Turkey. Int. J. Climatol. 
23, 1045-1055. 
 
Türkeş, M., Sümer, U. M. & Demir, İ. (2002). Re-evaluation of trends & changes in mean, maximum & 
minimum temperatures of Turkey for the period 1929-1999. Int. J. Climatol. 22, 947-977. 
 
Unal, T. K. & Karaca, M. (2003). Redifining climate zones for Turkey using cluster analysis. Int. J. 
Climatol. 23, 1045-1055. 
 
Van Belle, G. & Hughes, J. P. (1984). Nonparametric tests for trend in water quality. Water Resour. Res. 
20, 127-136 
 
Zhang, X., Vincent, L. A., Hogg, W. D., Niitsoo, A. (2000). Temperature and precipitation trends in Canada 
during 20th century. Atmosphere-Ocean 38, 395-429. 
 
Zhang, X., Harvey, K. D., Hogg, W. D. & Yuzyk, T. R. (2001). Trends in Canadian Runoff. Water Resour. 
Res.37, 987-998.