Abstract

Iran is not a rich freshwater country. Many dams have been constructed in Iran during the last decades. As better reservoir operation is always important, discovering different properties of water in the reservoirs can be used for developing operating policies. Shahid Rajaee dam with a large reservoir is located in Mazandaran Province, Iran. Shahid Rajaee reservoir falls in the warm monomictic stratified category. Available data from the period 2012 to 2015 have been used for modeling temperature, Dissolved Oxygen (DO) and Total Suspended Solids (TDS) in this reservoir using CE-QUAL-W2. Temperature–depth diagrams have been drawn and thermocline characteristics have been determined from the first-year data. Moreover, two operation scenarios have been defined. The effect of 20% reduction of the TDS loads in both upstream rivers or the major river on the TDS concentration of the reservoir at a specific point has been evaluated. The model's results show 20% and 14.2% reduction in the maximum values of the TDS time series, respectively and 19.4% and 13.4% reduction in the minimum values of the TDS time series, respectively. Moreover, the Iran Water Quality Index for Surface Water Resources – Conventional Parameters (IRWQISC Index) Index has been used for determining the reservoir's water quality. The index showed a ‘good’ water quality for the reservoir before and after the salinity reduction scenario.

INTRODUCTION

Human population growth is much more than the available water resources in the world (Kummu et al. 2010). Iran is a country with 1% of the world's population. However, only 0.37% of freshwater resources belongs to Iran (Zolpirani et al. 2015). Therefore, in this situation, reservoir operation is of complete importance. Additionally, water quantity and quality are among the most important parameters of a reservoir's operation (Chaves et al. 2004). For instance, the operation of upstream rivers or their basins seems to be important.

Before 1994, environmental impact assessment (EIA) was not supported by the law in Iran. Shahid Rajaee dam is among the projects for which preconstruction studies were finished before that year (Nazariha & Alinejad 2003). However, according to our previous unpublished study, Shahid Rajaee is a stratified reservoir, and it is classified as a warm monomictic reservoir according to its stratification (Rahimi-Movaghar 2017). One of the physical characteristics of stratified reservoirs is having a thermocline. The thermocline divides the water column into three distinct layers which have different water qualities. Therefore, the determination of each layer's depth is an essential parameter for better reservoir operation.

Another water quality variable that provides more physical features of the reservoir is total suspended solids (TDS). For different human usages, different amount of TDS is required. Management of TDS in the upstream river may have a significant effect on the reservoir TDS.

In this study, researchers aimed to determine the quality of Shahid Rajaee reservoir's water with a single index and estimate the effect of TDS reduction on the water quality.

METHODS

The geographical coordinates of the Shahid Rajaee dam are 53°13′43″ longitude and 36°14′57″ latitude. The dam is located 40 kilometres southeast of Sari city in Mazandaran Province, Iran, and it has been constructed on Sefid Rud River and Shirin Rud River, which are two branches of Tajan River (Figure 1).

Figure 1

The location of Shahid Rajaee dam in Mazandaran Province and the map of dam and reservoir and the data sampling stations.

Figure 1

The location of Shahid Rajaee dam in Mazandaran Province and the map of dam and reservoir and the data sampling stations.

The software used for modeling the reservoir is CE-QUAL-W2 (Deus et al. 2013; Torres et al. 2016). The data required for modeling with this specific software are divided into two groups, input–output and calibration data and they usually include topography, quality, hydrological, hydrodynamic and meteorological data categories. For collecting the calibration data, there were two groups of stations. The first group are stations 1 to 6 (Rahimi-Movaghar 2017) in Figure 1 and the second group are stations 7 to 9 (Figure 1). The calibration parameters were temperature, DO and TDS. The first six stations were recorded with a portable device named Hach HQ 40d (Shoaie 2016; Rahimi-Movaghar 2017) and the other three stations are data from the thesis written by Reza Mousavi (Mousavi 2017). The resolution of temperature and DO for the first six stations were 0.01 °C and 0.05 mg·L−1, respectively. The seasonal sampling was done for the first six stations during autumn 2012 and summer 2013 (Shoaie 2016). For summer 2015 the other calibration data (the second group of stations) were used to recalculate (somehow recheck) the previous calibration.

The meteorological data were obtained from three meteorological stations for the duration of the modeling. Gharakhil, Kiasar and Pol Sefid stations are the nearest stations around the reservoir. Gharakhil is located 22 km northwest of the center of the Shahid Rajaee dam (CSRD). Kiasar is located at 26.5 km east of CSRD. Moreover, Pol Sefid is located 48 km southwest of CSRD. The meteorological data were created from five different parameters, which were the velocity and direction of the wind, air and dew-point temperature and cloud cover. These data were taken from Iran Meteorological Organization. Moreover, the monthly data of inflows, nitrate, phosphate and TDS from October 2012 to October 2015 were collected from Iran Water Resource Management Company. But, we did not have the complete data of the two input streams' temperature and DO parameters.

Air and river temperature have strong correlation together. Therefore, a linear regression model is usually used to define river temperature. But, in a study of 584 gauging stations of rivers (Mohseni et al. 1998), there was some non-linearity and the stream temperature/air temperature diagram was almost in the shape of a ‘S’. This means that non-linearity was more obvious at both ends of the diagram in warm climatic zones. Therefore, in this research, based on the warm climatic zone and data published by Mohseni Mohseni & Stefan (1999), the average data of these streams were used for this case study. The air temperature and the relationship between the air and stream temperature were provided from the Mohseni data (Mohseni & Stefan 1999). In this way, the river water temperature was estimated.

To find the DO parameter, it was considered that there is a linear regression between water temperature and DO. First, the last linear regression equation was used to discover the missing data. Second, the calibration data just after the unknown DO duration was available. Considering that if a positive amount is added to the initial monthly DO value of the upstream rivers, the modeled DO will shift right and if a negative amount is added to the initial monthly DO value, the modeled DO will shift left. For this special case, after running the model, the first (initial) absolute error of calibrated DO had a small value, so by four steps running and adding and decreasing to the amount of the initial monthly DO (of the upstream rivers), the illustrated figure was found out (Figure 2). Furthermore, the daily volume of inflow and outflow of the dam were collected from Mazandaran Regional Water Co.

Figure 2

Comparison of the modeled and observed dissolved oxygen in the added data. The rows from top to bottom represent stations seven, eight and nine, respectively; the columns from left to right represent Julian days 1259, 1281 and 1350, respectively. HASL height above sea level.

Figure 2

Comparison of the modeled and observed dissolved oxygen in the added data. The rows from top to bottom represent stations seven, eight and nine, respectively; the columns from left to right represent Julian days 1259, 1281 and 1350, respectively. HASL height above sea level.

The depth sampling of the reservoir for the first six stations was as follows: for the first 30 metres, each 5 metres and after that from 30 to 50 metres each 10 metres (Rahimi-Movaghar 2017). Moreover, the depth sampling criteria for the three other stations were at the surface layer, 5-metre depth and after that each 10 metres' depth (Mousavi 2017).

For preparing the input data of CE-QUAL-W2 in bathymetry section, a MATLAB code was written that creates the bathymetry file from the surveyed topography data of the reservoir (Rahimi-Movaghar 2017).

The bathymetry of the reservoir has two branches (based on two upstream rivers). The first branch has 20 segments (1–20) and the second one has ten segments (21–30). Also, this bathymetry has 97 layers (Rahimi-Movaghar 2017) (Figure 3).

Figure 3

The orientation of the segments of the Shahid Rajaee reservoir (Rahimi-Movaghar 2017).

Figure 3

The orientation of the segments of the Shahid Rajaee reservoir (Rahimi-Movaghar 2017).

After the data input to the model, the calibration started. In the process, first of all the geometry calibration was done and after that sequentially, water surface level, temperature and DO or TDS calibrations were done.

In geometry calibration the mean absolute percentage error (MAME) equation was used based on Equation (1):  
formula
(1)
where n is the number of fitted points. is the actual value and is the forecast value.
The absolute mean error (AME) based on Equation (2) is another formula which is widely used in this article:  
formula
(2)
where n, and are as above.
The root mean square error (RMSE) based on Equation (3) is used in the text:  
formula
(3)
where n, and are as above.
Iran Water Quality Index for Surface Water Resources – Conventional Parameters has been used for determining the water quality of Shahid Rajaee reservoir. The index is calculated based on 11 parameters including fecal coliform, 5-d biochemical oxygen demand, nitrate, dissolved oxygen (saturation, %)(DOSP), electrical conductivity (EC), chemical oxygen demand, ammonium, phosphate, turbidity, total hardness and pH and each of them has its own weight. The equation used for calculating the index is written as Equation (4) (Sadeghi et al. 2015; Kabir et al. 2017):  
formula
(4)

In Equation (4), and n is the total number of parameters, is the normalized value assigned to parameters in the scale of 0–100, and Wi is the weight of each parameter. This index was developed by Iran Department of Environment in 2013 (Iran Department of Environment 2013). There is a footnote in the instruction of this index which states that if we do not have all of these parameters, we can still use the index with the same weights. In this research we had two of the above parameters including DOSP and EC.

In order to calculate DOSP, the maximum dissolved oxygen concentration saturation (MDOCS) was found in 4-day intervals for a cell in segment 19 and layer 58 from the top of the bathymetry (each layer depth is 1 metre). For the same days, DO was found from the model. Then DOSP was calculated using Equation (5):  
formula
(5)
MDOCS was used instead of the saturation amount, as we did not have the saturation value of dissolved oxygen for these days. EC was calculated using Equation (6):  
formula
(6)
In Equation (6), 1.56 was found from the average of 12 days of dividing EC by TDS from upstream river data.

The coefficients of EC and DO in Equation (4) are 0.096 and 0.097, respectively. In order to calculate the DO and EC indexes there are two figures in the instructions of the index which relate the values of DOSP and EC to their indexes, and these figures have been used for calculating the indexes. After using Equation (4) for finding the total index of the water body, we used Table 1 for determining the quality of the water body.

Table 1

Determining the water body quality from the index

Water body quality Water index 
Very good >85 
Good 70.1–85 
Relatively good 55.1–70 
Fair 45–55 
Relatively poor 30–44.9 
Poor 15–29.9 
Very poor <15 
Water body quality Water index 
Very good >85 
Good 70.1–85 
Relatively good 55.1–70 
Fair 45–55 
Relatively poor 30–44.9 
Poor 15–29.9 
Very poor <15 

The gradient criterion method is a widely used method for determining the thermocline (Fee et al. 1996; Zhang et al. 2015). This method requires that the vertical gradient of temperature in the water should be greater than a specific fixed value. However, there is no objective way to determine the appropriate value of the criterion, which ranges from 0.05 °C·m−1 to 2 °C·m−1 (Fee et al. 1996; Coloso et al. 2011; Hao et al. 2012; Zhang et al. 2015). For example, 0.2 °C·m−1 is the uniform criterion for the Chinese lake, Qiandaohu, water temperature (Zhang et al. 2015). In this study, we have used the uniform value of 0.2 °C·m−1 for determining the thermocline. The profiles used here are the resulting profiles of one complete year, from October 2012 to September 2013 (CE-QUAL-W2 software results). To do this, first, each day's profile was drawn in MATLAB software, then two tangent lines with the fixed slope of 0.2 °C·m−1 were moved throughout the profile to find the two primary and ending points in the profile with the exact slope of 0.2 °C·m−1 which represent the thermocline range. The midpoint of this range was chosen as the thermocline depth and the range as the thermocline thickness. After finding these two values, the monthly averages of these two values were calculated.

RESULTS AND DISCUSSION

The different calibration step results are as follow.

Geometry and water surface layer calibrations

The MAPE percentage of geometry calibration was 0.4% and the RSME of water surface layer calibration was 0.05 metres.

Temperature calibration

Based on the model's instructions (Cole & Wells 2015), we added recent data to the initial model data (Rahimi-Movaghar et al. 2017). Then we re-calibrated the temperature data (the data has not been demonstrated).

DO calibration

The DO data were also calibrated for the second time for the 2012–2015 duration (Cole & Wells 2015) (Figure 2 and Table 2).

Table 2

The absolute error for dissolved oxygen profiles (mg/L) (Figure 2)

Julian day S7(segment 15) S8(segment 13) S9(segment 10) 
1259 0.53 0.3 0.32 
1281 0.84 1.33 0.97 
1350 0.71 0.53 0.38 
Julian day S7(segment 15) S8(segment 13) S9(segment 10) 
1259 0.53 0.3 0.32 
1281 0.84 1.33 0.97 
1350 0.71 0.53 0.38 

The minimum AME for dissolved oxygen of the second series calibration data is 0.3 mg·L−1 and the maximum value is 1.33 mg·L−1. The calibration error range for dissolved oxygen data are completely in the acceptable range (compared with the dissolved oxygen models in the manual of the CE-QUAL-W2) (Cole & Wells 2015).

TDS calibration

The simulation model was also calibrated using TDS data (Figure 4 and Table 3).

Table 3

The absolute error for TDS profiles (mg/L) (Figure 4)

Julian Day S7(segment 15) S8(segment 13) S9(segment 10) 
1259 13.6 16.2 20.7 
1281 21.5 23 17.9 
1350 43 32.5 23.6 
Julian Day S7(segment 15) S8(segment 13) S9(segment 10) 
1259 13.6 16.2 20.7 
1281 21.5 23 17.9 
1350 43 32.5 23.6 
Figure 4

Comparison of the modeled and observed TDS: rows from top to bottom represent stations seven, eight and nine, respectively; columns from left to right represent Julian days 1259, 1281 and 1350, respectively. HASL is height above sea level.

Figure 4

Comparison of the modeled and observed TDS: rows from top to bottom represent stations seven, eight and nine, respectively; columns from left to right represent Julian days 1259, 1281 and 1350, respectively. HASL is height above sea level.

The minimum AME for TDS calibration data is 13.6 mg·L−1 and the maximum AME is 43 mg·L−1. The TDS calibration error range is in the range of Karkheh reservoir (Etemadi-Shahidi et al. 2009).

Monthly thermocline depth

During the second calibration of temperature, the first-year data of the temperature calibration error was less than for the other years, so this year was chosen for determining the thermocline depth. For this purpose, the monthly averaged thermocline (when it occurred) was determined (Figure 5).

Figure 5

Shahid Rajaee reservoir thermocline depth and thickness.

Figure 5

Shahid Rajaee reservoir thermocline depth and thickness.

Figure 5 shows that based on our definition, there is no thermocline in the reservoir from November to February. The maximum depth of thermocline occurs in October and the least occurs in March. The depth of thermocline increases from March to October. The same diagram with the same characteristics has illustrated Lake Qiandaohu in China (Zhang et al. 2015).

Thermocline thickness

Based on the first-year data of temperature, thermocline thickness was determined (Figure 5).

Based on Figure 5, the least thermocline thickness occurs in October and the maximum occurs in June. The thickness increases from March to June and decreases from June to October. The same thickness diagram has illustrated Lake Qiandaohu in China (Zhang et al. 2015).

Also, two scenarios were determined and modeled. The scenario was the effect of decreasing 20 percent of the inflow TDS load on the TDS concentration time series in the reservoir. In the second scenario, only the TDS load of the main upstream branch was reduced by 20 percent (Figure 6).

Figure 6

TDS (mg·L−1) time series before decreasing the monthly amount of TDS in upstream rivers (a) after decreasing 20% from two branches and (b) after decreasing 20% only from branch 1(c).

Figure 6

TDS (mg·L−1) time series before decreasing the monthly amount of TDS in upstream rivers (a) after decreasing 20% from two branches and (b) after decreasing 20% only from branch 1(c).

In decreasing the amount of TDS in two branches, one of the branches reached the maximum value of TDS. According to these scenarios, the maximum value of TDS concentration in the reservoir changed from 437 mg·L−1 (Figure 6(a)) to 350 mg·L−1 (Figure 6(b)) and 375 mg·L−1 (Figure 6(c)) (reduced by 20% and 14.2%), respectively. So, changing the decrease from one branch (Figure 6(c)) to two branches (Figure 6(b)) had different effects on the maximum value of TDS. Moreover, the minimum values are different with two branches decreasing (Figure 6(b)) and one branch decreasing (Figure 6(c)). The minimum value of the time series, as shown in Figure 6(a) is 310 mg·l−1, but the minimum values with two branches decreasing (Figure 6(b)) and one branch decreasing (Figure 6(c)) are 250 mg·L−1 and 267 mg·L−1 (19.4% and 13.4% reduction), respectively.

The IRWQISC index

The index value for the research was 58.8, which showed the relatively good water quality of the Shahid Rajaee reservoir. After decreasing 20% of TDS load in branch 1, the resulting value changed to 60.5 and after decreasing 20% of TDS in both branches it changed to 62.5, both of which are in the relatively good value group.

When the process was done in reverse and the EC value needed for changing the group of from relatively good to good was calculated, the amount of reservoir EC and TDS was calculated as 100 μSI·cm−1 and 64 mg·L−1, respectively, and this is not a reachable value.

The same index has been used with the 11 parameters defined in IRWQISC for Ghareh-chai River. The best quality was observed in autumn (‘Relatively good’ condition) and the worst was in summer (‘Fair’ condition) (Kabir et al. 2017). Moreover, the index has been used for determining the water quality of Karkheh reservoir to design the optimal water quality monitoring networks for reservoirs (Maymandi et al. 2018).

Our future study might be on monitoring, controlling and maybe, decreasing the TDS of upstream rivers and their basins to achieve more appropriate water of different usages including agriculture, drinking and pisciculture.

Limitations of the study are: first, lack of complete data for the modeling period; second, using two different data sources for calibration; third, not having calibration data in regular intervals; fourth, not having an exact saturated DO parameter.

CONCLUSION

Our temperature, DO and TDS models had acceptable ranges of error. We determined the monthly thermocline thickness and depth. Decreasing 20% of the TDS load in both upstream rivers or only the major river had different effects on both the maximum and minimum values of TDS in the time series of a cell near the dam. It changed the maximum value more than the minimum value in both situations. Moreover, the reduction of TDS in both branches had more effect on the reduction of TDS in a cell near the dam than the reduction of TDS load in one branch. Therefore, by controlling and reducing the TDS loads in the upstream rivers, we can reduce the reservoir’s TDS concentration. The IRWQISC Index showed that by reducing the amount of TDS, water quality will improve but the quality group of the water will not change (from relatively good to good).

ACKNOWLEDGEMENTS

We would appreciate the Iran Meteorological Organization, Mazandaran Regional Water Co. and Iran Water Resource Management Company for their valuable free data. In addition, we would appreciate the great help of Ms. Ameneh Hashemi and Mr Kazem Abbaszadeh.

REFERENCES

REFERENCES
Chaves
,
P.
,
Tsukatani
,
T.
&
Kojiri
,
T.
2004
Operation of storage reservoir for water quality by using optimization and artificial intelligence techniques
.
Mathematics and Computers in Simulations
67
(
4–5
),
419
432
.
doi: 10/1016/j.matcom.2004.06.005
.
Cole
,
T. M.
&
Wells
,
S. A.
2015
CE-QUAL-W2: A Two-Dimensional, Laterally Averaged, Hydrodynamic and Water Quality Model, Version 4
.
Department of Civil and Environmental Engineering, Portland State University
,
Portland, OR
,
USA
.
Coloso
,
J. J.
,
Cole
,
J. J.
&
Pace M
,
L.
2011
Short-term variation in thermal stratification complicates estimation of lake metabolism
.
Aquatic Sciences
73
(
2
),
305
315
.
doi:10.1007/s00027-010-0177-0
.
Deus
,
R.
,
Brito
,
D.
,
Mateus
,
M.
,
Kenov
,
I.
,
Fornaro
,
A.
,
Neves
,
R.
&
Alves
,
C. N.
2013
Impact evaluation of a pisciculture in the Tucuruí reservoir (Pará, Brazil) using a two-dimensional water quality model
.
Journal of Hydrology
487
,
1
12
.
doi:10.1016/j.jhydrol.2013.01.022
.
Etemadi-Shahidi
,
A.
,
Afshar
,
A.
,
Alikia
,
H.
&
Moshfeghi
,
H.
2009
Total dissolved solid modeling; Karkheh reservoir case example
.
International Journal of Environmental Research
3
(
4
),
671
680
.
doi:10.22059/IJER.2010.83
.
Fee
,
E. J.
,
Hecky
,
R. E.
,
Kasian
,
S. E. M.
&
Cruikshank
,
D. R.
1996
Effects of lake size, water clarity, and climatic variability on mixing depths in Canadian Shield lakes
.
Limnology and Oceanography
41
(
5
),
912
920
.
doi:10.4319/lo.1996.41.5.0912
.
Hao
,
J.
,
Chen
,
Y.
,
Wang
,
F.
&
Lin
,
P.
2012
Seasonal thermocline in the China Seas and northwestern Pacific Ocean
.
Journal of Geophysical Research Oceans (1978–2012)
117
(
C2
),
C02022
.
doi:10.1029/2011JC007246
.
Iran Department of Environment
.
A guide for estimation of Iran water quality index for surface water resources. 2013 https://www.doe.ir/Portal/file/?696074/
(in Persian).
Kabir
,
M.
,
Shahbazi
,
A.
,
Kouhanestani
,
Z. M.
&
Shahraki
,
F. R.
2017
Spatial–temporal variability in water quality in Ghareh-chai River, Golestan Province, Iran
.
AACL Bioflux
10
(
2
),
353
364
.
Kummu
,
M.
,
Ward
,
P. J.
,
de Moel
,
H.
&
Varis
,
O.
2010
Is physical water scarcity a new phenomenon? Global assessment of water shortage over the last two millennia
.
Environmental Research Letters
5
,
034006 doi: 10.1088/1748-9326/5/3/034006
.
Maymandi
,
N.
,
Kerachian
,
R.
&
Nikoo
,
M. R.
2018
Optimal spatio-temporal design of water quality monitoring networks for reservoirs: application of the concept of value of information
.
Journal of Hydrology
558
,
328
340
.
doi: 10.1016/j.jhydrol.2018.01.011
.
Mohseni
,
O.
&
Stefan
,
H. G.
1999
Stream temperature/air temperature relationship: a physical interpretation
.
Journal of Hydrology
218
,
128
141
. doi:10.1016/S0022-1694(99)00034-7.
Mohseni
,
O.
,
Stefan
,
H. G.
&
Ericson
,
T. R.
1998
A nonlinear regression model for weekly stream temperatures
.
Water Resources Research
34
(
10
),
2685
2692
.
doi:10.1029/98WR01877
.
Mousavi
,
R.
2017
Investigating and Comparing 2D and 1D Modeling Physical and Chemical Parameters of Reservoir Water
.
Master of Science thesis
,
Water Engineering Group, University of Tehran
, Tehran, Iran
(in Persian)
.
Nazariha
,
M.
&
Alinejad
,
S.
2003
Planning for retrofitting and reducing the negative environmental features of Shahid Rajaee dam
.
Mohitshenasi
30
,
9
18
(in Persian)
.
Rahimi-Movaghar
,
M.
2017
Modeling Dissolved Oxygen Stratification in Reservoirs (Shahid Rajaee Reservoir)
. Master of Science thesis, Environmental Engineering Group, K.N. Toosi University of Technology, Tehran, Iran.
Sadeghi
,
M.
,
Bay
,
A.
,
Bay
,
N.
,
Soflaie
,
N.
,
Mehdinejad
,
M. H.
&
Mallah
,
M.
2015
The survey of Zarin-Gol River water quality in Golestan Province using NSF-WQI and IRWQISC
.
Journal of Health in the Field
3
(
3
),
27
33
.
Shoaie
,
S. M.
2016
Simulating Heavy Metals in Reservoirs
.
PhD thesis
,
Water Engineering Group, K.N. Toosi University of Technology
,
Tehran
,
Iran
.
Torres
,
E.
,
Galván
,
L.
,
Cánovas
,
C. R.
,
Soria-Píriz
,
S.
,
Arbat-Bofill
,
M.
,
Nardi
,
A.
,
Papaspyrou
,
S.
&
Ayora
,
C.
2016
Oxycline formation induced by Fe(II) oxidation in a water reservoir affected by acid mine drainage modeled using a 2D hydrodynamic and water quality model – CE-QUAL-W2
.
Science of the Total Environment
562
,
1
12
. doi:10.1016/j.scitotenv.2016.03.209.
Zhang
,
Y.
,
Wu
,
Z.
,
Liu
,
M.
,
He
,
J.
,
Shi
,
K.
,
Zhou
,
Y.
,
Wang
,
M.
&
Liu
,
X.
2015
Dissolved oxygen stratification and response to thermal structure and long-term climate change in a large and deep subtropical reservoir (Lake Qiandaohu, China)
.
Water Research
75
,
249
258
. doi:10.1016/j.watres.2015.02.052.
Zolpirani
,
N. M.
,
Amirnejad
,
H.
&
Shahnazari
,
A.
2015
Calculating the economic value of water in paddy farms in the area of Alborz dam
.
Journal of Novel Applied Science
4
(
2
),
197
201
.