Science of the Total Environment 589 (2017) 107–116

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Introducing a water quality index for assessing water for irrigation

purposes: A case study of the Ghezel Ozan River
Farhad Misaghi a, Fatemeh Delgosha a, Mostafa Razzaghmanesh b,c,⁎, Baden Myers b
a
Water engineering Department, University of Zanjan, Iran
b
School of Natural and Built Environments, University of South Australia, Adelaide, Australia
c
ORISE Postdoc at US EPA Office of Research and Development

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• A simple direct index for assessing sur-

face water quality for irrigation demand
was developed.
• The selected parameters were provided
with a weighting based on the survey
outcomes.
• The resulting index was similar to the
original NSFWQI index.
• FAO 29 guideline for irrigation water
quality and an iterative survey employed.
• This study showed variation in water
quality from the upstream to down-
stream ends of the river.

a r t i c l e i n f o a b s t r a c t

Article history: Rivers are one of the main water resources for agricultural, drinking, environmental and industrial use. Water
Received 9 November 2016 quality indices can and have been used to identify threats to water quality along a stream and contribute to better
Received in revised form 20 February 2017 water resources management. There are many water quality indices for the assessment and use of surface water
Accepted 27 February 2017
for drinking purposes. However, there is no well-established index for the assessment and direct use of river
Available online xxxx
water for irrigation purposes. The aim of this study was to adopt the framework of the National Sanitation Foun-
Editor: D. Barcelo dation Water Quality Index (NSFWQI) and, with adjustments, apply it in a way which will conform to irrigation
water quality requirements. To accomplish this, the NSFWQI parameters for drinking water use were amended to
Keywords: include water quality parameters suitable for irrigation. For each selected parameter, an individual weighting
NSFWQI chart was generated according to the FAO 29 guideline. The NSFWQI formula was then used to calculate a final
Water quality index index value, and for each parameter an acceptable range in this value was determined. The new index was
Ghezel Ozan River then applied to the Ghezel Ozan River in Iran as a case study. A forty five year record of water quality data
Irrigation (1966 to 2010) was collected from four hydrometery stations along the river. Water quality parameters including
Na+, Cl−, pH, HCO− 3, EC, SAR and TDS were employed for water quality analysis using the adjusted NSFWQI for-
mula. The results of this case study showed variation in water quality from the upstream to downstream ends of
the river. Consistent monitoring of the river water quality and the establishment of a long term management plan
were recommended for the protection of this valuable water resource.
© 2017 Elsevier B.V. All rights reserved.

⁎ Corresponding author.
E-mail addresses: Farhad_Misaghi@znu.ac.ir (F. Misaghi), Mostafa.Razzaghmanesh@unisa.edu.au (M. Razzaghmanesh), Baden.Myers@unisa.edu.au (B. Myers).

http://dx.doi.org/10.1016/j.scitotenv.2017.02.226
0048-9697/© 2017 Elsevier B.V. All rights reserved.
108 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116
1. Introduction
The importance of rivers and surface water resources for society
has been recognized by several authors (Priscoli, 1998). For this
reason, most cities, industrial centres and agricultural activities
have been established very close to streams and other surface
water resources (FAO, 2011). However, population growth, climate
change, drought and increasing demand for water resources have
put greater pressure on water resources while simultanouesly
impairing water quality resulted from discharge of industrial and
agricultural activities in surface water resources is another barrier
(IPCC, 2007; Razzaghmanesh et al., 2005). Therefore, it is important
that monitoring is undertaken to ensure that water resources and
their quality remains within acceptable limits for sustainable end
use.
In general, for a comprehensive surface and groundwater quality
assessment, physical, chemical and biological water quality parame-
ters are required. One of the main problems of water quality studies
is the number of parameters which can be potentially monitored and
the associated time and cost of collecting, analysing and interpreting
this data (Hernández-Romero et al., 2004). To overcome this, end use
specific water quality indices have been adopted to conduct efficient
water quality classification based on a set of water quality parame-
ters which have been broadly accepted as informative for a desired
end use. Each of the many water quality indices available are
underpinned by a mathematical tool for converting the selected
quantitative water quality characteristics into a single number,
typically from 0 to 100, that provides an indication of the water qual-
ity with respect to the desired end use. Critical steps in this process
are the selection and widespread acceptance of water quality param-
eters to measure, and the consideration of a weighting factor for each
which reflects the importance of each parameter for a particular end
use (Mirzaei et al., 2005; Sharma et al., 2006) such as drinking water
supply, recreation, irrigation, livestock watering, or ecosystem
health.
There have been several water quality indexes (WQIs) proposed
and applied in practice worldwide. The first WQI was proposed by
Horton (1965) and was intended for drinking water supply assess-
ment. Brown et al. (1970) subsequently suggested a WQI which
could be used as a standard measure to compare the water quality
of different water bodies. Since these early WQIs were developed,
several more indexes have been proposed. These include the Unites
States National Sanitation Foundation Water Quality Index
(NSFWQI), the Florida Stream Water Quality Index (FWQI), the Brit-
ish Columbia Water Quality Index (BCWQI), the Canadian Water
Quality index (CWQI) and the Oregon Water Quality Index (OWQI)
as described by Cude (2002). Table 1 describs additional specifica-
tion about water quality inecies.
Each of these indices are frequently applied worldwide. WQIs are
structured to be interpreted in one of two ways. In the first, an increas-
ing number shows a higher level of pollution. This is called a ‘pollutant
index’, and the BCWQI is an example of this group. The alternative are
called ‘quality indices’ where an increasing index represents a lower
level of pollution. The NSFWQI and OWQI are examples of this category.
Of the indexes developed, the NSFWQI is one of the most common
comprehensive indices for surface water quality classifications world-
wide. The NSFWQI is used to compare the water quality of different
water bodies and is therefore generic in nature. The NSFWQI is deter-
mined based on nine water quality parameters including temperature,
−
pH, turbidity, PO2−
4 , NO3 , TDS, DO, BOD5 and faecal coliform concentra-
tions. The NSFWQI index has been used in several studies to indicate the
existing water quality condition of rivers (Brown et al., 1973;
Zandbergen, 1998; Mirzaei et al., 2005).
According to Jena et al. (2013), water quality indices may be cat-
egorized into four main groups. The first group do not consider the
end use of the water and are used for general water quality
assessment (Ott, 1978). Examples include the NSFWQI. The second
category is targeted at a specific type of water demand, such as
drinking, industrial, irrigation and ecosystem preservation. Exam-
ples of this type of index include the OWQI (DEQ, 2003). The third
category is aimed at designing or planning indices. These act as an in-
strument in planning water quality management projects and help-
ing decision makers to determine how to manage a water resource.
As described by Zandbergen (1998), the BCWQI is an example of
this category. These first three categories are all based on what is
known as the expert opinion approach or Delphi approach, where
water quality parameters are selected and different weightings
reflecting the importance of each are assigned based on feedback
from a panel of experts. The feedback is repeated over more than
two iterations. They therefore represent a subjective approach
(Harkins, 1974), where objectivity and comparability are possible
because of the different ratings given by experts. The fourth category
however differs because indices do not consider personal opinion,
but are supported by statistical methodology. The essential part of
the statistical approach in this fourth category of WQI is to assign a
relevant range of scores and make required assumptions to cover
all expected water quality observations (Terrado et al., 2010). Multi-
variate and generalized linear models are exaples of statistical
methods (Shrestha and Kazama, 2007).
Zandbergen (1998) conducted an ecological risk assessment of an
urban watershed using a GIS tool. The study focused on the Brunette
River watershed in British Columbia, Canada. A generic conceptual
model was developed and a set of key indicators were selected in-
cluding impervious areas, riparian habitat, pollutant loadings,
water quality, sediment quality, and public health. In this study the
BCWQI idex was employed for water quality analysis of the river. In
another study by Shokuhi et al. (2012) the water quality of a lake
retained by the Aydoghomosh Dam in north west Iran was assessed.
The retained water was used to supply potable water for nearby vil-
lages and irrigation water for agricultural lands downstream. Water
quality was assessed due to concenrs about pollutant discharges into
the reservoir. Standard field parameters including dissolved oxygen,
temperature, biochemical and chemical oxygen demand, most prob-
able number of coliforms, faecal coliforms, turbidity, total dissolved
solids, total solids, pH, electrical conductivity and others were mea-
sured at eight different stations during spring and summer in 2010.
Sampling points were selected on the basis of their importance.
The NSFWQI was calculated based on results of the sample collection
and analysis program. The NSFWQI of the collected samples ranged
from 84.89 in in July to 67.96 in May. The lower value of WQI has
been found mainly due to the slightly lower value of DO in the dam
reservoir water. Most of the water samples were found within the
good category of the NSFWQI. The results of sampling showed an ac-
ceptable water quality for a variety of purposes including drinking
water and irigation.
In another study by Sánchez et al. (2007) a two simple WQIs (an
empirical WQI and a dissolved oxygen deficit) recommended
by Pesce and Wunderlin (2000) were assessed along the
Guadarrama and Manzanares rivers in Spain. In this study, when
the values of the WQI were in the range of 0 to 25, the water must
be classified as “very bad”; for a WQI value in the range of 25 to 50
the water was classified as “bad”; for WQI values in the range of 51
to 70 the water classification was “medium” and finally, when the
WQI values were within the range of 71 to 90 and 91 to 100 the
water is classified as “good” and as “excellent”, respectively. The
results were a value of 70 (higher quality) at the upstream end of
the Guadarrama river which declined as water moved downstream
to a value of 64 (or ‘average’ quality) at the downstream end of the
river (Sánchez et al., 2007).
At present there is no established integrated index that considers
a combination of parameters (Ayers and Westcot, 1994; Stoner,
1978) for surface water quality assessment for the purposes of crop
 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116 109

Table 1
additional specification about water quality indeces.

Index Aim Surveyed people Parameters

Unites States National Sanitation
Foundation Water Quality
Index (NSFWQI)
Florida Stream Water Quality
Index (FWQI)
British Columbia Water Quality
Index (BCWQI)
Canadian Water Quality index
(CWQI)
Oregon Water Quality Index
(OWQI)
To provide a standardized method for More than 140 water quality DO, faecal coliform, pH, biochemical oxygen demand
comparing the relative quality of various scientists were surveyed about 35 (BOD), temperature change, total phosphate, nitrate,
bodies of water water quality tests turbidity, and total solids
• It is an arithmetic average of water clarity (turbidity and total suspended solids), dissolved oxygen; oxygen-demanding substances
(Biological Oxygen Demand [BOD], Chemical Oxygen Demand [COD], Total Organic Carbon [TOC]), nutrients (phosphorus and nitrogen),
bacteria (total and faecal coliform), and biological diversity (natural or artificial substrate macroinvertebrate diversity and Beck's Biotic
Index)
• The values for this index were determined as follows: 0 to b45 represents good quality, 45 to b60 represents fair quality, and 60 to 90
represents poor quality (SAFE, 1995)
• The BCWQI was found to be extremely sensitive to sampling design and highly dependent on the specific application of water quality
objectives.
• The BCWQI in its form has serious limitations for comparing water bodies and for establishing management priorities
The Index incorporates three elements: scope (the number of variables not meeting water quality objectives); frequency (the number of
times these objectives are not met); and amplitude (the amount by which the objectives are not met). The index produces a number
between 0 (worst water quality) and 100 (best water quality). These numbers are divided into 5 descriptive categories to simplify
presentation. The specific variables, objectives, and time period used in the index are not specified and indeed, could vary from region to
region, depending on local conditions and issues. It is recommended that at a minimum, four variables sampled at least four times be used in
the calculation of index values
The OWQI is calculated in two steps. The raw analytical results for each variable, temperature, DO, BOD, pH, ammonium nitrate nitrogen,
having different units of measurement, are transformed into unitless subindex total phosphates, total solids, and faecal coliform
values. These values range from 10 (worst case) to 100 (ideal). These subindices
are then combined to give a single WQI value ranging from 10 to 100. T

irrigation with consideration of soil salinity and drainage. In this
study, we describe the development of an irrigation specific WQI de-
veloped based on an adjustment of the existing NSFWQI index with a
specific focus on agricultural water needs.
The study was based in Iran which has been subject to periodic
drought salinity and growing concerns regarding climate change
(Amiri and Eslamian, 2010; Smedema and Shiati, 2002). In response,
the protection of surface water resources are of interest to authorities
and policy makers. A new index for irrigation water has been proposed
to evaluate the fitness of water quality specifically for irrigation both
historically and into the future. The Ghezel Ozan River is an important
source of irrigation water for the horticulture industry which requires
a good quality water supply to maintain productivity. This study there-
fore aims to develop a water quality index for agricultural waters, based

Fig. 1. Location of the hydrometry stations in the Ghezel Ozan River basin, indicating landuse.
110 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116

Table 2 Table 4
Guidelines for interpretation of water quality for irrigation (Ayers and Westcot, 1994). NSFWQI index range and associated water quality category.

Potential irrigation problem Units Degree of restriction on use NSFWQI index value Water quality category

None Slight to moderate Severe 0–25 Very bad

26–50 Bad
Salinity (affects crop water availability)
51–70 Average
Ecw dS/m b0.7 0.7–3.0 N3.0
71–90 Good
or
91–100 Very good
TDS mg/l b450 450–2000 N2000

Infiltration (affects infiltration rate of water into the soil. Evaluate using Ecw and
SAR together)
SAR =0–3 and Ecw = N0.7 0.7–0.2 b0.2
=3–6 = N1.2 1.2–0.3 b0.3
=6–12 = N1.9 1.9–0.5 b0.5 factors. Soils may become saline as a result of land use, including
=12–20 = N2.9 2.9–1.3 b1.3
the use of irrigation water with high levels of salt. By definition, a sa-
=20–40 = N5.0 5.0–2.9 b2.9
line soil contains excess soluble salts that reduce the growth of most
Specific ion toxicity (affects sensitive crops) crops. These soluble salts contain cations such as sodium (Na+), po-
Sodium (Na+)
tassium (K+), calcium (Ca2 +) and magnesium (Mg2 +) along with
Surface irrigation SAR b3 3–9 N9
Sprinkler irrigation me/l b3 N3 anions chloride (Cl-) and bicarbonate (HCO-3). Presence of these pa-
Chloride (Cl−) rameter can also reduce the soil cation exchangeable capacity or in-
Surface irrigation me/l b4 4–10 N10 crease soil dispersion that will reduce a soil infiltration capability.
Sprinkler irrigation me/l b3 N3 Therefore, considering the most important soil salinity parameters
Boron (B) mg/l b0.7 0.7–3.0 N3.0
Trace Elements
in a comprehensive representative index was aimed.
Seven important parameters regarding irrigation water quality as
Miscellaneous effects (affects susceptible crops) suggested in FAO 29 and not currently included in the NSFWQI were
Nitrogen (NO2− − N) mg/l b5 5–30 N30
3
Bicarbonate (HCO−
considered including Na+, Cl−, pH, HCO− 3 , Ec, SAR and TDS. A weighting
3 )
(overhead sprinkling only) me/l b1.5 1.5–8.5 N8.5 factor was also considered for each of the seven parameters of the new
pH Normal Range 6.5–8.4 index.
To determine the most appropriate parameters of the index for irri-
gation use and to establish a weighting system for each parameter, a se-
ries of questionnaires were prepared and provided to a group of 50
water specialists. The results of this survey were used for prioritizing
on a an existing and broadly accepted index, the NSFWQI. The study seven parameters, and subsequently to develop a weighting system
then applies the index to the Ghezel Ozan River as a case study of how for each. In a similar approach to the development of the NSFWQI
it may be used to assess the current status and any trend in water qual- index, after receiving and processing the first round of survey responses,
ity in an existing irrigation resource. Water quality data was examined a second survey was conducted to allow people to adjust their response
based on monthly and yearly trends and compared with the findings based on the outcomes of the first survey.
of the NSFWQI. The resulting data produced a complex multi-criteria analysis prob-
lem with several responses regarding appropriate paramters and their
weighting.
2. Material and methods
The results of this survey was used for prioritizing and developing a
weighting system. Also, analytical hierarchy process (AHP) and expert
2.1. Methodology
choice software was employed to extract NSFWQI index weighting's pa-
rameters as it has been illustrated in Table 3. Generally solving multi
The index in this study was developed based on the existing NSFWQI
criteria decision making problems is very complicated and in most of
index. The NSFWQI index was proposed by Brown et al. (1970) and sup-
the cases increased accuracy of one of the factors may effects on the ac-
ported by the United States National Sanitation Foundation. It was
curacy of the other parameters. However, multi criteria decision meth-
developed based on an iterative survey of experts from different back-
od and in specific multi attributes decision making is generated to solve
grounds in a process known as a Delphi study. The survey initially pro-
these sorts of problems. Multi criteria methods pose variety of tech-
posed 35 water quality parameters based on initial responses from the
niques during the different stages of decision making.
experts. After three survey rounds, parameters were progressively
According to Table 3, the maximum weight belongs to SAR (0.30)
discarded and the nine parameters previously stated were selected for
and the minimum one is for pH (0.05). Also depending on the values
the index.
of the recorded parameters in each station along the river the data
The development of the amended index used in this study and
was normalized and used for calculating NSFWQI index from Eq. (1).
specific to irrigation was undertaken by taking the existing
The final value of the irrigation water quality index was calculated
NSFWQI and adjusting it with consideration of FAO 29 recommenda-
from Eq. (1), which is similar to the original NSFWQI index (Brown
tions for irrigation waters. Guidelines for interpretations of water
et al., 1970).
quality for irrigation presented by FAO 29 were referred to in this
case and are presented in Table 1. Soil salinity is caused by several

Table 5
properties of the hydrometry stations along the Ghezel Ozan River.

Table 3 Station Latitude Longitude Height above sea level (m)

Factors used for calculating NSFWQI index. Bianloo 47° 51′ 36° 00 1540
Ostoor 47° 59′ 37° 31′ 1000
Parameters SAR Na pH Cl HCO3 EC TDS
Gharehgoni 47° 57′ 36° 14′ 1410
Weight factor 0.30 0.10 0.05 0.10 0.10 0.25 0.10 Laylan 47° 40′ 36° 45′ 1250
 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116 111

Table 6
Z-test of Mann-Kendall method along the Ghezel Ozan River from 1966 to 2010.

Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Bianloo −0.19 −0.17 008.0 24.0 68.0 39.1 45.0 59.0 03.0 −0.24 2.0 11.0
Ostoor 54.0 99.0 84.0 98.0 73.0 84.0 67.0 15.0 −0.26 08.0 19.0 07.0
Gharehgooni 87.0 06.0 1.0 37.0 19.0 25.0 0.18 −1.56 2.0 −0.1 22.0 22.0
Laylan 8.0 37.1 85.0 4.2 8.1 18.2 42.0 61.0 32.0 −2.23 −0.57 19.1

Table 7
Trend line slope of water quality parameters associated with proposed NSFWQI index.

Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Bianloo 1.76 −0.68 −0.38 −0.18 2.20 2.38 3.53 0.92 2.22 0.14 −0.73 0.81
Ostoor 2.32 2.20 2.17 2.75 3.05 1.53 3.22 0.32 0.57 −1.06 0.17 0.88
Gharehgooni −0.42 2.50 1.91 0.730 0.92 0.29 1.18 2.48 −2.47 1.39 −0.49 2.08
Laylan 0.18 0.31 – 1.74 1.09 2.08 2.41 2.14 – 0.18 – −0.47

Like the original NSFWQI, the proposed new index ranges from 0 to
100. In order to provide a qualitative description of the index outcome,
plain descriptions for index data were developed and these are shown
in Table 4.
n
∑ Wi Ii
NSFWQI ¼ i¼1
ð1Þ
∑Wi
In this formula, n: number of parameters, W: weight factor, I: pa-
rameter value.
2.2. Applying the irrigation water quality index – a case study
Following development of the index, it was applied to a case study
area to examine the current water quality and any historical trends.
The case study was carried out on the Ghezel Ozan River located in
the north west of Iran. The Ghezel Ozan River is one of the most impor-
tant sources of irrigation water for the agriculture and horticulture in-
dustries in Iran and it flows throughout the year. The river is 765 km
long with an average longitudinal slope of 0.30%. It originates in the
north west of Sanandaj, Kordestan Province and continues downstream
in a north east direction through Zanjan County and Tarom County in
Zanjan Province, and then eastwards through the Alborz mountain
range in Ardabil Province and Guilan Province. At Rudbar, south west
of the Alborz mountain, it joins the Shahrood river to form the Sefid-
Rud river. Then the Sefid-Rud flows northwards in Guilan Province,
through the Alborz range to the Caspian Sea. The case study applies
the proposed new irrigation water quality index to the Ghezel Ozan
River using data from four hydrometery stations, each measuring daily
flow, physical and chemical water quality.These include the Bianloo,
Gharehgoni, Laylan and Ostoor stations, eah maintained by the Iranian
Ministry of Energy. The properties and location of these stations are
shown in in Table 5 and Fig. 1.The water quality information used to cal-
culate the proposed NSFWQI index was a record of up to 44 years of
monthly water quality data from 1966 until 2010.
2.3. Data analysis
To examine the current water quality, the current value of the re-
vised NSFWQI for irrigation water was calculated. To examine trends

Fig. 2. Proposed NSFWQI index values at four hydrometry stations from year 1966 to 2009.
112 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116

Fig. 3. Monthly value of the proposed irrigation index at the Bianloo staion during the study period.

in water quality over the period of data analysis, the historical record of
the water quality index was plotted with time and visually assessed.
Further investigation was conducted on the historical record of the
revised NSFWQI index to quantitatively determine any trends in the
value of the proposed index over time. Generally, the availability of a
trend is defined as uniform changes within a series of hydrological pa-
rameters. Because of trend analysis capabilities, a majority of hydrolog-
ical series follow a non-normal distribution. Although there are various
methods for trend investigation, generally in hydro-metrological stud-
ies a non-parametric method is used (Razzaghmanesh et al., 2016).
Among available non-parametric methods, the Mann-Kendall trend
test has been suggested to be one of the best methods (Helsel and
Hirsch, 1992). As such, the Mann-Kendall trend test has been adopted
to identify trends among data from the Ghezel Ozan River.
3. Results and discussion
The results of the survey data analysis indicates that the seven most
appropriate paramters for the revised NSFWQI index suitable for
irrigation included the SAR, Na, pH, Cl, HCO3, EC and TDS. Using the sur-
vey data collected in conjunction with the analytical hierarchy process
in the selected expert choice software, a weighting was provided to
each parameter. The weightings are summarized in Table 3. According
to Table 2, the results of the survey suggest that the maximum
weighting belongs to SAR (0.30) and the minimum is for pH (0.05).
3.1. Developng proposed NSWQI index
3.1.1. Application of the proposed NSFWQI index to the Ghezel Ozan River
The results of the proposed irrigaiton index at the four stations along
the Ghezel Ozan River are shown in Fig. 2, showing the annual average
value of the index at each station. When compared with the guideline
data in Table 2, it is apparent that for the purposes of crop irrigation,
the river water quality at the Ostoor station has typically been in aver-
age condition, and the rising trend of NSFWQI index shows that becom-
ing good in more recent years. While the trend shows the Laylan station
has been historically bad, but becoming better similar to Ostoor station
in recent years. The index Bianloo and Gharegoni ranges form bad to

Fig. 4. Average monthly value of the proposed irrigation index based on data from the Bianloo station during the study period.
 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116 113

Fig. 5. Monthly value of the proposed irrigation index at the Ostoor staion during the study period.

average condition for irrigation. Comparison of these results with ob-
served water quality data in the stations are consistent.
3.1.2. Proposed NSFWQI for irrigation results for the Bianloo station
The results in Fig. 3 show that the maximum valu eof the proposed
irrigation index on a monthly basis for the river based on water quality
data at the Bianloo station. The index shows that water quality for irri-
gation ranges from very bad to average water quality in most months
before 1997. After this time, irrigation water quality reaches a good con-
dition in some months, but still reaches a very bad condition at times.
Fig. 4 shows the average value of the index on a monthly basis, consid-
ering all data throughout the sampling period (1970 to 2009). The
highest (best) and lowest (worst) water quality for irrigation has been
recorded in July and November, respectively. The maximum monthly
value of the index was based on data from July 2000.
3.1.3. Proposed NSFWQI for irrigation results for the Ostoor station
The maximum monthly value of the proposed irrigaiton WQI at the
Ostoor station are shown in Fig. 5. Similar to the Bianloo Station, the
data shows varying values of the index over the sampling duration.
The water quality of the river at the Ostoor station ranges from very
bad to good levels. The trend of the proposed proposed irrigation WQI
and its variablity is a function of drier and wetter years. Fig. 6 shows
the average value of the proposed irrigation WQI with respect to the
month it was measured throughout the sample period. At the Ostoor
station, due to snow melt and subsequent increased water flow, the
index showed higher values in December to January. Of particular
note that is the fact that the water quality of the river reaches a mini-
mum value in July, corresponding with a peak in demand for irriation
water at this time.
3.1.4. Proposed NSFWQI for irrigation results for the Ghareghoni station
The maximum monthly results for the proposed irrigation WQI at
the Ghareghoni station are shown in Fig. 7. Two phases are evident
from the data – one before 1995 and a period after 1995. The data be-
fore 1995 showed a very bad to average range in water quality of the
river. After 1995, the river water quality is in good condition and the
index was about 77 in 1977. Fig. 8 shows the average value of the
proposed irrigation WQI with respect to the month it was measured
throughout the sample period. The highest values were found to

Fig. 6. Average monthly value of the proposed irrigation index based on data from the Ostoor station during the study period.
114 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116

Fig. 7. Monthly value of the proposed irrigation index at the Gharehgoni staion during the study period.

occur in January and February and the lowest value has been report-
ed for June. This indicates that the value of the proposed irrigation
index was dependent on the season and precipitation regime.
3.1.5. Proposed NSFWQI for irrigation results for the Laylan station
Fig. 9 indicates that based on the proposed irrigation index, the
water quality at the Laylan Station ranges fomr very back to good for ir-
rigation in a cyclical manner. The highest recorded value was in 2002
where the proposed irrigation WQI reached 94. Fig. 10 shows the aver-
age value of the proposed irrigation WQI with respect to the month it
was measured throughout the sample period. The water quality for irri-
gation is generally best in September and at its worst in November as it
was one of the low rainall year.
3.2. Trend analysis
The Mann-Kendall test was employed to study trends on the pro-
posed NSFWQI index along the Ghezel Ozan River. For this purpose,
the previously presented long term water quality data from 1966 to
2010 was analysed at each of the monitoring stations. The results of
the Mann-Kendall trend analysis for the NSFWQI index variations
along the Ghezel Ozan River are reported in Table 6, representing
monthly averages of water quality parameters of the river. The positive
values showed better quality and the negative values showed the lower
water qualities.
The Trend line slopes of Ghezel Ozan River water quality are report-
ed in Table 6. The results showed varied trends in the proposed irriga-
tion index values among the months of the year at each station. The
highest values were 2.40 for the Laylan station in March and the lowest
value was −2.23 in October and also in the Laylan station. Furthermore,
the β coefficient for the trend line slope calculated from the Mann-
Kendall test is reported at the Table 7.
According to Table 7, in the months of March, April, May, June and
December, the values have an increasing trend. This indicates that in
these months water quality conditions for irrigation have been in an im-
proving or stable situation. These results reflect positively on the condi-
tion of the river for irrigation purposes.
Fig. 11 shows the water quality index trend in April for the Laylan
station. The steep slope of the trend line represents the improving con-
dition of the river in this month. As we can see from the Table 6, for

Fig. 8. Average monthly value of the proposed irrigation index based on data from the Gharehgoni station during the study period.
 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116 115

Fig. 9. Monthly value of the proposed irrigation index at the Laylan staion during the study period.

other months of the other station the data did not show an appropriate
trend and it is completely different. Based on these findings, it is not
possible to provide a reliable objective comment on improvements or
degradation of irrigation water quality in other parts of the year. There
is however some evidence that for the months of July, August, Septem-
ber, October and December the river water quality is improving, and in
some cases it has been suitable. However, variation in the data reveals
an overall unclear behaviour that may cause a serious problem for
river managers and end-users in coming years in terms of planning for
current irrigation management and expansion.
Trends of the poorest month with descending trend is illustrated in
Fig. 12. As it has shown the slope of the trend lines − 2.23 and this
value is showing a decreasing water quality at this month (October)
over time.
4. Conclusion
In this study a simple direct index for assessing surface water quality
for irrigation demand was developed and applied to a case study region
of Iran. To develop a proposed new irrigation index, the existing
NSFWQI index was adopted and the underlying paramters were param-
eters adjusted to comply with crop irrigation water quality require-
ments using recommendations of the FAO 29 guideline for irrigation
water quality and an iterative survey of irrigation experts. The selected
parameters were provided with a weighting based on the survey
outocmes. The resulting index was similar to the original NSFWQI
index, in that it provides a rating of between 0 and 100 to provide a sin-
gle indicative value of water quality. It is intended that the proposed
index can be of use to water managers for monitoring surface waters
in areas where irrigation represents an end use. It is possible to use
this index in other rivers and with using this method it would be quickly
possible to provide estimation of the water quality of the surface water
resources if the historical data is available. It is recommended that prior
of start of each year growing season, based on the calculated index for
that year, the crop pattern must be adjusted for sustainable irrigation.
Application to the Ghezel Ozan River in Iran revealed that water quality
has raged from very bad to good at four station along the river. Water
quality is cyclical, and clearly affected by seasonal patterns. There is ev-
idence of water quality improvement in some months of the year. It is
recommended that the index is appled to validate the results of this

Fig. 10. Average monthly value of the proposed irrigation index based on data from the Laylan station during the study period.
116 F. Misaghi et al. / Science of the Total Environment 589 (2017) 107–116
Fig. 11. Water quality variations associated with proposed NSFWQI index in Laylan
Hydrometry station in March for the period of this study (1971–2010).
Fig. 12. Water quality variations associated with proposed NSFWQI index in Laylan
Hydrometry station in October for the period of this study (1971–2010).
study. Further imporvmenet may be achieved by investigating other in-
dices on which to base an irrigation index.
Also, developing such an index will help to revise the crop pattern of
irrigating land to comply with the water quality of that year. More im-
portantly adoption of the available methods will increase the accuracy
of the method and therefore more reliable results will be concluded.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.scitotenv.2017.02.226.
References
Amiri, M.J., Eslamian, S.S., 2010. Investigation of climate change in Iran. J. Environ. Sci. Technol. 3
(4), 208–216.
Ayers, R., Westcot, D., 1994. Water quality for agriculture. FAO Irrigation and Drainage Paper 29
Rev. 1. Food and Agriculture Organization of the United Nations, Rome, Italy.
Brown, R.M., McClelland, N.I., Deininger, R.A., Tozer, R.G., 1970. A water quality index - do we
dare. Water Sew. Works 339–343.
Brown, R.M., McClelland, N.I., Deininger, R.A., O'Connor, M.F., 1973. A water quality index crash-
ing the psychological barrier. Indic. Environ. Qual. 173–182.
Cude, Curtis G., 2002. Oregon Water Quality Index: a tool for evaluating water quality manage-
ment effectiveness. J. Am. Water Resour. Assoc. 37 (1), 125–137.
DEQ, 2003. The Oregon Department of Environmental Quality, Oregon. Available from:. http://
www.deq.state.or.us/lab/WQM/WQI/Wqi main.htm (viewed 06/04/2015).
FAO, 2011. The State of the World's h Land and Water Resources for Food and Agriculture
(SOLAW) – Managing Systems at Risk. Food and Agriculture Organization of the United Na-
tions, Rome and Earthscan, London.
Harkins, R.K., 1974. An objective water quality index. J. Water Pollut. Control Fed. 46 (7),
578–588.
Helsel, D.R., Hirsch, R.M., 1992. Statistical Methods in Water Resources. Vol. 49. Elsevier.
Hernández-Romero, A.H., Tovilla-Hernández, C., Malo, E.A., Bello-Mendoza, R., 2004. Water
quality and presence of pesticides in a tropical coastal wetland in southern Mexico. Mar.
Pollut. Bull. 48 (11–12), 1130–1141.
Horton, R.K., 1965. An index number system for rating water quality. Water Pollut. Control Fed.
37, 300–306.
Jena, V., Dixit, S., Gupta, S., 2013. Assessment of water quality index of industrial area surface
water samples. Int. J. Chem. Technol. 5 (1), 278–283.
IPCC, 2007. Climate change 2007, impacts, adaptation and vulnerability. Working Group II to the
Fourth Assessment Report of the IPCC.
Ott, W.R., 1978. Environmental Indices: Theory and Practice.
Pesce, S.F., Wunderlin, D.A., 2000. Use of water quality indices to verify the impact of Córdoba
City (Argentina) on Suquı́a River. Water Res. 34 (11), 2915–2926.
Priscoli, J.D., 1998. Water and civilization: using history to reframe water policy debates and to
build a new ecological realism. Water Policy 1, 623–636.
SAFE, 1995. Strategic assessment of Florida’s environment, Florida stream water quality index,
statewide summary. available at:. http://www.pepps.fsu.edu/safe/environ/swq1.html.
Sánchez, E., Colmenarejo, M.F., Vicente, J., Rubio, A., García, M.G., Travieso, L., et al., 2007. Use of
the water quality index and dissolved oxygen deficit as simple indicators of watersheds
pollution. Ecol. Indic. 7, 315–328.
Sharma, M.P., Sharma, S., Goel, V., Sharma, P., Kumar, A., 2006. Water quality assessment of
Behta River using benthic macroinvertebrates. Life Sci. J. 3, 68–74.
Shokuhi, R., Hosinzadeh, E., Roshanaei, G., Alipour, M., Hoseinzadeh, S., 2012. Evaluation of
Aydoghomush Dam Reservoir water quality by National Sanitation Foundation Water Qual-
ity Index (NSF-WQI) and water quality parameter changes. Iranian J. Health Environ. 4 (4),
439–450.
Smedema, L.K., Shiati, K., 2002. Irrigation and salinity: a perspective review of the salinity haz-
ards of irrigation development in the arid zone. Irrig. Drain. Syst. 16 (2), 161–174.
Stoner, J.D., 1978. Water-quality Indices for Specific Water Uses. Department of the Interior,
Geological Survey.
Shrestha, S., Kazama, F., 2007. Assessment of surface water quality using multivariate statistical
techniques: a case study of the Fuji river basin, Japan. Environ. Model Softw. 22, 464–475.
Terrado, M., Barceló, D., Tauler, R., Borrell, E., de Campos, S., 2010. Surface-water-quality indices
for the analysis of data generated by automated sampling networks. TrAC Trends Anal.
Chem. 29 (1), 40–52.
Zandbergen, Paul A., 1998. Urban watershed ecological risk assessment using GIS: a case study
of the Brunette River watershed in British Columbia, Canada. J. Hazard. Mater. 61, 163–173.
Razzaghmanesh, M., Mohammai, K., Samani, J.M.V., 2005. A river quality simulation using
WASP6: case study. 9th Environmental Engineering Specialty Conference, Toronto, Ontario,
Canada. 2nd–4th June.
Razzaghmanesh, M., Beecham, S., Myers, B., 2016. Long term effects of green roofs on
stormwater quality from two sites in Australia. World Environmental and Water Resources
Congress 2016. May 2016, pp. 371–379.
Mirzaei, M., Nazari, A., Yari, A., 2005. Water quality zoning of Jajrood River. J. Environ. Stud. 37,
17–26.