Statistical Tools For Analyzing Water Quality Data

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/224829937
Statistical Tools for Analyzing Water Quality Data
Chapter · April 2012

DOI: 10.5772/35228 · Source: InTech
CITATIONS READS
23 17,044
2 authors:
Liya Fu You-Gan Wang

Xi'an Jiaotong University Queensland University of Technology
31 PUBLICATIONS 207 CITATIONS 236 PUBLICATIONS 4,501 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Monitoring Queensland Mackerel Fisheries View project
PBVRMF View project
All content following this page was uploaded by You-Gan Wang on 30 May 2014.
The user has requested enhancement of the downloaded file.

0
6
Statistical Tools for Analyzing Water Quality Data

Liya Fu1 and You-Gan Wang2 *
1 School of Science, Xian Jiaotong University, China and Centre for Applications in Natural
Resource Mathematics (CARM), School of Mathematics and Physics,
the University of Queensland
2 Centre for Applications in Natural Resource Mathematics (CARM), School of
Mathematics and Physics, The University of Queensland
Australia
1. Introduction
Water quality data are often collected at different sites over time to improve water
quality management. Water quality data usually exhibit the following characteristics:
non-normal distribution, presence of outliers, missing values, values below detection limits
(censored), and serial dependence. It is essential to apply appropriate statistical methodology
when analyzing water quality data to draw valid conclusions and hence provide useful
advice in water management. In this chapter, we will provide and demonstrate various
statistical tools for analyzing such water quality data, and will also introduce how to use
a statistical software R to analyze water quality data by various statistical methods. A
dataset collected from the Susquehanna River Basin will be used to demonstrate various
statistical methods provided in this chapter. The dataset can be downloaded from website
http://www.srbc.net/programs/CBP/nutrientprogram.htm.
2. Graphical analysis of water quality data

Graphs provide visual summaries of data, quickly and clearly describe important information
contained in the data, and provide insight for the analyst into the data under scrutiny. Graphs
will help to determine if more complicated modeling is necessary. In this section, three
particularly useful graphical methods are presented: boxplots, scatter plots, and Q-Q plots.
R codes for plotting graphs in the following subsections will be given in detail.
2.1 Boxplots
A boxplot is a very useful and convenient tool to provide summaries of a dataset and is
often used in exploratory data analysis. A boxplot usually presents a dataset through five
numbers: extreme values (minimum and maximum values), median (50th percentile), 25th
percentile, and 75th percentile. It also indicates the degree of dispersion, the degree of skew,
and unusual values of the data (outliers). Furthermore, boxplots can display differences
* Address
for correspondence: Centre for Applications in Natural Resource Mathematics (CARM), School of
Mathematics and Physics, the University of Queensland, St Lucia, QLD 4072, Australia
144
2 Water Quality Monitoring and Assessment
Will-be-set-by-IN-TECH
between different populations without making any assumptions of the underlying statistical
distribution. Boxplots of concentrations of total phosphorus (mg/L) at four stations from the
Susquehanna River Basin from 2005 to 2010 are constructed (Fig. 1). R codes for constructing
Fig. 1 are as follows:
> yl<- “Concentrations of total phosphorus (mg/L)”
> boxplot(TP ∼ Station, ylab = yl, data = dat, boxwex = 0.5, outline = TRUE)
In Fig. 1, it can be seen that the four stations have nearly identical median values, and outliers
could be present at all four stations. The distributions are right skewed. Further details on
construction of a boxplot can be found in McGill et al. (1978), and Tukey (1977). More details
for plotting boxplots are available in Adler (2009), Crawley (2007), and Venables & Ripley
(2002).
1.0
Concentrations of total phosphorus (mg/L)
0.8
0.6
0.4
0.2
0.0
Station 1 Station 2 Station 3 Station 4
Fig. 1. Boxplots for total phosphorus at four stations at the Susquehanna River Basin
2.2 Scatter plots
A scatter plot is a very useful summary of a set of bivariate data (two variables), usually
drawn before obtaining a linear correlation coefficient or fitting a regression line. It can be
used to detect whether the relationships between two variables are linear or curved, and aids
the interpretation of the correlation coefficient or a regression model. Fig. 2 is a scatter plot of
the concentration of total phosphorus (mg/L) versus instantaneous flow (feet3 /s) in log scale
at Station 1.
> xl<- “Instantaneous flow on log scale in cubic feet per second”
> plot(log(UNA$Flow), UNA$TP, xlab = xl, ylab = yl)
> points(log(UNA$Flow)[39], UNA$TP[39], col=2, pch =16)
Statistical Tools
Statistical Tools for for
Analyzing Analyzing
Water Quality Data Water Quality Data 1453
We can generate a scatter plot using the data from two stations and can also use the function
0.30
0.25
0.20
0.15
0.10
0.05
0.00
5 6 7 8 9
Instantaneous flow on log scale in cubic feet per second
Fig. 2. Scatter plot of total phosphorus and instantaneous flow at Station 1, three possible
outliers are in red
xyplot (Venables & Ripley, 2002) to split the data into different panels based on station (Fig. 3
(a), (b)).
> plot(log(UNA$Flow), UNA$TP, xlab = xl, ylab = yl, type = “p”, pch = 1, col = 1)
> points(log(CKL$Flow), CKL$TP, pch = 3, col = 2)
> legend(4.5, 0.3, c(“Station 1”, “Station 2”), pch = c(1, 3), col=c(1, 2))
> library(lattice)
> xyplot(TP ∼ log(Flow)|Station, data = dat, xlab = xl, ylab = yl, col = 1)
Concentration varies with natural log of instantaneous flow, as illustrated using a scatter plot.
A linear regression model could be used to fit the data in Fig. 2, but true changes in slope are
difficult to detect from only a scatter plot. Various methods have been developed to construct
a central line to detect variation of slope locally in response to the data themselves, such as the
locally weighted scatter plot smoothing (LOWESS) method (Fig. 4) (Cleveland et al., 1992).
> plot(log(UNA$Flow), UNA$TP, xlab = xl, ylab = yl)
> lines(lowess((UNA$TP) ∼ log(UNA$Flow)), col=2)
2.3 Q-Q plots

A Q-Q plot presents the quantiles of a dataset against the quantiles of another dataset
(Chambers et al., 1983; Gnanadesikan & Wilk, 1968). It can be used to determine whether
two datasets come from populations with the same distribution. The greater the departure
from the reference line, the greater the evidence to conclude that these two datasets come
from populations with different distributions. If their distributions are identical, the Q-Q plot
follows a straight line. Q-Q plots can be applied to compare the distribution of a sample
146
4 6 8 10
Station 3 Station 4
1.0
0.30
0.8
Station 1
Station 2
0.6

0.25
0.4
0.2
0.20
0.0
Station 1 Station 2
0.15
1.0
0.8
0.10
0.6
0.05
0.4
0.2
0.00
0.0
5 6 7 8 9
4 6 8 10
Instantaneous flow on log scale in cubic feet per second Instantaneous flow on log scale in cubic feet per second
(a) two stations (b) four stations
Fig. 3. Scatter plots of total phosphorus and instantaneous flow

0.30
0.25
0.20
0.15
0.10
0.05
0.00
5 6 7 8 9
Instantaneous flow on log scale in cubic feet per second
Fig. 4. The data of Fig. 2 fitted with the locally weighted scatter plot smoothing method
to a theoretical distribution (often a normal distribution). Therefore, Q-Q plots provide a

very efficient way to tell how a sample distribution deviates from an expected distribution.
The advantages of Q-Q plots are that (a) the sample sizes of two datasets do not need to be
equal; (b) many distributional aspects can be simultaneously tested, such as shifts in location
and scale and changes in symmetry; (c) the presence of outliers can also be detected. The
functions qqnorm and qqplot can be used to construct a Q-Q plot (Adler, 2009; Crawley, 2007;
Venables & Ripley, 2002). Fig. 5 (a) and (b) are two Q-Q plots to test whether the distributions
of total phosphorus concentrations and the values of on the natural log scale at Station 1 are
Statistical Tools
Analyzing Analyzing
normal, respectively.
> qqnorm(UNA$TP); qqline(UNA$TP, col = 2)
> qqnorm(log(UNA$TP)); qqline(log(UNA$TP), col = 2)
Normal Q−Q Plot Normal Q−Q Plot
−1.0
0.30
−1.5
0.25
−2.0
Sample Quantiles
Sample Quantiles
0.20
−2.5
0.15
−3.0
0.10
−3.5
−4.0
0.05
−4.5
0.00
−2 −1 0 1 2 −2 −1 0 1 2
Theoretical Quantiles Theoretical Quantiles
(a) (b)
Fig. 5. A Q-Q plot of total phosphorus concentrations at Station 1 versus the standard normal
distribution
Fig. 5 (a) indicates that the distribution of total phosphorus concentrations is skewed to the
right. Fig. 5 (b) shows an S-shape, but there is not sufficient evidence to prove that the
distribution of total phosphorus on the natural log scale is non-normal. Fig. 6 is a Q-Q plot
comparing whether two sample datasets are from populations with a common distribution.
Note that there are also a few outliers appearing (possible outliers are in red). Otherwise, the
plot suggests that the two samples have the same distribution.
> qq <-qqplot(UNA$TP, CKL$TP, plot.it = TRUE, xlab = “Concentrations of total phosphorus
at Station 1”, ylab =“Concentrations of total phosphorus at Station 2”)
> points(qq$x[85],qq$y[85], pch=16, col=2)
3. Water quality index

Sometimes it is difficult to assess water quality from a large number of water quality
parameters. Traditional methods to evaluate water quality are based on the comparison
of experimentally determined parameter values with an existing local normative, which
does not provide a global summary on the spatial and temporal trends in the overall water
quality (Debels et al., 2005; Kannel et al., 2007). To integrate complex water quality data and
provide a simple and understandable tool for informing managers and decision-makers about
the overall water quality status, various water quality indices (WQI) have been developed,
which can be used to give a global vision on the spatial and temporal changes of the water
quality. An early water quality index was proposed by Horton (1965), and then developed
by Brown et al. (1970), Dojlido et al. (1994), McClelland (1974), and Pesce & Wunderlin (2000).
148
0.5
Concentrations of total phosphorus at Station 2
0.4
0.3
0.2
0.1
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30
Concentrations of total phosphorus at Station 1
Fig. 6. A Q-Q plot comparing distributions of total phosphorus at Station 1 and Station 2
Water quality indices have been employed frequently in the public domain to assess water
quality, such as the US National Sanitation Foundation Water Quality Index (Brown et al.,
1970), the Canadian Water Quality index (CCME, 2001), the British Columbia Water Quality
Index (Zandbergen & Hall, 1998) and the Oregon Water Quality Index (Cude, 2001). Main
steps to derive a water quality index are as follows: select the most important water
quality parameters (such as dissolved oxygen, total phosphorus, temperature); transform
the parameters into a common scale; assign parameter weights; and aggregate scores to
a single score. In this section, various water quality indices, such as those based on the
weighted/unweighted arithmetic/geometric/harmonic mean functions, will be presented
and compared. Their uses and limitations will be also discussed.
3.1 Weighted water quality indices

Water quality indices are usually obtained by assigning a suitable weight to each water quality
parameter index and averaging them using some type of average functions. In this subsection,
we consider three different weighted water quality indices. The water quality index proposed
by Pesce & Wunderlin (2000) is:
n
WQISA = k ∑ ω i Si , (1)
i =1
where n is the number of the water quality parameters, Si is the score of the ith parameter,
and ω i is the relative weight given to Si satisfying ∑ ni=1 ω i = 1. k is a subjective constant
representing the visual impression of river contamination. The value of k ranges from 0.25
(for highly contaminated water) to 1 (for water without contamination). WQISA tends to
overestimate the pollution due to the use of a subjective constant, which is not correlated with
the measured parameters (Kannel et al., 2007).
Statistical Tools
Analyzing Analyzing
Let k = 1 in Equation (1). In general, we have the objective water quality index originally
proposed by Horton (1965), hereafter called the weighted arithmetic water quality index. It
has been used by many researchers (Brown et al., 1970; Prati et al., 1971; Sanchez et al., 2007):
n
WQIWA = ∑ ω i Si . (2)
i =1
The third water quality index is based on the weighted geometric mean function (Brown et al.,
1970; McClelland, 1974), which is always smaller than WQIWA if all values of Si are positive:
n
WQIW G = ∏ Siω . i
(3)
i =1
The above weighted water quality indices indicate that each water quality parameter may
have different weights based on the importance of the water quality situation. This
characteristic could be desirable when water quality indices are specific to the protection
of aquatic life. However, when sensitivity to changes in each water quality parameter
is more desirable than sensitivity to the most heavily weighted water quality parameter,
such weighting could be unnecessary (Cude, 2001; Gupta et al., 2003; Landwehr & Deininger,
1976). Some unweighted water quality indices were therefore explored (Cude, 2001;
Dojlido et al., 1994; Landwehr & Deininger, 1976) and are now introduced in the following
subsection.
3.2 Unweighted water quality indices

In this subsection, we introduce three unweighted water quality indices. The first two are
arithmetic/geometric water quality indices proposed by Landwehr & Deininger (1976),
n
WQI A = 1/n ∑ Si , (4)
i =1
n
WQIG = ( ∏ Si )1/n , (5)
i =1
which is a special case of (2) and (3) with ω i = 1/n for any i, respectively. As with the
relationship between WQIW G and WQIWA, WQIG is always lower than WQI A . The third is
the harmonic square water quality index,

n
WQIH = , (6)
∑i=1 S12
n
i
which has been suggested as an improvement over both WQIWA and WQIW G (Cude, 2001;
Dojlido et al., 1994). Compared to WQIWA and WQIW G, WQIH is the most sensitive to
changes in single water quality parameter (Cude, 2001).
3.3 Harkins’ water quality index

An objective water quality index was proposed by Harkins (1974), which is based on Kendall’s
nonparametric multivariate ranking procedure.
150
n
( Ri − Ric )2
WQIHR = ∑ var ( Ri )
, (7)
i =1
where
ki
1
var ( Ri ) = [( M3 − M ) − ∑ (t3ij − tij )],
12M j =1
Ri and Ric correspond to the rank and control values of the ith water quality parameter,
respectively. M is the number of water quality parameters plus the number of control
values, tij is the number of elements involved in the jth tie encountered when ordering
the measured values of the ith water quality parameter, and k i is the total number of ties
encountered in ranking the measurements of the ith parameter. Landwehr & Deininger
(1976) and Gupta et al. (2003) compared WQIHR with water quality indices WQIWA, WQIW G ,
WQIG , and WQI A . Their results indicated that these five indices are correlated well with the
opinions of experts, and although the five indices showed significant correlation with each
other, WQIHR was the lowest of the five. Therefore, they suggested adopting any of the four
indices except WQIHR.
4. Methods for handling data below detection limits

One feature of water quality measurement is that some data will fall above or below the
detection limit, and therefore not be captured, because of limitations of the measurement
procedures or the analytical tools used in the laboratories. Data below a detection limit are
also referred as left-censored data. There could also be multiple detection limits involved
if an instrument is upgraded during the project period or data are combined from multiple
laboratories. Even data below the detection limits are still of considerable importance because
of the potential health hazard. The data below the detection limits complicate the analysis of
the water quality data. Various strategies have been developed to analyze the data that fall
below detection limits (Fu & Wang, 2011; Helsel, 1990; Shumway et al., 2002). In the following
subsections, simple substitution methods, parametric methods, and nonparametric methods
will be introduced.
4.1 Simple deletion/substitution methods

Simple deletion/substitution methods delete/replace the measurements
√ below detection
limits (DL) with fixed values, such as zero, 1/2DL, 1/ 2DL or √ DL (Helsel, 1990;
Hornung & Reed, 1990). Hornung & Reed (1990) proposed using 1/ 2DL when the data
are not highly skewed
√ and 1/2DL substitution otherwise. Hewett & Ganser (2007) found
that 1/2DL and 1/ 2DL perform well when the sample size is less than 20 and the percent
censored is less than 45 percent. It is easy and convenient to use the substitution methods.
However, all tend to be biased and cause a loss of information (El-Shaarawi & Esterby, 1992;
Helsel & Cohn, 1988; Lubin et al., 2004). When the results strongly depend on the values being
substituted, particularly for data with multiple detection limits (Shumway et al., 2002), the
substitution methods are not generally suitable. In particular, when there is a high proportion
of data below detection limits, results for standard errors are also far less desirable, and the
biased standard errors may further distort the inference (Helsel, 1990; 1992; Shumway et al.,
2002).
Statistical Tools
Analyzing Analyzing
4.2 Parametric methods

Assume that the distribution of measurements is known, such as normal or lognormal.
The data below the detection limits can be filled using values randomly selected from the
distribution or replaced with their conditional expected values (conditional on being less
than the detection limits) (Helsel, 1990). Suppose that there are n detected measurements
(y1 , . . . , yn ) and m measurements below the detection limits (c1 , . . . , cm ). The likelihood
function is
n m
L (θ ) = ∏ f θ (yi ) ∏ Fθ (c j ), (8)
i =1 j =1
where θ is a vector of parameter, f (·) is the probability density function of y, and F (·) is the
cumulative density function (c.d.f.) of y. The parameter estimates of θ and summary statistics
can be obtained by the maximum likelihood method (ML) (Cohen, 1976; Cohn, 1988; Helsel,
1992). Results based on a lognormal distribution assumption by the maximum likelihood
method can be easily obtained using statistic software R (NADA package) (Lee & Helsel,
2005). If the distributional assumption is appropriate and the sample size is large, the
maximum likelihood method is the most efficient (Cohn, 1988; Helsel, 1992; Hewett & Ganser,
2007). To incorporate the covariate effects when analyzing the water quality data that fall
below detection limits, the following regression models can be considered.
4.2.1 Tobit regression

Tobit regression model (Tobin, 1958) has been widely used to analyze censored data. The
model can be written as
log(y∗i ) = β0 + β1 xi + i , (9)
where y∗ is a latent variable and yi = y∗i if y∗i > ci and yi = ci otherwise. Random error term
i follows a normal distribution N (0, σ2 ). The likelihood function (8) can be written as
δi 1−δi
1 log(yi ) − β0 − β1 xi log(ci ) − β0 − β1 xi
L ( β0 , β1 ) = ∏ σ
φ
σ ∏ Φ
σ
,
i i
where δi = 1 if y∗i
> ci and δi = 0 otherwise. The maximum likelihood estimates (MLE) of
parameters can be obtained from the function survreg (survival package) and vglm (VGAM
package) in R if the detection limit is a single number. An example to obtain MLE of
parameters in a Tobit regression model log(DP) = β0 + β1 log(Flow) + is given,
> library(survival)
> fit<- survreg(Surv(log(DP), DP>=0.01, type = ‘left’) ∼ log(Flow), data = UNA, dist =
‘gaussian’)
> summary(fit)
For multiple detection limits, the estimates can be derived by a Newton-Raphson algorithm.
The Wald type test or the likelihood ratio test can be applied to test the group difference
or covariate effects (by testing β = 0). Tobit regression is also applicable when both the
measurements of the response and covariate variable are with detection limits (Helsel, 1992).
When the distribution is known and the error terms are homeostatic, the estimate derived by
the maximum likelihood method is optimal (Helsel, 2005b).
152
4.2.2 Logistic regression

Let ỹ = 1 if the response is above a detection limit c and ỹ = 0 otherwise. Assume that the
probability of ỹ = 1 is p, then p = p(y > c). A binary logistic regression modeled as a linear
function of covariate x is given by
p
log( ) = α0 + α1 x.
1− p
The likelihood function is
(1 − pi )1−ỹi ,
ỹ i
L ( α0 , α1 ) = ∏ pi
i
where pi = exp(α0 + α1 xi )/[1 + exp(α0 + α1 xi )]. The maximum likelihood estimates of

parameters can be obtained from glm function in R. The significance of the covariate effect can
be tested using the likelihood ratio statistic (Helsel, 1992). For multiple detection limits, the
ordered logistic regression can be used. More details can be seen in Helsel (1992). An example
to obtain parameter estimates from a logistic regression log( p/(1 − p)) = α0 + α1 log(Flow) is
given as follows, where p = p(DP > 0.01).
> UNA$ DPd<- 1-(UNA$ DPrem ==“<”)
> logitfit<- glm(DPd ∼ log(Flow), data = UNA, family = binomial(“logit”))
> summary(logitfit)
Parametric methods generally perform well for summary statistics when the dataset is large
and the underlying distribution can be approximated by a known distribution. Specification
of the underlying distribution of a dataset may be difficult in practical problems. The ML
method does not work well when the distributional assumption is invalid or the sample size
is small (<20) (Gleit, 1985; Helsel, 2005b; Helsel & Cohn, 1988). Furthermore, the ML method
is sensitive to outliers, which usually exist in water quality data. An implementation of fully
parametric methods is a robust and efficient semi-parametric regression method on order
statistics (ROS) and will be introduced in the following subsection.
4.2.3 ROS method

The ROS method was provided by Helsel & Cohn (1988), which is a simple imputation
method that fills in data below detection limits based on a probability plot of detections
(Helsel & Cohn, 1988; Lee & Helsel, 2005; Shumway et al., 2002). It can be used to obtain
summary statistics, plot modeled distributions, and predict values based on the modeled
distributions (Fig. 7). The ROS method has been evaluated as one of the most
reliable approaches for estimating summary statistics of data with multiple detection
limits (Shumway et al., 2002). Lee & Helsel (2005) developed software implementation that
performs the ROS method, and it is a part of the NADA library in statistical software R. R
codes for Fig. 7 are as follows:
> library(NADA)
> UNA<- UNA[!is.na(UNA$OP), ]
> UNA$CenOP<- UNA$OPrem == “<”
> rosop<- cenros(UNA$OP, UNA$CenOP, forwardT =“log”, reverseT = “exp”)
> plot(rosop, plot.censored = TRUE)
> lines(rosop, col = 2)
Statistical Tools
Analyzing Analyzing
11
Percent Chance of Exceedance
95
90
75
50
25
10
5
0.100
0.050
0.020
Value
0.010
0.005
0.002
−2 −1 0 1 2
Normal Quantiles
Fig. 7. A normal Q-Q plot for a ROS model. Solid circles are detected data. Open circles are
modeled undetected values.
4.3 Nonparametric methods

Parametric and semi-parametric methods are based on the assumption of the underlying
distribution of the data. Nonparametric methods provide an alternative that does not require
specifying a distribution and filling in the data below detection limits. The nonparametric
methods are generally used to analyze the right censored data. Left censored data can
be converted into right-censored data by flipping the data over the largest observed value.
Lee & Helsel (2007) provided software tools for direct analysis of data with multiple detection
limits (left-censored data) by nonparametric modeling and hypothesis testing.
4.3.1 Kaplan-Meier
The Kaplan-Meier (K-M) method is the standard method for computing descriptive statistics
of data that fall below detection limits (Helsel, 2005; Lee & Helsel, 2007). K-M method
has been widely used in survival analysis, where it is employed with right-censored
time-to-failure data. The K-M method can estimate the percentiles or c.d.f. for a dataset,
and can test hypotheses. It can describe and compare the shapes of different datasets (Figs. 8
(a) and (b)).
> KM<- cenfit(UNA$OP, UNA$CenOP)
> plot(KM)
> dat2<- dat2[!is.na(dat2$OP), ]
> dat2$CenOP<- dat2$OPrem == “<”
> g2<- cenfit(dat2$OP, dat2$CenOP,dat2$Station)
> plot(g2,lty = c(1 : 3), col=c(1, 2, 4))
> legend(0.002, 0.8, c(“Station 1”,“Station 2”,“Station 4”), lty = c(1:3), col=c(1, 2, 4))
154
1.0
1.0
0.8
0.8
Station 1
Station 2
Station 4
0.6
0.6
Probability
Probability
0.4
0.4
0.2
0.2
0.0
0.0
0.002 0.005 0.010 0.020 0.050 0.100
0.002 0.005 0.010 0.020 0.050 0.100
Value Value
(a) Dashed lines are 95% confidence (b) Empirical c.d.f at three stations
limits
Fig. 8. Empirical cumulative distribution functions for datasets with multiple detection limits
Zhang et al. (2009) developed a nonparametric estimation procedure, and under a fixed
detection limit and some mild conditions, they established the theoretical equivalence of three
nonparametric test statistics: the Wilcoxon rank sum, the Gehan, and the Peto-Peto tests.
Their simulation studies indicated that nonparametric methods work well for a range of small
sizes and censoring rates (Zhang et al., 2009). For hypothesis testing with multiple detection
limits, one robust method is to censor all data at the highest detection limit and then perform
an appropriate nonparametric test (Helsel, 1992). This can result in a loss of information,
however, the accelerated failure time (AFT) model can integrate the Gehan and logrank tests,
incorporate covariate effects, and compare the differences between two/multiple data groups
with multiple detection limits (Jin et al., 2006; Wei, 1992; Zhang et al., 2009).
4.3.2 AFT model

Assume that {Yi , i = 1, . . . , N }, {Ci , i = 1, . . . , N } and { Xi , i = 1, . . . , N } are measurements,
detection limits and p × 1 covariate vector, respectively. Let Δ i = 1 if Yi is below the detection
limit Ci and Δ i = 0 otherwise. Let Z̃i = min{− log(Yi ), − log(Ci )}; therefore ( Z̃i , Δ i , Xi ) are
the observations. The accelerated failure time model is
Zi = Xi β + i ,
where Zi = log(Yi ), β is an unknown regression parameter vector, and i is the error

term. Suppose that {i , i = 1, . . . , N } are independent and identically distributed and their
underlying distribution is unknown.
Estimation and inference of the regression parameters are based on the estimating functions
given by

j =1 X j I ( e i ≥ e j )
∑N
N
−2
U ( β ) = N ∑ Δ i ω ( e i ) Xi − ,
i =1 ∑Nj =1 I ( e i ≥ e j )
where ω (ei ) is a weight function and ei = log(Yi ) − Xi β t , where β t is the true value of β.
Let ω (ei ) = 1 and ω (ei ) = ∑ Nj =1 I ( e i ≥ e j ); U ( β) correspond to the log-rank and Gehan
Statistical Tools
Analyzing Analyzing
13
statistics, respectively. The estimating functions U ( β) are step functions and discontinuous
in the regression parameters, which makes it difficult to find consistent estimators and their
asymptotic variance and covariance matrices. Much progress has been made to overcome
these difficulties (Brown & Wang, 2006; Heller, 2007; Jin et al., 2003; Lee et al., 1993), and the
function lss (lss package) can be used to obtain various statistics from an AFT model.
> library(lss)
> UNA$status<- 1-(UNA$OPrem==“<”)
> aftfit<- lss(cbind(log(OP), status)∼ log(Flow), data=UNA, gehanonly=FALSE, cov=TRUE)
> print(aftfit)
Jin et al. (2006b) extended marginal accelerated failure time models to multivariate censored
data. Their method, which is based on an independence “working" model, may ignore
the within-site correlations in obtaining parameter estimates, while taking account of
the correlation in calculating the standard errors. More efficient estimators with similar
computational complexity were developed for multivariate censored data analysis, when
measurements from the same site exhibit strong temporal correlations (Fu & Wang, 2011).
5. Trend detection
In recent years, concentrations of various water quality parameters have been collected.
Tests for trends specific to various water quality parameters have been of keen interest in
environmental science (Helsel, 1992). A number of methods have been proposed to detect
and assess changes in water quality. In this section, a variety of approaches will be introduced
and their strengths and weaknesses investigated. The exogenous variable effects and serial
dependence will be considered when testing water quality trends.
5.1 Parametric methods

Under the normality of residuals and constant variance assumptions, simple/multiple linear
regressions are preferable for detecting trends of water quality.
5.1.1 Simple linear regression

Let Y be the random variable of interest in a trend test, such as concentrations of water quality
parameters. T denotes time (often expressed in years). If Y is linear over time T, the linear
simple regression of Y on T is a test for trend.
Y = β0 + β1 T + , (10)
where β1 is the rate of change in Y. The null hypothesis for testing the trend of y can be stated
as a test for β1 = 0. The Wald type statistic (t-statistic) can be used. If the null hypothesis
is rejected, it indicates that there is a linear trend in Y over time. If Y is not linear over time
T, some transformation of Y, such as a log transformation, may be necessary. An example
using a linear regression to detect the trend of total phosphorus concentrations at Station 1 is
presented in Fig. 9. The results indicate that the trend of total phosphorus is not significant.
5.1.2 Multiple regression

Most concentrations of water quality parameters have strong seasonal patterns (see Fig.
10). They are influenced by the changes in biological activity, both natural and managed
156
0.30
0.25
0.20
0.15
0.10
0.05
0.00
2006 2007 2008 2009 2010 2011
Time
Fig. 9. Linear regression trend line for total phosphorus concentrations

Regression: C = 0.09 − 0.007Time, and p = 0.14
activities such as agriculture (Helsel, 1992; Hirsch et al., 1991). Therefore, it is important
to consider seasonal effects when evaluating changes in water quality data. In parametric
procedures, multiple regression with periodic functions can be used to describe seasonal
variation. Consider the following simple case,
Y = β0 + β1 T + β2 cos(2πT ) + β3 sin(2πT ) + , (11)
where T is expressed in years and β1 indicates the change rate of Y. Terms sin(2πT ) and
cos(2πT ) capture the annual cycle and account for seasonality. Residuals must follow a
normal distribution (or approximately normal). The trend test can be constructed by testing
β1 = 0.
If residuals still show a seasonal pattern (see Fig. 11), additional periodic functions should be
included in model (11) to remove the seasonal variation. A general multiple linear regression
is given by
K
Y = β0 + β1 T + ∑ [ β2k+1 cos(2πkT ) + β2k+2 sin(2πkT )] + . (12)
k =0
The cases of K = 0 and K = 1 correspond to model (10) and (11). If K = 2, a period of 1/2 year
is then also included in model (12). Fig. 11 shows that the residuals of the linear regression in
Subsection 5.1.1 represent a seasonal pattern, therefore periodic functions should be included
in the model.
When Y or some transformation of Y is linear with time T, and residuals follow a normal
distribution with a constant variance, the parametric regression is optimal. However, the
Statistical Tools
Analyzing Analyzing
15
0.30
0.25
0.20
0.15
0.10
0.05
0.00
2006 2007 2008 2009 2010 2011
Time in years
Fig. 10. Time series plot of total phosphorus concentration at Station 1

0.25
0.20
0.15
0.10
Residual
0.05
0.00
−0.05
2006 2007 2008 2009 2010 2011
Time
Fig. 11. Residuals of the linear regression mentioned in above subsection versus times in year.
distribution of water quality data is usually highly skewed, in particular, data related to
discharge, as well as biological indicators (biomass, chlorophyll) (Helsel, 1992; Hirsch & Slack,
158
1984). The test that depends on the normality assumption may be inappropriate. The
following subsection introduces several nonparametric methods that do not require the
normality assumption.
5.2 Nonparametric methods

Water quality data usually have the following characteristics: nonnormal data, missing values,
values below detection limits, and serial dependence. The nonparametric methods are robust
and can handle these problems easily.
5.2.1 Mann-Kendall test

Mann (1945) and Kendall (1975) proposed a nonparametric test for randomness against trend.
According to Mann (1945), the null hypothesis H0 states that ( x1 , . . . , xn ) are a sample of n
independent and identically distributed random variables. The alternative hypothesis H1 of a
two-sided test is that the distributions of xk and x j are not identical for all k, j ≤ n, and k = j.
The test statistic S is defined as
n −1 n
S= ∑ ∑ sgn( x j − xk ),
k =1 j = k +1
where
⎧
⎨ 1 if θ > 0
sgn(θ ) = 0 if θ = 0
⎩
−1 if θ < 0.
Under the null hypothesis, Mann (1945) and Kendall (1975) obtained the mean and variance
of S.
E (S ) = 0,
var(S ) = [ n (n − 1)(2n − 5) − ∑ t(t − 1)(2t − 5)] /18,

t
where t is the extent of any given tie (number of xs involved in a given tie) and ∑t denotes the
summation over all ties.
Both Mann (1945) and Kendall (1975) derived the exact distribution of S for n ≤ 10; proved
that the distribution of S is normal as n → ∞; and further showed that even for n = 10, the
normal approximate is excellent if one calculates the standard normal variate Z by
⎧ S −1
⎨ {var( S)}1/2 if S > 0
⎪
Z= 0 if S = 0
⎪
⎩ S +1
{var( S)} 1/2 if S < 0.
Hence, in a two-sided test for trend, the H0 should be rejected if | Z | ≥ zα/2 , where Φ (zα/2 ) =
1 − α/2, Φ (·) is the standard normal c.d.f. and α is the significance level for the test. A position
value of S indicates an “upward” trend, and a negative value of S presents a “downward”
trend. For an example from Station 21 at Susquehanna River basin (Fig. 10), the statistic
Statistical Tools
Analyzing Analyzing
17
S = −90, the var(S ) = 1096.67 under the null hypothesis and the p value is 0.0072, which
indicates a downward trend in the concentration of total phosphorus at Station 21.
> library(Kendall)
> TP<- ts(CONY$TP, frequency=1, start=1990)
> mk<- MannKendall(TP)
> summary(mk)
The seasonality is a common phenomenon, which indicates that the distributions differ
for different times of year. The Mann-Kendall test therefore is sensitive to seasonality.
Hirsch et al. (1982) developed a modified Mann-Kendall test to detect the trend of data with
seasonality.
5.2.2 The seasonal Kendall test

Hirsch et al. (1982) presented a modified Mann-Kendall test that detects trends in time series
with seasonal variation and called as a seasonal Kendall test. Let X = ( X1 , X2 , · · · , Xm ) and
Xi = ( xi1 , xi2 , · · · , xini ), where X is the entire sample consisting of m subsamples Xi , and m
is the number of seasons. Each subsample Xi contains n i annual values. The null hypothesis
H0 is that X is a sample of independent random variables ( xij ), and that Xi is a subsample of
independent and identically distributed random variables for i = 1, · · · , m. The alternative
hypothesis H1 is that the subsample is not distributed identically. The test statistic proposed
by Hirsch et al. (1982) is given as follows. For month i,
n i −1 ni
Si = ∑ ∑ sgn( xij − xik ). (13)
k =1 j = k +1
Under the null hypothesis, Si is a Mann-Kendall test statistic and
E (Si ) = 0,
var(Si ) = [ n i (n i − 1)(2n i − 5) − ∑ ti (ti − 1)(2ti − 5)] /18.

ti
The distribution of Si is normal as n i → ∞ (ti is the extension of a given tie in month i). Define
the seasonal Kendall statistic
m
S∗ = ∑ Si , (14)
i =1
and its expectation

m
E (S ∗ ) = ∑ E(Si ) = 0,
i =1
and variance
m m m
var(S ∗ ) = ∑ var(Si ) + ∑ ∑ cov(Si , S j ). (15)
i =1 i =1 j = i
160
Under the null hypothesis, Si and S j ( j = i ) are independent, therefore

m
var(S ∗ ) = ∑ [ni (ni − 1)(2ni − 5) − ∑ ti (ti − 1)(2ti − 5)]/18.
i =1 ti
The standard normal variate Z∗is defined as

⎧ S ∗ −1 ∗
⎨ {var( S∗ )}1/2 if S > 0
⎪
∗
Z = 0 if S ∗ = 0
⎪
⎩ S ∗ +1
{var( S∗ )}1/2
if S ∗ < 0.
The approximation is adequate for n i = 3 and m = 12 for all i (Hirsch et al., 1982). For the
example from Station 21 at Susquehanna River basin (Fig. 10), the statistic S ∗ = −360, the
var (S ) = 10779.67 under the null hypothesis, and the p value is 0.0005, which indicates a
downward trend in the concentration of total phosphorus at Station 21.
> library(Kendall)
> TPS<- ts(c(t(RCON[,-1])), frequency = 12, start = c(1990, 1))
> smk<- SeasonalMannKendall(TPS)
> summary(smk)
A limitation of the seasonal Kendall test is one observation per month. If there are multiple
observations in each of the months, Hirsch et al. (1982) suggested using the medians of
the multiple observations in the seasonal Kendall test. Another limitation is that the
seasonal Kendall test is not robust against serial dependence. When serial dependence exists,
cov(Si , S j ) in Equation (15) does not equal zero. Hirsch & Slack (1984) provided a modification
of the seasonal Kendall test which is robust against serial dependence, except when the data
have very strong long-term persistence or when the sample sizes are small. More details can
be found in Hirsch & Slack (1984) and Letternmatier (1988). In addition to detecting the trend,
the magnitude of such a trend may also be desirable. In model (10), an estimate of β1 can be
used to estimate the trend. For a seasonal Kendall test, calculate dijk = ( Xij − Xik )/( j − k) for
all pairs ( Xik , Xij ) and (k < j). Hirsch et al. (1982) proposed using the median of dijk as an
estimator of the slope, which is robust against extreme values.
5.2.3 Sen’s T test

Farrel (1980) proposed an aligned-rank test for detecting trends, which is distribution free
and not affected by seasonal fluctuations (Van Belle & Hughes, 1984; Yu et al., 1993). Let xij
be the measurement in the ith month of the jth year at a sampling station, and i = 1, . . . m,
j = 1, . . . , n. Let Rij be the rank of ( xij − xi+ ) among the mn values of differences, where
xi+ = ∑nj=1 xij . If ties occur, the average of the ranks is taken as the rank of each tie. The
statistic is
1/2
12m2 n
n+1 nm + 1
T=
n (n + 1) ∑i,j ( Rij − Ri+ )2
∑ ( i −
2
)( R +j −
2
) ,
i =1
where Ri+ = ∑ j Rij /n and R+ j = ∑i Rij /m. Under the null hypothesis of no trend, the
distribution of T tends to the standard normal distribution. Simulation results indicated
that the normal approximation for the statistic T was reasonable even for a small sample
(Van Belle & Hughes, 1984).
Statistical Tools
Analyzing Analyzing
19
The three nonparametric methods for detecting trends mentioned above have practically the
same power at a statistical significance level of 0.05 (Yu et al., 1993). It is worth noting that
there may exist water quality parameters which exhibit strong evidence of a download trend
in some months and then exhibit strong evidence of an upload trend (step trend) (Helsel, 1992;
Hirsch et al., 1991). The methods described above all assume a single trend across all seasons,
provide a summary statistic for the entire record (monotonic trend), and do not indicate
when there are trends in opposing directions in different months. Van Belle & Hughes (1984)
developed a statistic for testing homogeneity of trends. The statistic is
m
χ2homogeneous = χ2total − χ2trend = ∑ Zi2 − mZ̄2 ,
i =1
where Zi = Si / var(Si ), Si is the Mann-Kendall statistic in Equation (13), and Z̄ =

∑mi =1 Zi /m. Under the null hypothesis that the seasons are homogeneous with respect to
trend, χ2homogeneous approximates the chi-square distribution with m − 1 degree of freedom.
If χ2homogeneous exceeds the critical value, it indicates that there are different trends among
different seasons. In that case, the three nonparametric methods are not meaningful, and
the Mann-Kendall statistic can be used to test the trend for each individual season.
5.3 Adjusting covariate effects on trend tests

Several variables (X) other than time trend usually have considerable influence on water
quality parameters (Y) (see Fig. 12). These variables are natural and random phenomena such
as rainfall, temperature, and stream flow. To detect the trend of water quality parameters with
time (T), these variable effects on water quality parameters need to be removed. The removal
process includes modeling and explaining variable effects with regression methods and the
LOWESS method (Helsel, 1992).
> xyplot(log(TP)∼ log(Flow)|Year, col.line = 2, type=c("p", "r"), data = UNA, xlab = "Log
values of flow", ylab = "Log values of total phosphorus concentrations")
5.3.1 Parametric methods

Consider a linear regression of Y versus time T and one or more covariates X,
Y = β0 + β1 T + β2 X + .
For the trend test, the null hypothesis is β1 = 0. The t-statistic can be used for the trend test.
This model simultaneously explains the covariate effect and detects the trend with time. If the
covariate changes with time, the following regression can be considered.
Y = β0 + β1 T + β2 X + β3 T ∗ X + ,
where T ∗ X is the interactive term. For regression models, the relationship (linear function)
between Y and X must be checked. Residuals should have no outliers and a constant
covariance. The functions in R for testing these assumptions can be found in Subsection 5.4.
According to the previous analysis, we use the following multiple linear regression to detect
the trend of total phosphorus at Station 1. This model adjusts the flow effect and also captures
the annual cycle.
ln (C ) = 2.25 − 0.74Time − 0.43 sin(2πTime) + 0.22 cos(2πTime)
−1.81ln (Flow) + 0.096Time ∗ ln (Flow) + 0.15ln (Flow)2
162
5 6 7 8 9
2008 2009 2010

−1
−2
Log values of total phosphorus concentrations

−3
−4
2005 2006 2007

−1
−2
−3
−4
5 6 7 8 9 5 6 7 8 9
Log values of flow
Fig. 12. Log values of total phosphorus concentration (C) versus log values of flow (F)
The results indicate an annual cycle and flow effect exist (p<0.05). After adjusting exogenous
variables effects, the concentration of total phosphorus significantly decreases (p = 0.02).
5.3.2 Semi/nonparametric methods

Hirsch & Slack (1984) provided an adjusted seasonal Kendall test and proposed the following
mixture procedure to test trends: (a) Use a regression model Y = f ( βX ) + to find the
relationship between the concentration and covariates, where f (·) is a certain function of
covariate X; (b) If there exists a significant relationship, compute the adjusted concentration
Yik − Ŷik , where Ŷik = f ( β̂X ) is the estimated concentration of Yik ; (c) Then apply the seasonal
Kendall test for trend and slope estimator to the time series of Yik − Ŷik .
The nonparametric LOWESS method (Cleveland, 1979; Helsel, 1992) can be used to remove a
covariate effect without previously assuming the form of the relationship between Y and X.
It is solely determined by the dataset and therefore it is robust to the distribution of the data
pattern. The function lowess in the statistical software R can be used to obtain the fitted values
Ŷ of Y. The seasonal Kendall statistic (14) is calculated from Y − Ŷ and T data pairs.
The parametric regression method can simultaneously check the covariate effect and detect
the trend. When the linearity and normality assumptions are met, the parametric regression
method outperforms for detecting and estimating the magnitude of trends. Otherwise, the
LOWESS method is a desirable alternative. To examine the trend of a water quality parameter,
the covariate and seasonal effects need to be removed. According to the real datasets, choose
a reasonable statistical approach to test for trends. Various methods for trend tests are given
in Table 1. For water quality data with detection limits, parametric Tobit regression (9) can be
used. When a fixed detection limit exists, all the data below the fixed detection limit can be
Statistical Tools
Analyzing Analyzing
21
considered to be tied with each other. The nonparametric procedures such as Mann-Kendall,
and the seasonal Kendall statistics can be used directly. If multiple detection limits exist,
censor the data at the highest detection limit and then use an appropriate method to test the
trend. Some information is certainly lost by making this change.
No exogenous covariate (X) effects
No seasonality Seasonality
Parametric Regression of Y on T Regression of Y on T and Seasonal terms
Nonparametric M-K test S-K test
Mixed S-K test on residuals from regression of Y on X
Exogenous covariate (X) effects exist
No seasonality Seasonality
Parametric Regression of Y on ( T, X ) Regression of Y on ( T, X, S )
Nonparametric M-K of residuals from regression of Y on ( T, X ) S-K of residuals from lowess of Y on X
Mixed M-K of residuals from regression of Y on ( T, X ) S-K of residuals from regression of Y on ( T, X, S )
Table 1. Classification of various types of tests for monotonic trend. M-K indicates
Mann-Kendall test, S-K indicates Seasonal Kendall test, and S denotes seasonal terms.
5.4 Computational implementation for linear regression models using R

In this subsection, we will show how to use the statistical software R to fit, evaluate and
modify a linear regression model. More details can be seen in Adler (2009), Crawley (2007),
and Venables & Ripley (2002).
A linear regression model is one of the most classic and popular methods in statistical practice.
It is a very important tool for the statistical analysis of water quality data. It assumes that
there is a linear relationship between a response variable (continuous) and some covariate
variables. To fit a linear regression model to a dataset, the primary function is lm. We begin
with the dataset mentioned in Section 2 to show how to fit a linear model in R. R codes for
fitting the dataset are as follows.
> con.lm <- lm(log(TP) ∼ log(Flow) + pH, data = UNA)
To print a simple display of the fitted information, use the print function:
> print(con.lm)
To obtain conventional regression analysis results, use the summary function:
> summary(con.lm)
To extract the regression coefficients, use the coef or coefficients function:
> coef(con.lm)
To obtain the variance-covariance matrix for the model fitted above, use the vcov or Var
function:
> vcov(con.lm)
To calculate the confidence intervals for the coefficients in the fitted model, use the confint
function:
>confint(con.lm, level = 0.95)
To get the residuals, use the resid or residuals function:
> resid(con.lm)
To obtain the fitted values, use the fitted or fitted.values function:
> fitted(con.lm)
To return the deviance of the fitted model, use the deviance function:
> deviance(con.lm)
To refit the model, it is better to use the update function, which can save some typing. For
164
example, a slightly different model is used to fit the data above, which considers an extra
covariate “Temp” besides “Flow” and “pH”.
> con.lm2<-update(con.lm, . ∼ . + Temp)
To compare models con.lm and con.lm2 which are used to fit the same dataset, use the anova
function:
> anova(con.lm, con.lm2)
The main arguments to the function lm are
> lm(formula, data, weights, subset, na.action),
where formula is the model formula that specifies the form of the model to fit; data is an
optional data frame containing the variables in the model; weights is a positive numeric vector
containing weights to be used in the fitting process; subset is an optional vector specifying
a subset of observations to be used in the fitting process; and na.action is a function which
indicates how to handle missing values contained in the data.
The least-squares method performs well when the following key assumptions are satisfied: (1)
There is a linear relationship between any pair of covariate variables (linearity); (2) The error
terms are normally distributed (normality) with a constant variance (homoscedasticity); (3)
The error terms are not correlated with each other (autocorrelation). However, because these
assumptions may not be met in water quality data, linear regression is therefore not always
appropriate. The test functions can be used to check these assumptions in R. The function
ncv.test in the car package can be used to test the homoscedasticity. The function durbin.watson
(car package) is used to test autocorrelation in a linear regression model. Diagnostic plots can
also provide checks for homoscedasticity, normality, and influential observations (see Fig. 13),
which can be obtained using the function plot(con.lm).
Residuals vs Fitted Scale−Location

19
19
2
Standardized residuals
1.5
51
54
51
1
Residuals
1.0
0
0.5
−1
54
0.0
−4.0 −3.5 −3.0 −2.5 −2.0 −4.0 −3.5 −3.0 −2.5 −2.0
Fitted values Fitted values
Normal Q−Q Residuals vs Leverage

0.5
19
3
3
51 51
2
2
40
1
1
0
0
−1
−1
−2
38
Cook’s distance
−2
54
−2 −1 0 1 2 0.00 0.02 0.04 0.06 0.08 0.10
Theoretical Quantiles Leverage
Fig. 13. Diagnostic plots for a linear regression model

Statistical Tools
Analyzing Analyzing
23
6. Conclusions
Statistical methods are important in water quality analysis because much of what is known
about water quality comes from numerical datasets. In this chapter, various statistical
methods for analyzing water quality data have been introduced. Three typical graphs,
boxplots, Q-Q plots, and scatter plots, which contain appropriate summarized information
about datasets, are used to provide insight for analysts into datasets. A variety of classic
water quality indices are applied to give a global assessment of water quality. Weighted water
quality indices are relatively subjective; unweighted water quality indices and Harkins’ water
quality index are more objective. Other more advanced methods can be found in Raican et al.
(2011) and Qian et al. (2007). To handle water quality data with detection limits, simple
substitution methods as well as parametric and nonparametric approaches are investigated.
Substitution methods are simple but possibly biased. Nonparametric methods which do not
require the distributional assumption are robust and efficient (Helsel, 2005). Several popular
methods, such as Mann-Kendall, the seasonal Kendall test, and multiple regression methods,
are provided to detect and assess changes of various water quality parameters (Helsel,
1992). Meanwhile, nonlinear trends, serial dependence, covariate effects, and irregular
measurement patterns need to be considered (Abaurrea et al., 2011; Morton & Henderson,
2008). Computational implementation using R for linear regression models is introduced.
Examples using a real dataset are given to illustrate some very useful R functions.
7. Acknowledgments
Dr. Liya Fu is a postdoc fellow whose research was supported by the Centre for Applications
in Natural Resource Mathematics (CARM), School of Mathematics and Physics, the University
of Queensland, Australia.
8. References
Abaurrea, J., Asín, J., Cebrián, C. C. & García-Vera, M. A. (2011). Trend analysis of water
quality series based on regression models with correlated errors, Journal of Hydrology,
Vol. 400, 341–352.
Adler, J. (2009). R in a nutshell, O’Reilly Germany.
Brown, R. M., McClelland, N. I., Deininger, R. A. & Tozer, R. G. (1970). A water quality index:
Do we dare? Water and Sewage Works, 117, 339–343.
Brown, B. M. & Wang, Y-G. (2006). Induced smoothing for rank regression with censored
survival times, Statistics in Medicine, 26, 828–836.
Canadian Council of Ministers of the Environment (CCME). (2001). Canadian water quality
guidelines for the protection of aquatic life, CCME water quality Index 1.0,
Technical Report. In: Canadian environmental quality guidelines, 1999. Winnipeg:
Canadian Council of Ministers of the Environment. http://www.ccme.ca/assets/pdf/
wqi_techrprtfctsht_e.pdf.
Cleveland, W. S. (1979). Robust locally weighted regression and smoothing scatterplots,
Journal of the American Statistical Association, 74, 829–836.
Cleveland, W. S., Grosse, E. & Shyu, W. M. (1992). Local regression models, Chapter 8 of
Statistical Models in S eds Chambers and Hastie, Wadsworth & Brooks/Cole.
Chambers, J. M., Cleveland, W. S., Kleiner, B. & Tukey, P. A. (1983). Graphical methods for data
analysis, PWA-Kent Publishing Co., Boston.
166
Cohen, A. C. (1976). Progressively censored sampling in the three parameter log-normal

distribution, Technometrics, Vol. 18, No. 1, 99–103.
Cohn, T. A. (1988). Adjusted maximum likelihood estimation of the moments of lognormal
populations from type I censored samples, U.S. Geological Survey Open File Report,
88–350,
Crawley, M. J. (2007). The R book, John Wiley & Sons Inc.
Cude, C. G. (2001). Oregon water quality index: a tool for evaluating water quality
management effectiveness, Journal of the American Water Resources Association, 37,
125–137.
Debels, P. and Figueroa, R. and Urrutia, R. and Barra, R. and Niell, X. (2005). Evaluation of
water quality in the Chillán river (Central Chile) using physiochemical parameters
and a modified water quality index, Environmental Monitoring and Assessment, Vol.
110, No. 1301–322.
Dojlido, J. R., Raniszewski, J. & Woyciechowska, J. (1994). Water quality index applied to rivers
in the Vistula River Basion in Poland, Environmental Monitoring and Assessment, Vol.
33, 33–42.
El-Shaarawi, A. H. & Esterby, S. R. (1992). Replacement of censored observations by a constant:
an evaluation, Water Research, Vol. 26, Nol. 6, 835–844.
Farrel, R. (1980). Methods for classifying changes in environmental conditions, Technical
Report VRF-EPA7. 4-FR80-1, Vector Research Inc., Ann Arbor, Michigan.
Fu, L. & Wang, Y-G. (2011). Nonparametric rank regression for analyzing water quality
concentration data with multiple detection limits, Environmental Science and
Technology, Vol. 45, No. 4, 1481–1489.
Gleit, A. (1985). Estimation for small normal data sets with detection limits, Environmental
Science and Technology, 19, 1201–1206.
Gupta, A. K., Gupta, S. K. & Patil, R. S. (2003). A comparison of water quality indices for
coastal water, Journal of Environmental Science and Health, A38, 2711–2725.
Gnanadesikan, R. & Wilk, M. B. (1968). Probability plotting methods for the analysis of data,
Biometrika, Vol. 55, 1–17.
Harkins, R. D. (1974). An objective water quality index, Journal of Water Pollution Control
Federation, Vol. 46, No. 3, 500–591.
Heller, G. (2007). Smoothed rank regression with censored data, Journal of the American
Statistical Association, Vol. 102, No. 478, 552–559.
Helsel, D. R. (1990). Less than obvious-statistical treatment data below the detection limit,
Environmental Science and Technology, Vol. 24, No. 12, 1766–1744.
Helsel, D. R. (2005). Nondetects and data analysis: statistics for censored environmental data, Wiley:
New York.
Helsel, D. R. (2005b). More than obvious: better methods for interpreting nondetect data,
Environmental Science and Technology, 15, 419–423.
Helsel, D. R., Cohn, T. (1988) Estimation of descriptive statistics for multiply censored water
quality data, Water Resources Research, 24, 1997–2004.
Helsel, D. R. & Hirsch, R. M. (1992). Statistical methods in water resources, Elsevier Amsterdam,
The Netherlands.
Hewett, P. & Ganser, G. H. (2007). A comparison of several methods for analyzing censored
data, Applied Occupational and Environmental Hygiene, Vol. 51, No. 7, 611–632.
Statistical Tools
Analyzing Analyzing
25
Hirsch, R. M., Alexander, R. B. & Smith, R. A. (1991). Selection of methods for the detection
and estimation of trends in water quality, Water Resources Research, Vol. 27, No. 5,
803–813.
Hirsch, R. M., Slack, J. R. & Smith, R. A. (1982). Techniques of trend analysis for monthly water
quality data, Water Resources Research, Vol. 18, 107–121.
Hirsch, R. M. & Slack, J. R. (1984). A nonparametric trend test for seasonal data with serial
dependence, Water Resources Research, Vol. 20, No. 6, 727–732.
Hornung, R. W. & Reed, L. D. (1990). Estimation of average concentration in the presence of
nondetectable values, Applied Occupational and Environmental Hygiene, 5, 46–51.
Horton, R. K. (1965). An index-number system for rating water quality. Journal of Water
Pollution Control Federation, Vol. 37, No. 3, 300–306.
Jin, Z., Lin, D. Y. & Wei, L. J. (2003). Rank-based inference for the accelerated failure time
model, Biometrika, 90, 341–353.
Jin, Z., Lin, D. Y. & Ying, Z. (2006). On least-squares regression with censored data, Biometrika,
93, 147–161.
Jin, Z., Lin, D. Y. & Ying, Z. (2006b). Rank regression analysis of multivariate failure time data
based on marginal linear models, Scandinavian Journal of Statistics, 33, 1–23.
Kannel, P. R., Lee, S., Kanel, S. R. & Khan, S. P. (2007). Application of water quality indices
and dissolved oxygen as indicators for river water classification and urban impact
assessment, Environmental Monitoring Assessment, Vol. 132, 93–110.
Kendall, M. G. (1975). Rank correlation methods, Charles Griffin, London.
Landwehr, J. M. & Deininger, R. A. (1976). A comparison of several water quality indices,
Journal of Water Pollution Control Federation, Vol. 48, No. 5, 954–958.
Lee, L. & Helsel, D. R. (2005). Statistical analysis of water-quality data containing multiple
detection limits: S-language software for regression on order statistics, Computer &
Geosciences, 31, 1241–1248.
Lee, L. & Helsel, D. R. (2007). Statistical analysis of water-quality data containing multiple
detection limits II: S-language software for nonparametric distribution modeling and
hypothesis testing, Computer and Geosciences, 33, 696–704.
Lee, E. W., Wei, L. J. & Ying, Z. (1993). Linear regression analysis for highly stratified failure
time data, Journal of the American Statistical Association, 88, 557–565.
Lettenmaier, D. P. (1988). Multivariate nonparametric tests for trend in water quality, American
Water Resources Association, Vol. 24, No. 3, 505–512.
Lubin, J. H., Colt, J. S., Camann, D., Davis, S., Cerhan, J. R., Severson, R. K., Bernstein, L. &
Hartge, P. (2004). Epidemiologic evaluation of measurement data in the presence of
detection limits, Environmental Health Perspectives, 112, 1691–1696.
Mann, H. B. (1945). Non-parametric tests against trend, Econometrica, 13, 245–259.
Alternative for handling detection limits data in impact assessments, Ground Water Monitor
Remediation, 4, 42–44.
McClelland, N. I. (1974). Water quality index application in the Kansas river basin. U.S.
Environmental Protection Agency Region 7, Kansas City, Missouri.
McGill, R., Tukey, J. W. & Larsen, W. A. (1978). Variations of box plots, The American Statistician,
Vol. 32, No.1, 12–16.
Morton, R. & Henderson, B. L. (2008). Estimation of nonlinear trends in water quality: An
improved approach using generalized additive models, Water Resourse Research, Vol.
44, W07420, doi:10.1029/2007WR006191.
168
Prati, L., Pavenello, R. & Pesarin, F. (1971). Assessment of surface water quality by single index
of pollution, Water Research, 5, 741–751.
Pesce, S. F. & Wunderlin, D. A. (2000). Use of water quality indices to verify the impact of
Cordoba city (Argentina) on Suquýa river, Water Research, 34, 2915–2926.
Qian, Y., Migliaccion, K. W., Wang, Y. S. & Li, Y. C. (2007). Surface water quality evaluation
using multivariate methods and a new water quality index in the Indian River
Lagoon, Florida. Water Recourses Research, 43, 1–10.
Raican, S. M., Wang, Y-G., Harch, B. (2011). Water quality assessments for reservoirs using
spatio-temporal data from balanced/unbalanced monitoring designs. submitted
Sanchez, E., Colmenarejo, M. F., Vicente, J., Rubio, A., Garcia, M. G., Travieso, L. & Borja,
R. (2007). Use of the water quality index and dissolved oxygen deficit as simple
indicators of watersheds pollution, Ecological Indicators, 7, 315–328.
Shumway, R. H., Azari, R. S. & Kayhanian, M. (2002). Statistical approaches to estimating
mean water quality concentrations with detection limits, Environmental Science and
Technology, 36, 3345–3353.
Tobin, J. (1958). Estimation of relationships for limited dependent variables, Econometrica, Vol.
26, No. 1, 24–36.
Tukey, J. W. (1977). Exploratory data analysis, Addison-Wesley Pub., Reading, MA.
Van Belle, G. & Hughes, J. P. (1984). Nonparametric tests for trend in water quality, Water
Resources Research, Vol. 20, No. 1, 127–136.
Venables, W. N. & Ripley, B. D. (2002). Modern applied statistics with S, Springer.
Wei, L. J. (1992). The accelerated failure time model: a useful alternative to the Cox regression
model in survival analysis, Statistics in Medicine, Vol. 11, 1871–1879.
Yu, Y.-S., Zou, S. & Whittemore, D. (1993). Non-parametric trend analysis of water quality
data of rivers in Kansas, Journal of Hydrology, Vol. 150, 61–80.
Zhang, D. H., Fan, C. P., Zhang, J. & Zhang, C-H. (2009). Nonparametric methods for
measurements below detection limits, Statistics in Medicine, 28, 700–715.
Zandbergen, P. A. & Hall, K. J. (1998). Analysis of the British Columbia water quality index for
watershed managers: a case study of two small watershelds. Water Quality Research
Journal of Canada, 33, 519–549.
View publication stats

Statistical Tools For Analyzing Water Quality Data

Uploaded by

Copyright:

Available Formats

Statistical Tools For Analyzing Water Quality Data

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Statistical Tools For Analyzing Water Quality Data

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Statistical Tools for Analyzing Water Quality Data

Chapter · April 2012

Liya Fu You-Gan Wang

SEE PROFILE SEE PROFILE

Monitoring Queensland Mackerel Fisheries View project

PBVRMF View project

The user has requested enhancement of the downloaded file.

Statistical Tools for Analyzing Water Quality Data

2. Graphical analysis of water quality data

Station 1 Station 2 Station 3 Station 4

2.2 Scatter plots

Instantaneous flow on log scale in cubic feet per second

2.3 Q-Q plots

Concentrations of total phosphorus (mg/L)

(a) two stations (b) four stations

Fig. 3. Scatter plots of total phosphorus and instantaneous ﬂow

Instantaneous flow on log scale in cubic feet per second

to a theoretical distribution (often a normal distribution). Therefore, Q-Q plots provide a

Normal Q−Q Plot Normal Q−Q Plot

Theoretical Quantiles Theoretical Quantiles

3. Water quality index

0.00 0.05 0.10 0.15 0.20 0.25 0.30

Concentrations of total phosphorus at Station 1

3.1 Weighted water quality indices

3.2 Unweighted water quality indices

3.3 Harkins’ water quality index

4. Methods for handling data below detection limits

4.1 Simple deletion/substitution methods

4.2 Parametric methods

4.2.1 Tobit regression

4.2.2 Logistic regression

The likelihood function is

where pi = exp(α0 + α1 xi )/[1 + exp(α0 + α1 xi )]. The maximum likelihood estimates of

4.2.3 ROS method

Percent Chance of Exceedance

4.3 Nonparametric methods

4.3.2 AFT model

where Zi = log(Yi ), β is an unknown regression parameter vector, and i is the error

5.1 Parametric methods

5.1.1 Simple linear regression

5.1.2 Multiple regression

2006 2007 2008 2009 2010 2011

Fig. 9. Linear regression trend line for total phosphorus concentrations

Y = β0 + β1 T + β2 cos(2πT ) + β3 sin(2πT ) + , (11)

2006 2007 2008 2009 2010 2011

Fig. 10. Time series plot of total phosphorus concentration at Station 1

2006 2007 2008 2009 2010 2011

5.2 Nonparametric methods

5.2.1 Mann-Kendall test

var(S ) = [ n (n − 1)(2n − 5) − ∑ t(t − 1)(2t − 5)] /18,

5.2.2 The seasonal Kendall test

Under the null hypothesis, Si is a Mann-Kendall test statistic and

var(Si ) = [ n i (n i − 1)(2n i − 5) − ∑ ti (ti − 1)(2ti − 5)] /18.

and its expectation

Under the null hypothesis, Si and S j ( j = i ) are independent, therefore

The standard normal variate Z∗is deﬁned as

5.2.3 Sen’s T test

where Zi = Si / var(Si ), Si is the Mann-Kendall statistic in Equation (13), and Z̄ =

where Zi = log(Yi ), β is an unknown regression parameter vector, and i is the error

Y = β0 + β1 T + β2 cos(2πT ) + β3 sin(2πT ) + , (11)

Under the null hypothesis, Si and S j ( j = i ) are independent, therefore