Available online at www.banglajol.

info
Bangladesh J. Sci. Ind. Res. 55(2), 113-130, 2020

Seasonal variation of water quality and waste loads in Buriganga river with
GIS visualization

M. K. Alam, M. A. Uddin, M. F. Satter and N. Majed*

Department of Civil Engineering, University of Asia Pacific, 74/A Green Road, Dhaka-1205

Abstract

Pollution status of the Buriganga river has been assessed in this study using GIS tool to analyze the
spatial and seasonal distribution of several water quality parameters from six surface water sampling
stations. In addition, characterization of water quality from certain selected outfalls discharging
wastewater along the river was accomplished as well. The observed surface water quality
concentration was higher during dry season for BOD5 (13.68-22.8 mg/l), COD (22-30 mg/l), TDS
(280- 303 mg/l), PO4(5.6- 7.3 mg/l) while an exception was found for NH3-N (0.67- 0.89 mg/l) and
Turbidity (25.19 to 91.18). The DO level was found noticeably lower during dry season (1.73 to
2.36 mg/l) than wet season (3.27 to 4.68 mg/l). Moreover, Principal Component Analysis
identified high ionic distributions during wet season and more organic pollution during dry season.
Water quality parameters TDS, TSS, EC and COD were obtained in high concentration levels
from industrial outfalls towards the downstream of the river while BOD5, NH3-N were higher from
municipal wastewater outfalls towards the upstream of the river. The observed pollution loading
from the outfalls revealed the location at Dholaikhal as the major pollution source in the river with
Received: 29 August 2018
a BOD5 (11681- 12243 kg/day) and TDS (9749- 32348 kg/day) loading. Accordingly, influence of
Revised: 05 September 2019
nearest wastewater sources on the surface water quality was evident through GIS representation
Accepted: 11 November 2019 of the parameters.

DOI: https://doi.org/10.3329/bjsir.v55i2.47632 Keywords: Buriganga river; GIS; Outfalls; Seasonal variation; Spatial distribution; Waste load

Introduction

Buriganga River, the lifeline of Dhaka city is considerably sewerage coverage (DWASA, 2014). Lack of adequate
most polluted among the surrounding four rivers around the sewerage system results in direct discharge of municipal
city (Sarker, 2005). In recent years, both the width and the waste into the river or nearby water body which is eventually
flow rate of the river have reduced due to the building up of carried out to the Buriganga river. These untreated wastes are
illegal manmade structures and waste dumping in the banks harmful for existing aquatic ecosystem and water quality in
of the river. The changes in hydrodynamic properties of the the river. The situation is even worse during dry season with
river is also influencing the self-cleansing property of the reduced water flow and low rainfall flushing. In orderto
river. With increasing pollution load, the sign is very ascertain the overall impact on surface water quality,
alarming for the existence of the river. The most significant identification of wastewater sources and characterization of
sources of wastewater in the river are industries along the discharge sources are important. In addition, it will be of
bank of the river. There are 627 dying industries located interest to analyze the distribution of water quality parameters
beside the Buriganga River and near 5000 cubic meter along the river to evaluate the water quality profile of the river
effluents are discharged every day from these industries stretch and identify pollution zones in the river.
(Ahammed et al., 2016). Previous studies also identified 22
large industrial outlets along the Buriganga river (Rahman Efforts have been made in previous studies to investigate the
and Bakri, 2010). The existing sewerage system is quite old surface water quality in the Buriganga river to understand
and major northern parts of the mega city Dhaka is still out of seasonal and temporal variation of water quality parameters

*Corresponding author e-mail: nehreen-ce@uap-bd.edu

114 Water quality of Buriganga river with GIS visualization 55(2) 2020

(Rahman and Bakri, 2010; Saifullah et al., 2013; Ahammed the midstream of the study area, does not exist anymore and
et al., 2016; Sarkar et al., 2015). The waste loading from now converted into a box culvert (Rahman and Rana, 1996).
potential sources in the Buriganga River was estimated in the Many textile industries are mushroomed in downstream of
study by Kamal et al. (1999) along with the investigation of the study area near Dhaka-Narayanganj highway at
pollution status in the river. Some researchers used GIS tools Narayanganj.
to specify sampling locations in the Buriganga river for
analyzing water quality parameters (Moniruzzaman et al., Sampling stations
2009; Rahman and Hossain, 2008). One previous study by
Ouyang et al. (2006) showed the use of principal component Six wastewater outfalls and six surface water sampling
analysis to associate between water quality parameters and stations were selected near individual outfalls for water
variation of the levels of parameters between different sampling as delineated in Figure 1. The outfalls were
seasons, whereas Gao et al. (2015) used principal component identified through field trips around the Buriganga river using
analysis and hierarchical cluster analysis to understand boat and were selected depending on their magnitude of
pollution status and source of the Pu river.This study (Gao et discharge, size and location. Sampling stations were selected
al., 2015)also exhibited spatial distribution of water quality considerable distance away towards downstream from the
parameters of Pu river in China using spatial interpolation selected outfalls to assess the impact of the outfall discharge
method in GIS to classify different pollution zones in the on the river water quality. The water samples were collected
river. However, no such elaborated study using GIS has been from six outfalls and six water sampling stations for two
performed for Buriganga river so far to understand the spatial different seasons. The wet season sampling was done in
distribution of water quality parameters in the river. August 2017 and the dry season sampling was done in
December 2017.
The present study attempted to identify major wastewater
outfalls and characterize outfalls’ discharge along with Sample collection and analysis
surface water quality of the river in two different seasons.
One-liter size pre washed PVC bottles were used for the
Through the evaluation of water quality parameters from the
sampling and before sampling the sample bottles were
selected outfalls and surface water quality in the proximity of
washed with sample water to maintain the accuracy of
the sources, the intention was to understand the impact of
sampling. The wastewater samples from the outfalls were
outfalls discharge on the river water quality. The observed
collected at the discharging point. The surface water samples
levels of water quality parameters were also compared with
were collected considerable distance away towards
the Bangladesh Environment Conservation Rule 1997
downstream in the river from each of the outfalls and at one
(ECR,1997). Moreover, GIS analysis was performed along
feet depth from the water surface to avoid floating pollutants
with the statistical analysis of water quality to represent
and scums. Water samples were analyzed for twelve water
important water quality parameters in the river stretch.
quality parameters during wet and dry seasons which are pH,
Materials and methods color, turbidity, electrical conductivity (EC), total dissolved
solids (TDS), total suspended solids (TSS), dissolved oxygen
Study area (DO), 5-day biochemical oxygen demand (BOD5), chemical
oxygen demand (COD), ammonia-nitrogen (NH3-N),
Buriganga river flows at the north central region of nitrate-nitrogen (NO3-N) and phosphate (PO4).The dissolved
Bangladesh near Dhaka city with a total length of 25 km from oxygen, pH, TDS, TSS and EC were analyzed in laboratory
Mirpur to Fatulla (Haque, 2008). Originating from the river setup using multiparameter (Model: Hach HQ40D) as per
Turag near Aminbazar Bridge at Mirpur, it falls at standard methods (APHA, 1992). Ammonia nitrogen was
Shitalakhya river near Naraynganj (Haque, 2008). Flow measured by Nessler method while phosphate and
characteristics of the Buriganga river is non-tidal during dry nitrate-nitrogen were analyzed by Molybdovanadate and
season and tidal in wet season. In dry season, the flow of the Cadmium reaction method respectively using the
river is near nonexistent and water level remains lower spectrophotometer; these experiments were also done
(SWMC, 1996). For evaluating the distribution of wastewater following the standards (APHA, 1992). The color and
sources along the river stretch, the study was performed along turbidity were measured using platinum-cobalt and
6 km of the river stretch from Mitford to Pagla near Nephelometric method respectively as per the standard
Naraynganj (Figure 1). With densely populated area in both method (APHA, 1992). The BOD5 was analyzed with BOD
sides of the river at upstream of the study area, most of the trak II (Model: Hach) and BOD incubator (Model: Hach
wastewater sources are from municipal sewage connections. FOC120E) using potassium hydroxide and BOD nutrient
The large wastewater outlet Dholai khal used to discharge at buffer pillow reagent by BODTRAK method mentioned in
 Alam, Uddin, Satter and Majed 115

Fig. 1. Selected outfalls and surface water sampling stations in the study area of Buriganga river

the study (Hach et al., 1997). The COD was measured using for total three outfalls (O1, O3, O5) were measured through
COD reactor (Model: Hach DRB200) and Spectrophotometer direct measurements during the field visit. However, flow rate
(Model: Hach DR 6000) by reactor digestion method (Jirka of outfall O4 was not measured due to the inconsistent and
and Carter, 1975). insignificant flow during the flow measurement period. Thus,
outfall O4 was omitted from waste loading estimation.
Waste loading estimation
Data analysis: GIS and statistical methods
Flow rate of the outfalls were ascertained during the dry
season in December 2017. The dry season was chosen for the Spatial interpolation involves estimating the values of
measurement due to low water level which would facilitate in properties at non sampledsites within an area under existing
identifying the route of water discharge from the outfalls. The observations (Ian et al., 2001). In this study, spatial
whole waste loading method was done using distance method interpolation method was used to show the spatial
from the study by Sabit (2011) rather than involving any distribution of water quality parameters in the study area
special flow measurement equipment. The first step is using the data collected from six surface water sampling
measurement of cross sectional area of the flow through the stations. Total seven water quality parameters from six
outfalls. Later, the velocity through the outfalls were different sampling stations were processed using krigging
measured using a flowing object by the time it travels a method in Arcmap 10.2 as described in Gao et al. (2015). For
specific distance (Sabit, 2011). Eventually flow rate was using krigging approach, a georeferenced GIS map of the
obtained multiplying the cross sectional area of flow and study area was prepared and attribute table was created with
measured velocity of discharge. After the flow measurement, measured water quality parameter data.
the waste loading rate pertaining to respective pollutant was
estimated by multiplying the concentration of the respective Cluster analysis is known for unsupervised pattern detection
pollutant with the estimated flow rate stated by Kamal (1996). method which partitions different cases that are dissimilar to
Waste loading rate of the outfalls were measured for wet and other groups (Zhang et al., 2009). Hierarchical cluster
dry seasons, only using the dry season flow rate. Flow rates analysis abbreviated as HCA hereafter was applied in the
116 Water quality of Buriganga river with GIS visualization 55(2) 2020

study using squared Euclidean distance and ward method for are discharging. Outfalls O1 and O2 along the upstream zone
forming group of outfalls according to the discharge quality. were municipal outfalls and the rest (O3, O4, O5 and O6)
were industrial outfalls. Compared to municipal waste water
Principal component analysis (PCA) involves simple and outfalls, water quality parameters such as COD, TDS, TSS
meaningful representation of the factors by decreasing the and EC were found to be higher in general at the industrial
contribution to PC’s of variables with minor significance and outfalls (438 %, 157 %, 517 % between average levels for the
increasing the more significant ones (Vega et al., 1998). This mentioned parameters respectively). Suffice it to say, the
study used PCA analysis to identify major pollutant factors in chemical contents used in the industries seem to be
different seasons of Buriganga river. responsible for the increase in ionic distributions (TDS, EC)
whereas untreated effluents with particulate matter might
In this study, the mathematical and statistical computation have caused the high TSS concentration. The seasonal
was performed using the statistical software SPSS 20. variation of the water quality parameters revealed an
increasing trend from wet to dry season for most of the
Results and discussions
industrial outfalls except at O6. The reduced concentration
Seasonal variation of outfall discharge might be caused due to production rate on nearby factories
and reduced rainfall which carried more wastewater during
Table I displays the analysis results of water quality the wet season from surrounding industries. Meanwhile, at
parameters from the outfalls discharge in two different municipal outfalls, BOD5 and PO4 levels were obtained at
seasons. The selected outfalls in this study could be high concentrations mainly during the dry season (more than
categorized either as municipal or industrial outfalls 5% and 47 % respectively than wet season). In contrast,
depending on the sources of the wastewater that the outfalls NH3-N was surprisingly higher (more than 5%) during the

Table I. Water quality parameters of outfalls discharge in two different season

Para. O1 O2 O3 O4 O5 O6
Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry

pH 7.78 7.04 7.05 7.03 7.35 7.12 7.42 7.91 9.39* 7.87 9.94* 8.67

BOD5
mg/l 9.98 136.8* 104* 109.44* 121* 209.76* 27 45.6 58* 108.3 * 129 * 42.6

COD
mg/l 14 197 252 174 477 * 661 * 105 126 80 310 * 295 * 199

PO4
mg/l 3.2 41.6* 34 50 * 75.2 76 24.9 27.3 17.1 58.5 37.7 43.2

NO3-N
mg/l 0 3.2 0 0.2 7.3 5.8 0 0.3 0 0 0 0

NH3-N
mg/l 0.74 0.7 26.05 0.55 22.15 1.55 11.2 14.75 6.4 22.05 3.85 3.1

TDS
mg/l 63 462 337 409 504 750 161.8 280 275 776 766 502

TSS
mg/l 17 141* 119* 122* 791 * 207 * 238 * 108 54 84 285 * 54

EC
mg/l 132.4 930 692 824 1026 1483 * 336 565 568 1512 * 1541 * 995

According to the standard (ECR, 1997), designated unsafe for *: discharge into water body
 Alam, Uddin, Satter and Majed 117

wet season compared to dry season. The reason could be due first group comprised wastewater outfalls O2, O4, O3 and O1
to the high chance of augmented surface runoff through in the Buriganga river. Although, the outfalls O3 and O4 were
municipal outfalls during wet season, which brings off more identified and categorized as industrial outfalls through field
ammonia compounds in the river. Among the municipal observation, however the HCA analysis grouped these
outfalls, the highest seasonal variation was observed at outfall outfalls in the similar category as the municipal outfalls based
O1 (varied 300% to 1307%) whereas the highest on the discharge characteristics. This outcome indicates that
concentration was obtained at outfall O2 for most of the these outfalls also discharge municipal wastewater along with
parameters. Comparing of observed data from the municipal industrial effluents. Accordingly, the outfalls were observed
and industrial outfalls revealed that, DO level was higher at to contain high amount of BOD5, COD and PO4, indicative of
the industrial outfalls discharge during both the seasons organic pollution. On the other hand, the other two industrial
(Mainly O5 and O6). The higher DO level is indicative of wastewater outfalls O5 and O6 were grouped as another
lesser amount of organic pollutants at industrial outfalls than cluster. Accordingly, this cluster could be characterized to
that in the municipal outfalls. Concentration levels of most of contain discharge of similar water quality as both outfalls
the parameters exceeded the standards for wastewater within the cluster carry industrial wastewater from local
discharge into water body as specified in the standard (ECR, textile industries. Field observation also confirmed that these
1997) which are indicated in the table with asterisks. outfalls usually carry heavy textile dye.

Grouping of outfalls by discharge quality Seasonal variation of surface water quality

Data of two different seasons with total 12 water quality Figures 3 through 12 show the spatial and temporal variation of
parameters from six different wastewater outfalls were water quality parameters pH, BOD5, COD, DO, PO4, NO3-N,
processed for HCA, analysis was attempted with ward NH3-N, TDS, turbidity and TSS along the selected stretch of
method for clustering and Squared Euclidean distance was the river. The pH of the river water varied from 6.92 to 7.73
used as computed range. A dendogram was created grouping during wet season and ranged from 7.02 to 7.71 during dry
the six wastewater outfalls into two clusters (Figure 2). The season at surface water sampling stations (Figure 3).

Rescaled Distance Cluster Combine

Fig. 2. Dendogram showing the clusters of outfalls based on discharge quality

118 Water quality of Buriganga river with GIS visualization 55(2) 2020

pH level

Distance (km)

Fig. 3. Seasonal variation of pH at surface water sampling stations

BOD in mg/1

Distance (km)

Fig. 4. Seasonal variation of BOD5 at surface water sampling stations

 Alam, Uddin, Satter and Majed 119

Observed results of water quality parameters in two different dye and chemical effluents that are regularly discharged from
seasons suggested no seasonal variation of pH level in river nearby wastewater sources. However, among the sampling
water. However, during both the seasons, the highest pH level stations, the highest seasonal variation was observed
was recorded at downstream of the river near industrial zones atstationsS5 and S6 (more than 200%) and minimum level
likely due to the high chemical content discharge from local was obtained (below 15%) at S4.
industrial sources. The pH level in two different seasons were
within the permissible standard for aquatic living standard The DO concentration varied from 3.27 to 4.68 mg/l during
(ECR, 1997). wet season, whereas during dry season, the concentration
varied from 1.73 to 2.36 mg/l (Figure 6). Due to the high
Compared to wet season data, the observed level of precipitation and tidal effects during the wet season in the
BOD5was considerably higher during dry season at most of river, DO level remained higher. However, with the
the sampling stations (Figure 4). The highest BOD5 (22.8 progression of dry season, the water level goes lower and
mg/l) between two different seasons was recorded during increasing level of pollution causes depletion of DO in river
dry season at three different sampling stations S2, S5 and water due to insufficient natural purification. During the wet
S6 respectively whereas during wet season, high season, the DO concentration was higher at upstream of the
concentration was obtained at the midstream of the river river and lower at downstream of the river. In contrast, the
near the sampling stations S2 and S4. The highest DO level was found higher at downstream of the river and
concentration along the downstream occurred allegedly due lower at upstream of the river during dry season. Lesser
to the number of industries along this area. Whereas during amount of organic wastes from the industrial sources
wet season the high BOD5 level at midstream was observed compared to municipal sources are likely to increase DO
possibly due to the high municipal waste discharge from level at downstream during dry season. According to the
Dholaikhal (O2) point. However, the BOD5 value was standard for aquatic living, during both the seasons, the DO
above the permissible range for aquatic life standard for levels were mostly lower than the required specified level
both the seasons (ECR, 1997). (ECR, 1997).

The COD level varied from 8 to 24 mg/l during wet season The observed levels of phosphate (PO4) were comparatively
and from 22 to 30 mg/l during dry season at the selected higher during dry season than those in wet season (Figure 7).
sampling stations (Figure 5). The highest COD (30 mg/l) was During the wet season, the highest value (4.9 mg/l)was
recorded along the midstream of the river at sampling station observed at S2 sampling station towards upstream of the
S3 during dry season and the lowest (22mg/l) was obtained at river. In contrast, the highest concentration of phosphate was
sampling station S4 during dry season. These levels of COD observed at the downstream of the river near industrial zones
could have occurred possibly due to high amount of textile during dry season. Most of the industries in these areas
COD in mg/1

Distance (km)

Fig. 5. Seasonal variation of COD at surface water sampling stations

120 Water quality of Buriganga river with GIS visualization 55(2) 2020

DO in mg/1

Distance (km)

Fig. 6. Seasonal variation of DO at surface water sampling stations

Phosphate in mg/1

Distance (km)

Fig. 7. Seasonal variation of phosphate (PO4) at surface water sampling stations

discharge their effluents without treatment in the river, which concentration was found at the upstream of the river,
is the reason of high amount of phosphate in this stretch of changing to increased concentration in midstream which
the river. However, due to the high precipitation in wet gradually decreased towards the downstream of the river.
season, the concentration of phosphate was not obtained as Meanwhile, during dry season, higher concentration was
high as in dry season along the downstream of the river. observed at the upstream of the river (0.5 to 0.6 mg/l), where
municipal wastewater sources are dominating. During both
The nitrate-nitrogen (NO3-N) concentration ranged from 0 to the seasons, the nitrate-nitrogen concentration was lower at
0.6 mg/l at surface water sampling stations during both the downstream of the river near industrial zones, while the
seasons (Figure 8). During wet season, no nitrate highest concentration was obtained at sampling station S3.
 Alam, Uddin, Satter and Majed 121

Nitrate-nitrogen in mg/1

Distance (km)

Fig. 8. Seasonal variation of nitrate-nitrogen (NO3-N) at surface water sampling stations

Ammonia-nitrogen in mg/1

Distance (km)

Fig. 9. Seasonal variation of ammonia-nitrogen (NH3-N) at surface water sampling stations

Unlike the other water quality parameters, ammonia-nitrogen surface water runoff that brings more ammonia compounds in
(NH3-N) concentration was observed to be high during the river water. However, during dry season insignificant
wet season than that in the dry season (Figure 9). The concentration was observed at downstream of the river.
ammonia-nitrogen concentration varied from 0.67 to 0.89
mg/l during wet season and from 0 to 0.15 mg/l during dry Total dissolved solids were observed at 63.8 to 86.8 mg/l
season. During the wet season, highest ammonia during wet season and at 280 to 303 mg/l during dry season
concentration was obtained at the downstream of the river (Figure 10). Meanwhile, total suspended solidwas measured
probably due to the highest precipitation and augmented at 16 to 23 mg/l during wet season and 16 to 37 mg/l during
122 Water quality of Buriganga river with GIS visualization 55(2) 2020

Total dissolved solids in mg/L

Distance (km)

Fig. 10. Seasonal variation of total dissolved solids (TDS) at surface water sampling stations
Turbidity in Ntu

Distance (km)

Fig. 11. Seasonal variation of turbidity at surface water sampling stations

dry season (Figure 11). TDS concentrations increased almost (37 mg/l) during dry season. Notably, the TSS values were
4 times in all the sampling stations during dry season than the obtained at larger levels at downstream of the river during
wet season data. The highest TDS were observed 86.8 mg/l dry season similar to TDS values.
during wet season near S2 (86.8 mg/l) while the highest was
at S4 (303) during dry season. Industrial activities at Turbidity was observed at 25.19 to 91.18 mg/l during wet
downstream of the river might have caused this high season and 41.9 to 70.79 mg/l during dry season in Buriganga
concentration near S4. On the other hand, highest value of River (Figure 12). Comparing both the seasons, high
TSS was observed at S3 (23 mg/l) during wet season and S6 concentration was observed near the midstream around S3
 Alam, Uddin, Satter and Majed 123

Total suspended solids in mg/L

Distance (km)

Fig. 12. Seasonal variation of total suspended solids at surface water sampling stations

(81.05 mg/l) and S4 (91.18 mg/l) during the wet season. explained by 21.63% of the total variance and had strong
Apart from this midstream, dry season data was higher than positive loadings on COD (0.787) and strong negative
the wet season data at all other sampling stations. Municipal loading on color (0.882).This is due to the organic pollution of
sewage discharges coming from Dholaikhal which is located the water which affected color of the water quality. On the
at downstream near the sampling station S3, S4 in addition to other hand, component 3 had a positive loading on NO3-N and
extra rainfall carrying huge amount wastes from local TSS and explained by 21.16 % which suggests the influence
livelihoods might be the reasons behind increased turbidity of inorganic constituents. These inorganic constituents might
during wet season. have occurred due to the anthropogenic inputs in the river
(Ouyang et al., 2006). Whereas, the component 4 having
Principal component analysis of surface water quality negative loading on DO also suggests organic pollution in the
river that caused depletion of DO level.
PCA is a multivariate statistical analysis method which
reduces the dimensionality of a data set with large number of During the dry season, total variance of component 1 had
inter related variables, it also allows to assess the found 47.51% variance and had positive loadings on pH, DO,
participation of individual chemicals by several influence TDS, EC, PO4 (>0.80) and strong negative correlation NH3-N
factors (Kebede and Kebedee 2012). In this research, PCA (Table IV). Thus, component 1 is reflecting the mineral and
were applied for 12 water quality parameters for six different inorganic nutrients (PO4, NH3-N) (Kebede and Kebedee
locations of the Buriganga river in two different seasons, the 2012). These constituents were mostly influenced by the DO
results of which are shown in Tables III and IV. The PC’s level in the river during dry season. Component 2 was found
with eigenvalues over 1 were considered as essential to to have strong positive loading on COD, Turbidity and strong
complete the analysis. The wet seasonal datasets were yielded negative loading on color and explained by 30.32 %. This
by 4 PCs, while the dry seasonal dataset had 3 PCs and explains the degradation of the river water quality in terms of
respectively had a total variance 95.09 % and 95.357 % color and turbidity due to the organic pollution. While the
respectively. component 3 had a strong negative loading on BOD5 and
NO3-N and explained by 17.54%.The component 2 and 3
During the wet season, total variance of component 1 had again displaying sign of high organic pollution sign and
36.60 % variance which had strong positive loadings on TDS, influence of domestic pollution sources.
PO4, EC (>0.95) and negative loading on pH (Table III).
Thus, component 1 represented the ionic distribution in the To summarize, results indicated that during the wet season,
river water and influenced the pH level. Component 2 is the parameters TDS, PO4, DO and EC had a high loading
124 Water quality of Buriganga river with GIS visualization 55(2) 2020

Table II. PCA analysis of surface water quality parameters for wet season

Indicator Component
1 2 3 4

pH -0.742 -0.590 -0.298 -0.105

BOD5 0.511 0.509 0.456 0.455
COD 0.263 0.787 0.341 0.360
DO -0.151 -0.271 0.007 -0.934
NO3-N -0.215 -0.048 0.816 0.491
NH3-N 0.637 -0.284 -0.604 0.347
TDS 0.983 0.172 0.007 0.043
TSS 0.353 0.137 0.849 -0.175
Turbidity 0.108 0.591 0.601 0.502
PO4 0.981 0.082 0.097 0.116
Color -0.079 -0.882 0.060 -0.061
EC 0.983 0.172 0.006 0.043
Eigenvalue 4.393 2.595 2.539 1.884
Total variance % 36.609 21.626 21.155 15.70
Cumulative variance % 36.609 58.235 79.390 95.090

Table III. PCA analysis of surface water quality parameters for dry season

Indicator Component
1 2 3

pH 0.968 -0.185 0.098

BOD5 0.077 0.149 -0.718
COD 0.015 0.931 -0.084
DO 0.905 0.292 0.187
NO3-N -0.612 0.149 -0.718
NH3-N -0.945 0.278 0.049
TDS 0.848 0.434 0.243
TSS 0.664 0.606 0.438
Turbidity -0.132 0.916 0.261
PO4 0.868 0.131 0.386
Color -0.335 -0.793 -0.485
EC 0.792 0.556 0.206
Eigenvalue 5.701 3.638 2.104
Total variance % 47.508 30.316 17.534
Cumulative variance % 47.508 77.824 95.357
 Alam, Uddin, Satter and Majed 125

(>0.9) and constituted the main pollution factors in the river. NH3-N and high PO4 levels were observed. In the downstream
Meanwhile, during dry season, the parameters pH, DO, zone, high COD, PO4, TDS level is due to the increasing
NH3-N, Turbidity, COD (>0.9) could be identified as the number of industries and high DO level is the result of less
main pollution factors in the river. Though during wet season, organic pollutant from these industrial sources. However,
ionic parameters (EC, TDS) and PO4 were explained by after the water quality distribution map it can be concluded
as high as 36.6% variance but during dry season DO, pH, that water quality in river water was highly influenced by the
NH 3-N were explained by 47.50% variance. It eventually discharging wastewater sources along the river bank.
appeared from the PCA analysis that during the dry
season organic pollution was the main pollution factor in Waste loading characteristics of outfalls
the river than wet season as DO and NH3-Nwere
explained by high variance. The estimated waste loading data for two different seasons in
a map of study area along with the outfalls locations are
Spatial distribution of water quality parameters shown in Figure 14. With a BOD5 loading rate of 11,681
kg/day during wet season and 12,243 kg/day during dry
The spatial distributions of water quality parameters within season, outfall O2 discharges more than 500 times of BOD5
the study area were obtained through processing of surface during wet season and more than 35 times during dry season
water quality data obtained during two different seasons in a compared to the other municipal outfall O1 considered in this
GIS based krigging method. The GIS outputs are shown in study. Though the previous loading rate of BOD5 reported for
Figure 13. The GIS representation enabled a comparative O2 in the study (Kamal, 1999) was much higher (29,980
assessment of the distribution of the water quality parameters kg/day) than the present study. Concentration of BOD5 was
through the entire reach of the river and graphical also at220 mg/l in the previous study which is much higher
categorization was possible from severely polluted zone to than the present data. Thus, the pollution scenario changes
comparatively less polluted zone. Most of the water quality substantially with the sources of wastewater discharge into
parameters were found in lowest concentrations at the the river especially alongside the relocation of industries.
upstream of the river during wet season. The high level of DO Apart from BOD5, the outfall O2 is also the highest waste
(4.40-4.68 mg/l), low BOD5 and Turbidity during wet season contributor for TDS (32348 kg/day), TSS (13703 kg/day),
indicated that this portion of the river was comparatively COD (19544 kg/day) and PO4 (5616 kg/day) during dry
pollution free and that wastewater sources were less polluted. season. However, observed results for PO4 loading in present
In contrast, during dry season, high concentration of NO3-N study are observed to be higher than the mentioned loading
and lowest DO level (1.72-1.85 mg/l) indicated pollution for O2 in previous study (Kamal, 1999). The observation
from organic matter from the nearest wastewater sources, made among the municipal outfalls led to the result that the
most likely due to increased number of municipal wastewater major municipal pollution source in the study area was outfall
sources in this area. GIS distribution revealed high O2. Meanwhile, outfall O6 was the highest waste load
concentrations ofBOD5, NO3-N, COD and lower DO in both contributor at the river among all the industrial sources in this
the seasons between the stations S2 and S4, signifying the study. Among the industrial sources, highest BOD5 (1700
mid portion of the river to be the severely polluted zone kg/day), COD (3886 kg/day), TSS (3754 kg/day)was
within the study area. During the wet season, BOD5, NO3-N, obtained at O6 during the wet season, whereas highest levels
Turbidity, COD and PO4 were highest in this region and of TDS (3978 kg/day) and PO4 (570 kg/day) were obtained
showed reduction in DO level from the upper reaches of the during dry season. Similar to outfall O6, O5 was found to
river. While, during the dry season BOD5, COD, NO3-N, contribute high amount of waste loading in the Buriganga
NH3-N found highest for this region. Moreover, considering river. High waste loading rate from the outfall O6 and O5 is
seasonal variation, notable reduction was observed during due to the large flow rate and size of the out falls than the
dry season in BOD5 and DO level than the wet season. The other industrial wastewater source O3. High COD and TDS at
water quality parameters NH3-N, BOD5 and lower DO of this the outfalls is likely due to the chemical wastes and dye
region is affected by the huge amount of municipal waste injected from the local textile industries. Most of the outfalls
discharge from the Dholaikhal (O2) and NO3-N, COD was were found to discharge high concentration of wastewater
affected by the textile industry effluents from the source O3 during dry season than the wet season, with the exception at
and O4. The distribution of DO levels at downstream of the O6. The reduced waste loading rate during dry season is
river between S4 to S6 were acceptable than the upper probably due to the large fluctuation in flow from wet to dry
reaches during both of the seasons. Meanwhile, during the season. Overall, the waste loading data suggested outfall O2
dry season, BOD5, COD, Turbidity, TDS and PO4 were found (Dholai khal) to be the major pollution route in the Buriganga
to be higher at downstream and unlike wet season, low river than the other selected outfalls in the study area.
126 Water quality of Buriganga river with GIS visualization 55(2) 2020

(a)

(b)

(c)

(d)
 Alam, Uddin, Satter and Majed 127

(e)

(f)

(g)

(h)

Fig. 13. Spatial distribution of water quality parameters during dry and wet season through GIS mapping: (a)
BOD5; (b) COD; (c) DO; (d) PO4; (e) NO3-N; (f) NH3-N; (g) TDS; (h) Turbidity
128 Water quality of Buriganga river with GIS visualization 55(2) 2020

Fig. 14. Waste loading rates in kg/day for BOD5, COD, TDS, TSS and PO4 of selected outfalls during wet (W)
and dry season (D) [outfall O4 was omitted during load calculation]

However, dry season flow rate was utilized for waste load exceeded the discharge standard. Assessment of seasonal
calculation for both the seasons which might have caused an variation of surface water quality concluded that during dry
underestimation in the waste loading estimation for wet season the pollution status of the river becomes severe.
season. According to the guidelines of water quality established in
Environment Conservation Rules, the river water is in
Conclusion vulnerable situation for existence of aquatic living. The PCA
analysis identified high ionic distribution (TDS, EC)during
This study combines the statistical technique and GIS wet season while organic pollutants were the main pollution
analysis work to identify current pollution status of factors (DO, NH3-N, COD) during dry season. Moreover, the
Buriganga River in the selected study area. The research GIS spatial distribution data and waste loading data revealed
work also identified major wastewater outfalls and that major pollution source in the study area is Dholaikhal
characterized outfalls discharge in the selected study area. discharging point (O2) and the major outfall discharging
The characterization included the chemical analysis of water industrial effluent is O6 at downstream of the river near
quality parameters and waste loading measurement of the Pagla. The surface water quality and the outfall discharge
outfalls. Water quality parameters TDS, TSS, EC and COD observation concluded that the surface water quality
were obtained in high concentration from industrial outfalls parameters are highly influenced by nearest wastewater
at downstream of the river and BOD5, PO4, NH3-N were sources. Thus, proper treatment of wastewater before
higher mainly from municipal wastewater outfalls at discharging into the river water is strongly recommended.
upstream of the river. In addition, lack of treatment of However, as pollution status shows significant improvement
industrial effluents before discharging into the river was during wet season, the study of hydrological parameters
observed as most of the discharge quality of the outfalls during the wet season may be sought to learn about the
 Alam, Uddin, Satter and Majed 129

impact of river hydrological characteristics on pollution Water science and technology 40(2): 129-136.
status of the river. Finally, adequate coverage of sewerage
Kebede YK and Kebedee T (2012), Application of principal
system all around the city is the ultimate solution to prevent
component analysis in surface water quality
the current pollution level along the river.
monitoring. In Principal Component
References Analysis-Engineering Applications. In Tech Open.
Moniruzzaman M, Elahi SF and Jahangir MAA (2009),
Ahammed SS, Tasfina S, Rabbani KA and Khaleque MA
Study on temporal variation of physico-chemical
(2016), An Investigation into The Water Quality of
parameters of Buriganga river water through GIS
Buriganga River Running Through Dhaka.
(Geographical Information System) Technology.
International Journal of Scientific & Technology
Bangladesh Journal of Scientific and Industrial
Research 5(3): 36-41.
Research 44(3): 327-334.
APHA (1992), Standard Methods for the Examination of
Water and Wastewater, eighteenth ed. American Ouyang Y, Nkedi-Kizza P, Wu QT, Shinde D and Huang CH
Public Health Association, Washington, DC. (2006), Assessment of seasonal variations in surface
water quality. Water research 40(20): 3800-3810.
DWASA (2014), Annual Report 2013-14, Dhaka Water and
Sewage Supply Authority, p 48. Rahman MA and Bakri DA (2010), A Study on Selected
Water Quality Parameters along the river Buriganga,
ECR (1997), Environment Conservation Rules, Department Bangladesh. Iranica Journal of Energy & Environment
of Environment. Ministry of Environment and Forest, 1(2): 81-92.
People’s Republic of Bangladesh.
Rahman MR and Rana MY (1996), Pollution Assimilation
Gao H, Lv C, Song Y, Zhang Y, Zheng L, Wen Y and Yu H Capacity of Buriganga River. J Civil Eng (Bangladesh)
(2015), Chemometrics data of water quality and CE24, pp 85-95.
environmental heterogeneity analysis in Pu River,
China. Environmental Earth Sciences 73(9): Rahman S and Hossain F (2008), Spatial Assessment of
5119-5129. Water Quality in Peripheral Rivers of Dhaka City for
Optimal Relocation of Water Intake Point. Water
Hach CC, Klein JR, RL and Gibbs CR (1997), Introduction to Resources Management 22(3): 377-391.
Biochemical Oxygen Demand. Technical information
series, Booklet no. 07 Sabit MI (2011), Evaluation of water quality and pollution
sources of Gulshan Lake. M. Sc. Engineering Thesis,
Haque MI (2008), Water Resource Management in Bangladesh University of Engineering and
Bangladesh. 1st Ed. (Anushilanprakashani, Dhaka), p Technology, Dhaka.
80.
Saifullah AS M, Kabir MH, Khatun A, Roy S and Sheikh MS
Ian H, Sarah C and Steve C (2001), An Introduction to (2013), Investigation of Some Water Quality
Geographical Information Systems. 2nd Ed. (Pearson Parameters of the Buriganga River. Journal of
Education, India), p 117. Environmental Science and Natural Resources 5(2):
Jirka AM and Carter MJ (1975), Micro semiautomated 47-52.
analysis of surface and waste waters for chemical Sarkar M, Rahman AL, Islam JB, Ahmed KS, Uddin MN and
oxygen demand. Analytical chemistry, 47(8): Bhoumik N C (2015), Study of hydrochemistry and
1397-1402. pollution status of the Buriganga River, Bangladesh,
Kamal MM (1996), Assessment of Impact of Pollutants in the Bangladesh Journal of Scientific and Industrial
River Buriganga Using a Water Quality Model. M. Sc. Research 50(2): 123-134.
Engineering Thesis, Bangladesh University of Sarker MAK (2005), Study on the Accumulation and
Engineering and Technology, Dhaka. Seasonal Variation of Trace Metals in Water, Sediment
Kamal MM, Malmgren-Hansen A and Badruzzaman ABM and Some Fishes of the Buriganga River. M.Sc thesis
(1999), Assessment of pollution of the River (1998-1999), Department of Zoology, Dhaka
Buriganga, Bangladesh, using a water quality model. University, Dhaka, Bangladesh, pp 1-4.
130 Water quality of Buriganga river with GIS visualization 55(2) 2020

SWMC (Surface Water Modelling Centre) (1996), Water Zhang Y, Guo F, Meng W and Wang XQ (2009), Water
Quality Modelling of the Buriganga-Lakhya River quality assessment and source identification of Daliao
System, Calibration Report. River Research Institute, river basin using multivariate statistical methods.
Government of Bangladesh. Environmental Monitoring and Assessment 152(1-4):
105.
Vega M, Pardo R, Barrado E and Deban L (1998),
Assessment of seasonal and polluting effects on the
quality of river water by exploratory data analysis.
Water Research, Vol. 32: No. 12, pp 3581-3592.