Professional Documents
Culture Documents
Assessment of Water Quality of River Netravathi Using Fuzzy Inference System
Assessment of Water Quality of River Netravathi Using Fuzzy Inference System
1
CONTENTS
• Introduction
• Aim and Objective
• Literature review
• Study area
• Methodology
• Results
• Discussion
• Conclusion
• Reference
2
INTRODUCTION
AIM
• To use Fuzzy Inference System (FIS) as a tool for classifying water quality status and perform a
comparison between FIS and the conventional weighted arithmetic WQI
OBJECTIVE
3
LITERATURE REVIEW…
• Good quality water is of vital concern for humanity since it is directly linked with human welfare. The
quality of water, surface/ground, have been worsened due to some critical factors like increasing
population, industrialization, and urbanization. Therefore, it is essential to test the quality of water
(Tyagi et al., 2020)
• The water quality evaluation requires collection and analysis of large complicated data's of
compound parameters that can be difficult to evaluate (Uddin et al., 2021)
• WQI transforms the complex information of raw water quality data into a single unitless number that
reflects the overall water quality status of the selected area (Kothari et al., 2021)
• The key parameter to be selected in a particular study is very much dependent on the significance
of that particular parameter in the given environmental conditions, and this is a point that needs to
be given much consideration while making use of the WQI method (Costa et al., 2020)
4
LITERATURE REVIEW…
• Different parameter exerts a different effect on water classification. (Vandna Pathak and A.K.
Banerjee, 2006), (Abrahão et al., 2007), (Dwivedi & Pathak, 2007), (Karakaya & Evrendilek, 2010),
(Kannel et al., 2007), (Boyacioglu, 2007) and (Pesce & Wunderlin, 2000) have assigned weights to
each parameter with values ranging from 1 to 4
• After assigning the respective values ranging from 1 to 4, (Alobaidy et al., 2010) in their study
calculated relative weight using the following equation:
• As WQI involves steps like normalising the parameters and weighting of parameters, experts
frequently disagree on the relative importance of parameters and give different weightage to same
parameter (Zagatto et al., 2017)
• Also, when water quality reports are generated, they are written for specialists, showing survey data
with emphasis on individual parameters, without providing the institutional players (managers and
decision-makers), who seldom are specialists, a bird’s eye view of the quality of a hydric resource
(Lermontov et al., 2009).
5
LITERATURE REVIEW…
• Owing to these difficulties alternate methods have emerged from artificial intelligence (AI). One such
method is the fuzzy inference system which uses fuzzy logic as a language that allows translation of
sophisticated statements from natural language into a mathematical formalism (Kumaravel et al.,
2020)
• The Fuzzy logic has been used to assess water quality based on fuzzy reasoning. The linguistic
aspect of fuzzy logic could be based on two different approaches in river water quality management:
(a) expert knowledge and or (b) actual water quality data are available in a linguistic format (Ross,
2010)
• Mamdani Fuzzy Inference System is based on the collections of IF-THEN rules with both fuzzy
antecedent and consequent parameters
• In his book (Ross, 2010) says Mamdani FIS incorporates the following main four steps: (a)
Fuzzification of Inputs and Output (b) Selection of Membership functions (c) Determination of
application rule base (d) Defuzzification to crisp value
6
• A publication by (Sastry, 2017) in the newspaper “The Hindu” said, “Netravati, the lifeline of
Dakshina Kannada, the only source of drinking water for Mangalore, gets polluted along her path by
the discharge of wastewater.” The entire sewage generated in Bantwal town (located close to the
Thumbe vented dam, built across the Netravati) is let into the river at 15 locations with no
underground drainage system. Sewage and wastewater enter Netravati and its tributary, the
Kumaradhara upstream, including Dharmasthala and Kukke Subramanya
• There is enough evidence that River Netravati is polluted due to various factors. Therefore, there is
a need for systematic water quality monitoring to assess water quality for ecosystem health and
hygiene.
7
STUDY AREA
9
METHODOLOGY...
10
Table 1: WQI Rating scale
METHODOLOGY…
WQI Value Water Quality Rating
Data collection 0-25 Excellent
• Karnataka State Pollution Control Board, 26-50 Good
51-75 Poor
Baikampady, Mangalore
76-100 Very Poor
• 3 study stations – Dharmasthala, Uppinangady, ˃100 Unfit for consumption
Reference: Tyagi, S., Sharma, B., Singh, P., &
Thumbe Dobhal, R. (2020). Water Quality Assessment in
Terms of Water Quality Index. American Journal of
• Data of year 2020 Water Resources, 1(3), 34–38.
https://doi.org/10.12691/ajwr-1-3-3
Parameters and their significance
• 30 parameters monitored by KSPCB
• pH, DO, TDS, Chloride, Sulphate, Nitrate, BOD, Total
Coliform – chosen for the study
WQI rating scale
• Rating scale developed by R.M. Brown in 1970
11
METHODOLOGY…
Fuzzy water quality index
Fuzzy logic toolbox - MATLAB R2020b package
Fuzzification: Identification of input and output and
converting it to linguistic variables
Membership Functions: Triangular Membership
Functions, Membership value in the range 0-1
Inference Rules:
• IF-THEN Conditions
• AND operator,
• If (pH is Medium) and (DO is Medium) and (TDS Figure 3: Flowchart representing the structure of a
Fuzzy Inference System
is Medium) and (Chloride is Low) and (Sulphate
is Low) and (Nitrate is Low) and (BOD is Low)
and (TC is Low) THEN (WQI is Excellent)
12
METHODOLOGY…
• If (pH is Medium) and (DO is Medium) and (TDS is Medium) and (Chloride is Low) and (Sulphate
is Low) and (Nitrate is Low) and (BOD is Low) and (TC is Medium) THEN (WQI is Good).
• If (pH is Medium) and (DO is Medium) and (TDS is Medium) and (Chloride is Low) and (Sulphate
is Low) and (Nitrate is Low) and (BOD is Medium) and (TC is High) THEN (WQI is Very Poor).
13
METHODOLOGY…
Table 2: Input parameter ranges for triangular membership function
14
METHODOLOGY…
…
Figure 4: Fuzzy Inference System Figure 5: Membership Function plot of Dissolved Oxygen
(Extracted from MATLAB R2020b)
(Extracted from MATLAB R2020b)
15
METHODOLOGY…
Figure 6: The rule editor in MATLAB R2020b Figure 7: The rule viewer in MATLAB R2020b
16
METHODOLOGY…
Weighted arithmetic water quality index
• Adopted as a reference to compare the values obtained from the Mamdani Fuzzy Inference System
Qn = × 100
Where,
Vn = Estimated value of the nth parameter
of the given sampling station.
Sn = Standard permissible value of the nth
parameter.
For pH and DO, Qn is found out using the
following equation: Qn = 100 x
Where,
Figure 8: Flowchart representing the steps involved in the Vi for pH and DO is 7 and 14, respectively
calculation of WQI
WQI =
Where Wn is factor weight
17
WEIGHTED ARITHMATIC WATER QUALITY INDEX
Factor weights: Scale 1-4.
Standard permissible limit: ICMR, BIS
18
RESULTS…
Table 6: FWQI AND WAWQI results of Dharmasthala
DHARMASTHALA
UPPINANGADY
Month FWQI Water Quality WAWQI Water Quality
Status Status
January 40.6 Good 41.7836 Good
February 40.1 Good 47 Good
March 41.8 Good 37.2928 Good
April 40.6 Good 47.1522 Good
May 14.1 Excellent 28.3104 Good
June 15 Excellent 24.7017 Excellent
July 15 Excellent 28.1076 Good
August 14.1 Excellent 29.3534 Good
September 96.7 Very Poor 40.2304 Good
October 96.7 Very Poor 95.1595 Very Poor
November 96.7 Very Poor 94.2159 Very Poor
December 39.7 Good 97.3602 Very Poor
Average 45.925 Good 50.8890 Good
20
RESULTS…
Table 8: FWQI AND WAWQI results of Thumbe
THUMBE
Month FWQI Water Quality Very Poor
94.5789
Status
January 40.6 Good 38.6947 Good
February 40.1 Good 40.3373 Good
March 41.8 Good 41.6898 Good
April 40.6 Good 24.8374 Excellent
May 14.1 Excellent 25.6526 Excellent
June 15 Excellent 24.9974 Excellent
July 15 Excellent 24.9205 Excellent
August 14.1 Excellent 92.7777 Very Poor
September 96.7 Very Poor 90.5507 Very Poor
October 96.7 Very Poor 92.1412 Very Poor
November 96.7 Very Poor 94.0321 Very Poor
December 39.7 Good 57.1009 Poor
Average 45.925 Good 94.5789 Very Poor
21
RESULTS
Variation of WQI at Uppinangady
Variation of WQI at Dharmasthala 120
120
100
100
80
80
60
60
40
40
20
20
0
0 ry ry ch ril ay ne Ju
ly st be
r er be
r
be
r
ry ry ch ril ay ne ly st r er r r ua ua ar Ap M Ju gu ob
ua ua ar Ap Ju gu be ob be be n
eb
r M u m t m m
n br
M Ju u m t m m J a A te Oc ve ce
J a M A pt
e c ve ce F ep No De
Fe O
No De
S
Se
FWQI WAWQI
FWQI WAWQI
FWQI WAWQI
• Selection of parameters in conventional methods - E.g. TC and its permissible limit for drinking
water (Zero) cannot be used in the equation : Qn = × 100
• ‘Good’ and ‘Very Poor’- TC, Post monsoon (rainfall runoff and agricultural runoff), Festivals,
Vacation, Soak pits and septic tanks
• Only a few studies have been conducted on River Netravati to assess the water quality and its
status. Ramananda, 2017 in his study has used Statistical sampling theory to study the quality of
River Netravati. Lekeshmanaswamy & Pooja, 2019 has done the evaluation of physico chemical
parameters on water quality of River Netravati. Avvannavar & Shrihari, 2008 has evaluated water
quality of Netravati using Bhargava Water Quality Index Method. Gayathri et al., 2021 analyzed
the water quality of Netravati using National sanitation foundation water quality index
• The present study - different yet competent approach, less complex and straightforward
• No single systematic approach. The rules have to be updated with changing inputs
• Results may vary different membership functions (e.g., Trapezoidal, Gaussian membership functions)
and defuzzification methods (e.g., Center of sums, Center of largest area)
CONCLUSION…
• Water quality of Dharmasthala and Uppinangady is ‘Good,’ and that of Thumbe is Poor and are not
suitable for drinking 25
CONCLUSION
• Results of FWQI are almost consistent with that of the conventional WAWQI
• Analysis suggests improving the sewage treatment rate, controlling the pollution from agricultural
sources
• Therefore, it can be used as a substitute for conventional methods for the assessment of surface
water quality
• Fuzzy logic concepts, if used rationally, could be a potent tool for some of the environmental policy
subjects 26
LIMITATIONS OF THE STUDY
• Testing of samples should be performed on the same day. When the testing is prolonged, changes
occur in the collected water and misdiagnosing can happen
• Study cannot be entirely replicated for a study with parameters of different value
• The results of this study can be used as a reference when performing water quality analysis in
similar environmental conditions.
FUTURE WORK…
• Adaptive Network based Fuzzy Inference System (ANFIS) as it can construct an input – output
mapping
• Mean absolute error (MAE), Mean square error (MSE), and Root mean square error (RMSE)
27
FUTURE WORK
• FIS method with other methods for more clarity in its efficiency in assessing the water quality.
REFERENCE…
• Costa, D. D. A., Paulo, J., Azevedo, S. De, & Aurélio, M. (2020). Water quality assessment based on
multivariate statistics and water quality index of a strategic river in the Brazilian Atlantic Forest.
Scientific Reports, 1–13. https://doi.org/10.1038/s41598-020-78563-0
• Deka, P. C. (2020). A Primer On Machine Learning Applications In Civil Engineering. Taylor and Francis
Group
• Kothari, V., Vij, S., Sharma, S. K., & Gupta, N. (2021). Correlation of various water quality parameters
and water quality index of districts of Uttarakhand. Environmental and Sustainability Indicators,
9(December 2020), 100093. https://doi.org/10.1016/j.indic.2020.100093
• Ross, T. J., & Ross, T. J. (2010). Fuzzy logic with Engineering, Third edition. 28
REFERENCE
• Tyagi, S., Sharma, B., Singh, P., & Dobhal, R. (2020). Water Quality Assessment in Terms of Water
Quality Index. American Journal of Water Resources, 1(3), 34–38. https://doi.org/10.12691/ajwr-1-
3-3
• Uddin, M. G., Nash, S., & Olbert, A. I. (2021). A review of water quality index models and their use
for assessing surface water quality. Ecological Indicators, 122, 107218.
https://doi.org/10.1016/j.ecolind.2020.107218
• Vandna Pathak and A.K. Banerjee. (2006). Mine water pollution studies in chapha incline, Umaria
coalfield, Eastern Madhya Pradesh, India. 11(2), 27–35.
29