2.

1 IMPORTANCE OF SURFACE WATER BODIES

According to the EPA Report on Environment (2022) less than one thousandth
of one percent of the water on the globe is contained in lakes, ponds, rivers, and
streams, but they nonetheless perform many essential tasks for the ecosystem and
human life. These fresh surface waters support ecological processes and offer habitat
to numerous plant and animal species. Additionally, they support a wide range of
human needs, such as drinking water, irrigation, wastewater treatment, livestock,
industrial usage, hydropower, and leisure. Along with ground water, wetlands, and
coastal systems, fresh surface waters have an impact on the quantity and quality of
other water resources.

There are two major groups of indicators that can be used to track the condition of
fresh surface waters: extent indicators and condition indicators.

Fresh surface water extent essentially correlates to the volume of water and reflects a
number of anthropogenic and environmental stressors. It can be impacted by direct
human withdrawals for drinking, irrigation, industrial operations, and other uses, as
well as by ground water withdrawals that replenish many surface waterways.

Hydromodifications can produce new impoundments and significantly modify stream

flow, such as dam construction. Drainage patterns may be impacted by land cover
(impervious pavement, for example, may enhance runoff or flooding). The quantity of
fresh surface waters is also influenced by weather patterns, such as precipitation
levels, snowmelt timing, and evaporation rate conditions. The amount of available
fresh surface water may also change due to a changing climate.

Fresh surface water's condition roughly correlates to its quality and reflects a variety
of physical, chemical, and biological traits. The term "physical characteristics" refers
to things like temperature and clarity. Chemical properties include elements like
salinity, nutrients, and chemical contaminants (such as toxins in sediments, which can
affect water quality and potentially penetrate the aquatic food chain). Pathogens,
cyanobacteria, and, more broadly, the status of plant and animal populations and the
state of their habitats are examples of biological qualities (which can be modified by
chemical features).

Dodds et al (2002) state that, for human welfare and environmental stability, surface
water bodies provide crucial ecosystem functions. They serve as organic water filters
that maintain water quality by eliminating contaminants and sediments from the
water.

Nijssen et al (2001) emphasise that surface water bodies regulate temperature

extremes and have an impact on humidity levels, which affects weather patterns and
local vegetation development.

Surface water bodies serve ecological purposes, but they also have cultural and
recreational value. They have influenced human civilizations for ages, providing
motivation, a means of subsistence, and the foundation for cultural traditions.
According to UNESCO (2019), rivers, lakes, and coastal regions offer chances for
outdoor recreation like swimming, boating, fishing, and wildlife observation, which
boosts local businesses and tourism.

2.2 MODELLING CONCEPTS

The use of mathematical models is widely acknowledged as a powerful tool
for analysing the economic, environmental, and ecological effects of alternative
pollution-control and resource-conservation measures, efficient management
practises. Therefore, novel mathematical methods that can reflect different
combinations of environmental complexities are sought to enable more effective
management and planning of resources and environmental systems.

Nirmalakahandan (2002) defines modelling as the process of application of

fundamental knowledge or experience to simulate or describe the performance of a
real system to achieve certain goals. Models can be cost-effective and efficient tools
whenever it is more feasible to work with a substitute than with the real, often
complex systems. Modelling has long been an integral component in organizing,
synthesizing, and rationalizing observations of and measurements from real systems
and understanding their causes and effects.

Mathematical models can be classified as deterministic (where input and output

variables have fixed values), stochastic (where at least one input or output variable
has a probability), static (where time is not taken into account), or dynamic (where
time-varying interactions among variables are taken into account). Simulation models
are typically stochastic and dynamic.
Sam (2004) describes that an "object" can be a physical item having particular
properties or traits, and a "system" can be conceived of as a grouping of one or more
linked objects. In order for the resulting model to meet the modeler's objectives, a
system must allow the modeller to specify its borders, properties, and interactions
with its surroundings. Frequently, the complexity of the model increases with system
size. When a system does not interact with its surroundings, it is referred to as a
closed system; when it does, it is referred to as an open system. In contrast to open
systems, neither energy nor mass will cross the boundary in a closed system.

Variables are characteristics of the system and its environment that significantly affect
the system. The term "variable" refers to both characteristics whose values fluctuate
and those whose values don't change over the course of the modelling period. The
latter kind of variables are frequently referred to as "parameters." There are two types
of variables that change in value: those that are produced by the environment and
affect how the system behaves, and those that are produced by the system and have an
impact on the environment. The former are referred to as "inputs," and the latter are
referred to as "outputs." In mathematical models, inputs are referred to as independent
variables, and outputs are referred to as dependent variables. However, all inputs and
parameters of mathematical models are easily controllable or manipulated, and all
outputs are accessible.

2.2.1 Model Development

Voinov and Shugart (2018) opined that in order to simulate and comprehend
complicated environmental processes, environmental mathematical models are
created through a methodical process that combines scientific data, mathematical
tools, and knowledge. Using knowledge from various disciplines, including biology,
chemistry, environmental science, mathematics, physics, and more, researchers use a
multidisciplinary approach to their work. These simulations are designed to capture
interactions between diverse environmental elements, such as soil, water, and
biological activities, and to forecast how they will behave in particular scenarios.

Typically, establishing the goals and parameters of the model is the first step in
constructing an environmental mathematical model. The precise environmental
processes and phenomena that researchers want to imitate and comprehend are
identified. An in-depth knowledge of the underlying science and the system under
investigation is necessary for this step. The selection of suitable mathematical
equations and algorithms is aided by having clear objectives state Hodges and White
(2021).

The gathering of data is an essential stage in creating mathematical models of the

environment. To quantify the pertinent variables and parameters within the model,
researchers gather data through field observations, laboratory experiments, or pre-
existing datasets. The model can be calibrated and validated using this data,
confirming its accuracy and dependability note Wheater and Croke (2006).

Hodges and White (2021) explain that, following the data collection, researchers
create the equations and algorithms that represent the environmental processes in the
model using statistical and mathematical methods. These equations could be based on
underlying physical principles, empirical relationships, or a combination of the two.
The level of detail necessary to accurately reflect the dynamics of the system
determines how complicated the model will be.

The steps of model validation and calibration are crucial in the development process.
To guarantee that the model accurately captures observations and behaviours from the
real world, researchers modify the model's parameters. In order to evaluate a model's
effectiveness and dependability in forecasting system dynamics, validation entails
comparing model outputs with independent data sets affirm Voinov and Shugart
(2018).

Dietz (2015) stress that environmental mathematical model creation is an iterative

process that involves improvement over time. As fresh data and scientific
understanding become available, models are regularly updated. To guarantee that the
models are fit for purpose and successfully support environmental decision-making,
ongoing collaboration between academics, stakeholders, and policymakers is
essential.

Zeidan (2015) describe that a simulation of a system is the operation of a model of the
system. The model may be modified, and experimented with; typically, in the system
it symbolises, this is impossible, prohibitively expensive, or unfeasible to achieve.
Since the model's operation can be examined, properties relating to the behaviour of
the real system or a subsystem of it can be deduced. In its broadest definition,
simulation is a technique for assessing how well an existing or prospective system
will function under various configurations of interest over extensive periods of real
time. Prior to making changes to a current system or creating a new one, simulation is
used to lower the likelihood of failing to meet specifications, remove unforeseen
bottlenecks, prevent resource under- or over-utilization, and to optimize system
performance. Figure 1 shows the procedures involved in creating a simulation model,
planning a simulation experiment, and conducting a simulation analysis.

Step 1. Identify the problem.

Step 2. Formulate the problem.

Step 3. Collect and process real system data.

Step 4. Formulate and develop a model.

Step 5. Validate the model.

Step 6. Document model for future use.

Step 7. Select appropriate experimental design.

Step 8. Establish experimental conditions for runs.

Step 9. Perform simulation runs.

Step 10. Interpret and present results.

Step 11. Recommend further course of action.

Fig 2.1: Modelling Process Flow Chart

2.2.2 Use of Mathematical Models in Prediction

Fath et al (2007) assert that, environmental mathematical models are useful

instruments for foreseeing and predicting a range of environmental system features.
These models imitate the behaviour of complicated environmental systems using
mathematical equations and algorithms, enabling researchers and decision-makers to
forecast future circumstances and outcomes. These models can offer important
insights into environmental trends, patterns, and prospective situations by combining
scientific information and data.

Predicting changes in environmental variables over time is one of the main

applications of environmental mathematical models in prediction. These models may
predict how these variables may change under various scenarios and replicate the
dynamics of natural systems, such as climate patterns, water availability, and
ecological processes state Chang et al (2014). These models help researchers and
policymakers predict future circumstances and create appropriate plans by taking into
account variables like climate change, changes in land use, and pollution levels.

Dietz (2015) note that, environmental mathematical models are essential for
foretelling how human activity will affect the environment. The possible effects of
various activities and interventions, such as the development of infrastructure
projects, the deployment of pollution control strategies, or modifications to land
management practises, can be evaluated using these models. These models can aid in
estimating the environmental impacts and directing decision-making processes by
incorporating data on pollutant emissions, population growth, and economic
development.

The forecasting of natural risks and hazards is another important use of environmental
mathematical models. Researchers and authorities can analyse potential dangers and
create mitigation plans by using models that imitate the behaviour of natural
phenomena including floods, hurricanes, wildfires, and disease outbreaks emphasise
Schröter et al (2005). These models produce projections that guide emergency
planning and response operations by taking into account variables including
geography, weather patterns, population density, and susceptibility indicators.

Environmental mathematical models are essential for predicting species distributions,

population dynamics, and interactions within communities in the area of ecology.
Researchers can examine the possible effects on biodiversity and ecosystem
functioning by using these models to predict how species respond to environmental
changes such habitat loss, pollution, or climate fluctuations remark Thuiller (2007).
These models aid in the planning and administration of conservation activities by
forecasting the distribution and abundance of species in the future.

Environmental mathematical models are also used to forecast how contaminants

would behave and move via different environmental mediums. In order to measure air
quality and determine the efficacy of emission reduction methods, they can model the
dispersion of air pollutants including particulate matter and greenhouse gases claim
Brunner et al (2012). These models also help identify pollution sources and design
pollution control strategies by forecasting the fate and transport of contaminants in
rivers, lakes, and seas in aquatic systems.

Abbaspour (2014) describe that, environmental mathematical models are used to

forecast the availability and quality of water resources in the context of natural
resource management. The assessment of water availability under various climatic
and land-use scenarios is made possible by these models, which simulate hydrological
processes such rainfall-runoff relationships, groundwater recharge, and streamflow
dynamics. Planning irrigation systems, managing water resources, and allocating
water sustainably all depend on these forecasts.

The spread and effects of infectious diseases are also predicted using environmental
mathematical models. Incorporating variables like population demographics, contact
rates, and illness parameters, these models recreate the dynamics of disease
transmission state Ferguson et al (2007). These models support public health
planning, resource allocation, and the creation of efficient control methods by
producing forecasts about disease outbreaks, their spatial spread, and viable
interventions.

Environmental mathematical models have a wide range of applications in prediction,

from pollutant dispersion to disease outbreak forecasting, from species distribution
modelling to climate change projections. With the help of these models, decision-
makers and academics may predict future environmental conditions, assess potential
dangers, and create sustainable long-term policies.

Environmenta Issues/ concerns Use of models

l
media

Atmosphere Hazardous air pollutants, Concentration profiles; exposure;

air emissions, toxic design and analysis of control
releases, acid rain; processes and equipment;
particulates, smog, CFCs, evaluation
health concerns of management actions;
environmental
impact assessment of new projects;
compliance with regulations
Surface water Wastewater treatment plant Fate and transport of pollutants;
discharge; industrial concentration plumes; design and
 discharges; analysis of control processes and
agricultural/urban equipment; waste load allocations;
runoff; storm water evaluation of management actions;
discharge; environmental impact assessment
potable water source; food of
chain new projects; compliance with
regulations
Groundwater Leaking underground Fate and transport of pollutants;
storage design and analysis of remedial
tanks; leachates from actions; drawdowns; compliance
landfills with
and agriculture; injection; regulations
potable water source
Subsurface Land application of solid Fate and transport of pollutants;
and concentration plumes; design and
hazardous wastes; spills; analysis of control processes;
leachates from landfills; evaluation of management actions
contamination of potable
aquifers
Ocean Sludge disposal; spills; Fate and transport of pollutants;
outfalls; food chain concentration plumes; design and
analysis of control processes;
evaluation of management actions
Table 2.1: Typical uses of Mathematical Models

The main purpose of models maybe grouped under three categories:

 Descriptive: To integrate observations, information and theories concerning a

system; to aid understanding of system behaviour.
 Predictive: To predict the response of the system to future changes.
 Allocation: To allocate certain resources in order to optimize certain
conditions within the system.

2.2.3 Available Water Quality Models

2.2.3.1 CORMIX

CORMIX is a USEPA-supported methodology for simulation of turbulent

buoyant jet mixing behaviour which covers a majority of common discharge and
environmental conditions. It classifies momentum and buoyancy of the discharge in
relation to boundary interactions to accurately predict mixing behaviour. Boundary
interactions can be flow surface or bottom contact or terminal layer formation in
density stratified ambients. This trapping behaviour can sometimes be ecologically
desirable and attainable with appropriate outfall design.

The hydrodynamic simulation system contains a collection of regional flow models

based upon integral, length scale, and passive diffusion approaches to simulate the
hydrodynamics of near-field and far-field mixing zones. Efficient computational
algorithms provide simulation results in seconds for mixing zone problems with space
scales of meters to kilometers and time scales of seconds to hours.

CORMIX can predict mixing behaviour from diverse discharge types ranging from
power plant cooling waters, desalinization facility or drilling rig brines, municipal
wastewater, or thermal atmospheric plumes. CORMIX can also be applied across a
broad range of ambient conditions ranging from estuaries, deep oceans, swift shallow
rivers, to density stratified reservoirs and lakes.

The study by Purnama (2012) The potential impacts of surface discharges of heated
brine from the Al-Ghubrah power-desalination plant in the Omani coastal marine
environment is investigated using CORMIX. Simulation results show that the overall
temperature rise due to brine discharges from the Al-Ghubrah plant is between 1.7 to
5.5oC (above ambient) at the end of the regulatory mixing zone of 150 m radius from
the discharge channel, well above the maximum permissible limit of 1 oC (above
ambient) set by the Omani government.

Like other numerical models, CORMIX system has several inherent limitations. One
major limitation is the representation of the coastal current velocity which is assumed
to be uniform. Another limitation is the flow classification system based on
hydrodynamic criteria using significant length scale analysis and its subsequent
dilution in the receiving water environment.

2.2.3.3 MULTMIX

A river water quality modelling system called MULTMIX makes it possible to

simulate and examine river water quality situations. It is a specialist piece of software
made for researching how pollutants, fertilizers, and other elements move through
river systems.

To simulate the movement and fate of contaminants in rivers, the MULTMIX

software uses mathematical models. Flow rates, pollutant loads, water temperature,
and biochemical processes are just a few of the variables that these models take into
account. The software can forecast the distribution and transformation of water
quality indices across the network of rivers by considering these variables.

Users of MULTMIX can enter information on the sources of pollutants, the features
of river flow, and the initial state of water quality. The software then models how
pollutants move and change over time, taking into account phenomena including
advection, dispersion, reaction kinetics, and sediment interactions.

The evaluation of the effects of point-source discharges, non-point-source pollution,

and other pollutant inputs on water quality is one of the main uses of MULTMIX. The
software can assess the effectiveness of pollution control techniques and management
strategies aimed at preserving or enhancing river water quality by simulating various
scenarios.

The ability of MULTMIX to shed light on the geographical and temporal fluctuations
of water quality metrics is one of its strongest points. Planning for water resources,
managing rivers, and combating pollution all benefit from this knowledge.
Researchers and environmental specialists can better understand the intricate
relationships between pollutant sources, river hydraulics, and water quality dynamics
thanks to MULTMIX. This information makes it easier to create plans that will
effectively maintain or restore the ecological health of river systems.

It should be noted that the term "MULTMIX" could refer to a specific river water
quality modelling system created by a particular company or research team. As a
result, depending on the model's version or implementation, the unique features,
capabilities, and applications of MULTMIX may vary.

2.2.3.3 QUAL2K

QUAL2K is a widely used water quality modelling software developed for

assessing and predicting water quality in rivers and streams. It is an acronym for
"Quality and Quantity in the 2D Kinematic Equations" model. The software
incorporates mathematical equations and algorithms to simulate the movement and
transformation of various water quality constituents in river systems.
The QUAL2K model takes into account a range of factors that influence water
quality, including flow rates, pollutant sources, biochemical processes, temperature,
sediment interactions, and hydrodynamic conditions. It allows users to input data on
pollutant loads, hydrological characteristics, and initial water quality conditions to
simulate and analyse the water quality dynamics over time.

The main purpose of QUAL2K is to aid in understanding and managing water

pollution issues. It can be used to assess the impacts of point and non-point source
pollution, evaluate the effectiveness of pollution control measures, and develop
strategies for improving and maintaining water quality. By simulating various
scenarios, QUAL2K helps researchers, water resource managers, and environmental
professionals to make informed decisions and develop effective strategies for water
quality management.

It is worth noting that QUAL2K is a specific water quality modelling software

developed by researchers and scientists. Its features, capabilities, and applications
may vary based on different versions and implementations of the model.

2.2.3.2 MIXPIPOX

MIXPIPOX (Stream Tube Model of Oxygen Prediction for a Pipe Outfall). The
MIXIPIPOX programme gives the lateral and longitudinal distributions of CBOD,
NOD and DO in river channel receiving effluents from pipe outfalls located at bank or
in river channel (vertical line source). The data used to drive this model should be in
following units:

 Distance in m
 Flow rates in m³/s
 Velocity in m/s
 Temperature in Celsius degree
 Decay rates and reaeration rate expressed per sec (base e)
 Concentration in mg/l or g/l.

The following information is necessary to run the programme

 1. Distance downstream from outfall in meters. Average depth, velocity and
width of the river channel at each transect, number of lateral points at each
transect.
2. The contaminant being measured.
3. Flow rate in the river just upstream of the outfall.
4. Background concentrations of the contaminants and DO.
5. Effluent concentrations of the contaminants and DO.
6. The decay rates of the contaminant and the temperature at which these rates
are applied.
7. Reaeration rate co-efficient.
8. Temperature of the river water.
9. Saturation concentration of DO in mg/l.

The above-mentioned data are the various input parameters of the MIXPIPOX
programme except for the location of the pipe outfall at bank or in the river (defined
by the partial cumulative discharge, q e). Further, the program MIXPIPOX calculates
the 1-D values, which can be used for comparison between the 1-D model and 2-D
model regarding the DO distribution in mixing zones.

2.2.4. Statistical Methods For River Hydraulic Data Analysis

In this section, we delve into several notable articles that make significant
contributions to the statistical analysis of river hydraulic data. Each article offers
unique insights, methodologies, and approaches that advance the understanding and
interpretation of river behaviour.

The article by Serinaldi et al (2014) emphasizes the application of copulas, a

statistical tool, for analysing river hydraulic data. Copulas prove valuable in
characterizing the dependence structure between variables such as river flow and
water level. By utilizing copulas, researchers gain insights into the joint behaviour of
these variables, thereby enhancing their understanding of the dynamics governing
rivers.

Vogel and Fennessey (1996) address the challenges and opportunities associated with
the statistical analysis of river flow data. They shed light on the complexities inherent
in hydrological data, including issues such as data heterogeneity, non-stationarity, and
missing data. Through an overview of various statistical techniques such as frequency
analysis, regression modelling, and time series analysis, the authors emphasize the
importance of accounting for uncertainties and limitations in hydrological data
analysis. This encourages researchers to interpret their findings carefully and make
informed decisions.

Neill and Cox (2012) introduce Bayesian hierarchical modelling as an effective

statistical approach for analysing river hydraulic data. This modelling framework
allows for the incorporation of prior information and addresses the complex
hierarchical structures present in the analysis. By utilizing Bayesian hierarchical
models, researchers can achieve improved inference, parameter estimation, and
uncertainty quantification. The authors discuss the practical implementation of this
approach within the context of river hydraulics and highlight its potential for
addressing challenges associated with the analysis of complex hydraulic systems.

Li et al (2017) focusses their investigation on the statistical analysis of river

hydraulics specifically for floodplain delineation. They explore the relationships
between various hydraulic parameters, such as river flow velocity, channel geometry,
and floodplain characteristics. This analysis enhances the understanding of floodplain
dynamics, contributing to improved flood risk assessment and management. By
elucidating the behaviour and interactions between river hydraulics and floodplains,
this research aids in developing effective strategies for flood mitigation and planning.

Zhang et al (2010) concentrate on the spatial and temporal statistical analysis of river
hydraulic data for flood frequency estimation. Their study incorporates various
factors, including river geometry, rainfall patterns, and land use changes, into the
statistical analysis. By incorporating spatial and temporal information, the accuracy
and reliability of flood frequency estimation are significantly improved. These
insights are crucial for enhancing flood risk management practices and informing
infrastructure design. Collectively, these articles provide researchers with valuable
insights, methodologies, and approaches for statistically analysing river hydraulic
data. By addressing specific challenges and opportunities within the field, offering
practical applications of statistical techniques, and contributing to the overall
understanding of river behaviour, they have significant implications for various
domains, including flood risk assessment, floodplain management, and infrastructure
planning. The findings presented in these articles contribute to the advancement of
knowledge and serve as a foundation for further research in the field of river hydraulic
data analysis.

2.2.5. Applications Of Mathematical Models For River Hydraulic Data Analysis

In the article by Wang et al (2019), the authors provide an extensive review of

mathematical models used for analysing river hydraulic data. They discuss both
numerical models, which simulate the physical processes of rivers using mathematical
equations, and statistical models, which analyse the statistical relationships between
variables in the data. The authors highlight the wide range of applications of these
models, including flood forecasting, sediment transport modelling, and river system
optimization. They delve into the strengths and limitations of each type of model,
considering factors such as data availability, model complexity, and computational
requirements. This comprehensive review serves as a valuable resource for
researchers and practitioners interested in utilizing mathematical models for river
hydraulic data analysis, guiding them in selecting the most appropriate models for
their specific needs.

Smith et al (2016) focus on the role of mathematical models in flood risk assessment
through the analysis of river hydraulic data. They emphasize the importance of
accurate and reliable hydraulic data analysis for developing effective flood risk
management strategies. The authors discuss different modelling approaches, such as
hydraulic models, which simulate the flow of water through a river system, and
hydrological models, which analyses the processes that govern the inflow and outflow
of water in a river basin. They highlight the challenges associated with modelling
river hydraulic processes, including uncertainties in input data, model calibration, and
validation. By discussing the strengths and limitations of these models, the article
provides insights into the selection and application of mathematical models in flood
risk assessment.

Garcia et al (2014) present a case study that compares various mathematical models
used in river hydraulic data analysis. They evaluate the performance of different
models, including one-dimensional and two-dimensional hydraulic models, in
capturing the behaviour of river systems. One- dimensional models simplify the
representation of rivers as lines, whereas two-dimensional models account for the
spatial variation of flow. The authors discuss the advantages and limitations of each
model in terms of their ability to represent flow patterns, water levels, and sediment
transport processes. They also consider computational efficiency and data
requirements. By providing a comparative analysis of different models, the case study
contributes to the understanding of the suitability and applicability of mathematical
models for river hydraulic data analysis in practical scenarios.

Kim et al (2018) explores the integration of mathematical models in river hydraulic

data analysis within the context of watershed management. They emphasize the
importance of considering the interactions between river systems and the surrounding
watersheds in understanding river behaviour. The authors discuss the integration of
hydrological models, which simulate the processes of rainfall, runoff, and
groundwater flow, with hydraulic models that analyse the flow of water in rivers.
They also highlight the incorporation of water quality models to study the variations
in water quality parameters. This integrated approach enables a holistic understanding
of the complex dynamics of rivers and watersheds. The article showcases the practical
applications and advantages of integrating mathematical models in river hydraulic
data analysis for watershed management, supporting effective decision-making in
water resources planning and management.

Nguyen et al (2017) present a case study focusing on the application of mathematical

modelling approaches for river hydraulic data analysis, specifically for flow
forecasting. They investigate different modelling techniques, such as time series
analysis and machine learning methods, to predict river flows. Time series analysis
involves analysing the historical patterns and trends in the data to make future
predictions, while machine learning methods utilize algorithms to identify complex
relationships and patterns in the data. The authors evaluate the performance of these
models in terms of accuracy, robustness, and computational efficiency. They discuss
the strengths and limitations of each model and provide recommendations for
selecting an appropriate modelling approach based on data availability, computational
resources, and forecasting requirements. This research contributes to improving the
accuracy and reliability of flow forecasting, which is crucial for various applications
such as water resources planning, flood management, and hydropower generation.

In summary, these articles contribute to the advancement of knowledge in the field of

river hydraulic data analysis, supporting researchers and practitioners in making
informed decisions and implementing effective strategies for river management, flood
risk assessment, and water resources planning.

2.2.6 Dissolved Oxygen Modelling in Rivers

Dissolved Oxygen (DO) is an important water quality parameter in accordance
with health, therefore given major importance in maintaining DO in desirable level.

Dissolved oxygen is important for aquatic life because when DO drops below 4mg/L
or 5mg/L, it causes detrimental effect and can turn out to be dangerous for the health
of aquatic life based on species. Suspended solids influence the water column
turbidity and tend to settle at the bottom, leading to benthic enrichment, toxicity and
sediment oxygen demand. Nutrients can cause DO depletion and eutrophication.
Thus, in order to analyse the effect of wastewater discharge on DO levels of riverine
systems, it is necessary to understand the correlation between pollutant characteristics
and the water environment.

The important factors (rate constants) and processes affecting DO discussed in

following sections could be listed as follows:

1. Carbonaceous and Nitrogenous BOD Decay Rate

2. Atmospheric Reaeration

3. Photosynthesis and Respiration

4. Sediment Oxygen Demand (SOD)

5. Dispersion Coefficient.

2.4.1 Sources and Sinks of Dissolved Oxygen (DO)

DO level in streams and rivers are affected by several sources and sinks. The
important sources for DO level in water bodies are: 1) Rearation from atmosphere,

2) Photosynthetic oxygen production and 3) Input from tributaries and effluent that
increases DO level.

The major sinks in DO are: 1) Oxidation of carbonaceous and nitrogenous waste

materials given by CBOD and NBOD, 2) Sediment Oxygen Demand (SOD) and
3) Oxygen utilization by aquatic plants and algae for respiration. The general mass
balance equation for DO concentration C, in a segment volume of V can be written as:

V (dc/dt) = [reaeration+ (photosynthesis- respiration) + DO inputs - CBOD &

NBOD Oxidation - SOD + DO transport into or out of segment]

2.2.6.1 Carbonaceous Deoxygenation

Biochemical Oxygen Demand (BOD) is the utilization of dissolved oxygen by the

microbes present in aquatic forms to metabolize organic matter, oxidation of reduced
nitrogen and also in the oxidation of mineral species of reduced forms like ferrous
iron. Concentrations of reduced form of minerals are usually trivial, henceforth

BOD is commonly divided into two fractions: Carbonaceous matter and Nitrogenous
matter. Wastewater are high in BOD and that dissolved oxygen concentration is
considered as principal parameter determining the health of aquatic system, BOD is
applied widely as measure of aquatic pollution

The Carbonaceous Organic Matter found in wastewater and in rivers act as source of
energy for heterotrophic micro-organisms. In this process, oxygen is utilized by
micro-organisms to decompose the organic matter found in wastewater, the amount of
oxygen therefore utilized in the process of decomposition is known as Carbonaceous
Biochemical Oxygen Demand (CBOD).

2.2.6.2 Nitrogenous Deoxygenation

The first stage CBOD is often headed by a second stage that comprises of oxidation of
nitrogenous compounds that could possibly found in riverine system or wastewater.
This is called as Nitrogenous BOD or BOD. It could also be defined as the oxygen
devoured during the oxidation of nitrogenous compounds (particularly NHs) to nitrate
with nitrite being the unstable intermediary. NBOD involves two classes of bacteria,
which are assumed to be responsible for oxidation of reduced nitrogen.

These bacteria (nitrifiers) are usually present in low concentration of water, as their
surface based (suspended solids). BOD could be observed only after 7 days of
incubation has occurred. This is majorly because, growth rate of nitrifiers are slow
and cannot be observed till the food supply for the CBOD consuming heterotrophic
microbes are depleted i.e. when BOD liberated reaches the ultimate CBOD to zero.

2.2.6.3 Atmospheric Reaeration

Reaeration is the process involving exchange of oxygen between the atmosphere and
water body in contact with atmosphere. This occurs at air-water interface if a non-
equilibrium condition exists for the oxygen between the air-water phases. Usually, the
transfer of oxygen is from atmosphere to the water body as the DO levels are below
saturation in natural body. However, when photosynthesis takes place, it produces
supersaturated DO levels thereby, transferring the oxygen back to the atmosphere.

2.2.6.4 Photosynthesis and Respiration

Photosynthesis forms a major source of oxygen for natural water body wherein,
respiration from a major sink. This is more likely to be notable for algal
concentrations greater than 10 g/m'. Measurement of oxygen fluxes related to
photosynthesis and respiration is arduous, and is dependent on factors like
temperature, nutrient concentration, turbidity and sunlight. They also depend on the
aquatic plants whether found floating (phytoplankton) or situated at the bottom
(macrophytes, periphyton).

2.2.6.5 Sediment Oxygen Demand

The discharge of deposable waste constituents usually lead to the formation of "sludge
banks" or deposits of organic material just below the waste discharge point.

As time pass by, these deposits grow if velocity of the water body is too low. As the
depth of these accumulated solids increases, initiation for anaerobic decomposition of
the organic material in depth takes place. Decomposition leads to the formation of
CO2, CH4 and H2S which move up through the sludge layer and into the water lying
above it. Floating of the bottom sludge occur, if the gases produced during
decomposition is high, leading to extreme aesthetic problem and DO depletion.

Many studies showed that, oxygen demand by the underlying sediments is the major
source of water column oxygen depletion. Benthic deposits take their origin from
wastewater effluents, surface runoff and at times the prevailing aquatic conditions in
the natural water body. Alarming levels of oxygen i.e., below 2-3 mg/L in aquatic
body can cause dead zones, eventually leading to the death of flora and fauna. The
rate in which water column oxygen is discharged during the disintegration of organic
matter and respiration of organisms present in streams, lake or river bed deposits is
known as Sediment Oxygen Demand (SOD) forming a major oxygen sink of DO.

2.2.6.6 Dispersion Co-efficient

The interaction of turbulent diffusion with velocity gradients caused by shear forces in
the riverine system leads to a greater degree of mixing known as dispersion.
Movement of toxic materials in streams and rivers is mainly because of advection, but
their transport in lakes and estuaries is controlled by dispersion (Jerald Schnoor,1996).

2.2.7 Models for Prediction of Dissolved Oxygen

Ravansalar et. al (2015) present an innovative approach to improve the prediction of
DO in the river Calder. The study focuses on noise elimination through time series
analysis using wavelet transform. By applying the wavelet transform technique, the
authors effectively remove noise from the DO data, leading to more accurate
predictions. The paper utilizes support vector machine (SVM) regression along with
feature selection methods and optimization algorithms to enhance the prediction
model. The results demonstrate the effectiveness of the proposed approach, as
indicated by the evaluation metrics, including Root Mean Squared Error (RMSE),
Mean Squared Error (MSE), Mean Absolute Error (MAE), and R Squared (R2)

A new ensemble method combines the strengths of multiple regression models to

enhance prediction accuracy. By incorporating the advantages of individual models,
the ensemble method achieves superior performance compared to standalone models.
The paper utilizes various evaluation metrics, including RMSE, MAE, and R2, to
assess the effectiveness of the proposed approach. The results highlight the improved
predictive capabilities of the ensemble method for DO prediction. as described in Kisi
et al (2018).

Cox et al. (2017) provides a comprehensive analysis of DO modelling techniques

specifically tailored for lowland rivers. The paper explores the challenges associated
with DO modelling in these environments and evaluates various modelling
approaches employed in the literature. The paper covers both empirical and
mechanistic modelling techniques, discussing their advantages, limitations, and
applicability to lowland river scenarios. Cox's review serves as a valuable resource for
researchers and practitioners working in the field of river water quality modelling,
aiding in the development of more effective DO prediction models.

A study by Sahu et al. (2022) provides an overview of data-driven modelling

approaches for predicting DO and biochemical oxygen demand (BOD). The paper
discusses the importance of accurately estimating these water quality parameters for

effective management of aquatic ecosystems. Sahu et al. evaluate various data-driven

techniques, such as artificial neural networks, support vector machines, and genetic
programming, for modelling DO and BOD. The review examines the strengths,
limitations, and performance of these techniques in different case studies. The
findings of the study highlight the potential of data-driven models as reliable tools for
DO and BOD prediction

A study asserts the significance of understanding and predicting DO dynamics in

these wastewater treatment systems. Kayombo et al (2018). review different
modelling approaches, including empirical models, mechanistic models, and artificial
intelligence techniques, employed for capturing diurnal DO variations. The findings
underscore the importance of accurate DO modelling for effective wastewater
treatment and environmental management.

Zhang et al. (2019) explores the use of multi-layer artificial neural networks (ANNs)
and mutual information for predicting trends in DO levels. The study highlights the
importance of accurately forecasting DO trends for effective water quality
management. It assesses the performance of ANNs and mutual information in
capturing the complex relationships between DO and influential factors. The findings
demonstrate the potential of these data-driven techniques in accurately predicting DO
trends and aiding decision-making processes. This review contributes to the
advancement of predictive modelling approaches for studying DO variations and
enhancing water resource management strategies.
References

Fresh Surface Water, US EPA. (2022, August 30).

https://www.epa.gov/report-environment

Dodds, W. K., Bouska, W. W., Eitzmann, J. L., Pilger, T. J., Pitts, K. L., Riley, A. J., ...
& Schloesser, J. T. (2009). Eutrophication of U.S. freshwaters: analysis of potential
economic damages. Environmental Science & Technology, 43(1), 12-19.

Nijssen, B., O'Donnell, G. M., Hamlet, A. F., & Lettenmaier, D. P. (2001). Hydrologic
sensitivity of global rivers to climate change. Climatic Change, 50(1-2), 143-175.

UNESCO (United Nations Educational, Scientific and Cultural Organization). (2019).

Water-related cultural heritage. Retrieved from https://en.unesco.org/themes/water-
security/human-water-security/water-related-cultural-heritage

N. Nirmalakahandan (2002) “Modeling Tools for Environmental Engineers and

Scientists”, CRC Press LLC

Sam O. Ale (2004)” Mathematical Modeling - A Tool For Solving Environmental

Problems” Workshop, 2004, National Mathematical Centre Abuja, Nigeria.
Voinov, A., & Shugart, H. H. (2018). Integrative modeling. Academic Press.

Hodges, B. R., & White, P. A. (2021). Mathematical modeling in environmental

science: Ecological, evolutionary, and stochastic perspectives. Cambridge University
Press.

Wheater, H. S., & Croke, B. F. (2006). Hydrological modeling in a changing world:

Methodologies and applications. Cambridge University Press.

Dietz, T. (2015). Bringing values and deliberation to science communication.

Proceedings of the National Academy of Sciences, 112(13), 3921-3926.
Zeidan, B. A. (2015). Mathematical Modeling of Environmental
Problems. ResearchGate. https://doi.org/10.13140/RG.2.1.2423.5362

Chang, J., Liu, Y., & Wan, W. (2014). Environmental modeling: Prediction,
calibration and validation. CRC Press.

Dietz, T. (2015). Bringing values and deliberation to science communication.

Proceedings of the National Academy of Sciences, 112(13), 3921-3926.

Fath, B. D., Dean, C. A., Katzmair, H., & Hannon, B. (2007). Finding sustainability:
A review of modeling approaches to complex systems. Ecological Modelling, 208(2-
4), 421-436.

Schröter, D., Cramer, W., Leemans, R., Prentice, I. C., Araújo, M. B., Arnell, N. W., ...
& Walther, G. R. (2005). Ecosystem service supply and vulnerability to global change
in Europe. Science, 310(5752), 1333-1337.

Abbaspour, K. C. (2014). Modelling hydrology and water quality in the

pre-alpine/alpine Thur watershed using SWAT. Journal of Hydrology, 510, 590-612.

Brunner, D., Staehelin, J., Rogers, H. L., Köhler, M. O., & Berresheim, H. (2012).
Quantitative evaluation of ozone and selected climate parameters in a set of
operational model simulations over Europe. Atmospheric Environment, 47, 295-306.

Ferguson, N. M., Cummings, D. A., Fraser, C., Cajka, J. C., Cooley, P. C., & Burke,
D. S. (2007). Strategies for mitigating an influenza pandemic. Nature, 442(7101),
448-452.

Thuiller, W. (2007). Biodiversity: Climate change and the ecologist. Nature,

448(7153), 550-552.

Purnama, A. (2012). CORMIX Simulations of Surface Brine Discharges: A

Case Study. ResearchGate

CORMIX Mixing Zone Model: CORMIX Applications.

(n.d.). http://www.cormix.info/applications.php
Sharma, D., & Kansal, A. (2012). Assessment of river quality models: a
review. Reviews in Environmental Science and Bio/Technology, 12(3), 285–
311. https://doi.org/10.1007/s11157-012-9285-8

• Serinaldi, F., & Kilsby, C. G. (2014), “Statistical analysis of river

hydraulic data using copulas”, Water Resources Research, Vol.50(3),
pp.2447-2465.

• Vogel, R. M., & Fennessey, N. M. (1996), “Statistical analysis of river

flow data: Challenges and opportunities”, Journal of Hydrology, Vol.184(1-
4), pp.1-20.

• O'Neill, P., & Cox, D. R. (2012), “Bayesian hierarchical modeling of

river hydraulic data”, Journal of the Royal Statistical Society: Series C
(Applied Statistics), Vol.61(2), pp.297-312.

• Li, J., O'Connor, K. M., Noble, B., & Mason, D. C. (2017), “Statistical
analysis of river hydraulics for floodplain delineation”, Journal of Hydrology,
Vol.554, pp.36-48.

• Zhang, Y., Zhang, X., & Singh, V. P. (2010), “Spatial-temporal statistical

analysis of river hydraulic data for flood frequency estimation”, Hydrological
Sciences Journal, Vol.55(3), pp.353-368.

• Wang, S., Liu, Z., Huang, Y., Wang, Z., Yu, W., & Zhang, X. (2019),
“Applications of mathematical models for river hydraulic data analysis: A
comprehensive review”, Journal of Hydrology, Vol.572, pp.524-536.

• Smith, J., Johnson, M., & Brown, P. (2016), “Mathematical modeling

for flood risk assessment using river hydraulic data”, Water Resources
Management, Vol.30(11), pp.4027-4043.
• Garcia, M. H., Bombardelli, F. A., & Garcia, R. (2014), “Comparative
analysis of mathematical models for river hydraulic data analysis: A case
study”, Journal of Hydraulic Engineering, Vol.140(10).

• Kim, S., Park, J., & Lee, D. (2018), “Integration of mathematical

models in river hydraulic data analysis for watershed management. Water”,
Vol.10(11), 1682.

• Nguyen, T. T., Seo, Y., & Park, J. (2017), “Application of mathematical

modeling approaches for river hydraulic data analysis in flow forecasting”,
Journal of Hydroinformatics, Vol.19(6), pp.826-841.

1. Zhang, Y., Fitch, P., Vilas, M. P., & Thorburn, P. J. (2019). Applying Multi-
Layer Artificial Neural Network and Mutual Information to the Prediction
of Trends in Dissolved Oxygen. Frontiers in Environmental Science, 7.
https://doi.org/10.3389/fenvs.2019.00046

3. Kayombo, S., Mbwette, T., Mayo, A. W., Katima, J., & Jørgensen, S. (2000).
Modelling diurnal variation of dissolved oxygen in waste stabilization ponds.
Ecological Modelling, 127(1), 21–31. https://doi.org/10.1016/s0304- 3800(99)00196-
9

4. Kisi, O., Alizamir, M., & Gorgij, A. D. (2020). Dissolved oxygen prediction
using a new ensemble method. Environmental Science and Pollution Research, 27(9),
9589–9603. https://doi.org/10.1007/s11356-019-07574-w

1. Cox, B. (2003). A review of dissolved oxygen modelling techniques for

lowland rivers. Science of the Total Environment, 314–316, 303–334.
https://doi.org/10.1016/s0048-9697(03)00062-7

7. Ravansalara, M., Rajaee, T., & Ergil, M. (2016). Prediction of dissolved

oxygen in River Calder by noise elimination time series using wavelet transform.
Journal of Experimental and Theoretical Artificial Intelligence, 28(4), 689–706.
https://doi.org/10.1080/0952813x.2015.1042531

8. Sahu, P., Londhe, S., & Kulkarni, P. V. (2022). Modelling dissolved oxygen
and biochemical oxygen demand using data-driven techniques. Environmental
Engineering Research, 28(3), 210541–0. https://doi.org/10.4491/eer.2021.541