Mathematical analysis and control optimization of soluble and insoluble water pollutants dispersion Authors: O.Y. Oludoun, S.O. Salawu, S.O. Adesanya, O.E. Abiodun Published: 2024 ABSTRACT One of the major and abundant environmental elements is water, which life existence on earth is depending on. However, degradation of water quality arising from natural occurrences due to anthropogenic that resulted in eutrophication needs to be guided. Indiscriminate harmful waste discharge and dispersion into the water bodies leads to water pollution. As such, constant management and monitoring of water resources for an enhanced quality and quantity is essential. Therefore, this study employed a coupled derivative mathematical model to investigate soluble and insoluble water pollutants. The insoluble pollutants are converted to soluble pollutants through an applied control or treatment. To optimize the water quality, parametric sensitivities in relation to reproduction number are examined. From the computational analysis carried out, it was revealed that treatment still remains the only remedy for safe water. Introduction A concrete abstract description of systems with the help of mathematical language and concepts defined by mathematical modeling. Mathematical formulations or models are applicable in social sciences, agricultural administration, medicine, engineering, and other areas of specialization [1], [2]. Deterministic or stochastic, nonlinear or linear models have been developed to forecast the epidemic outbreaks and control, Bonyah et al. [3]. Mathematical models remain an effective way for infectious diseases to be dynamically studied in a population and host. According to Yavuz and Ozdemir [4], the contagious diseases mathematical model needs a combination of vaccination, containment, quarantine and medication. A compartmental, epidemiological, coupled derivative model has been developed for diseases such as malaria, Ebola, COVID-19, HBV, HIV, and so on, which has attracted attention due to an effective control prediction [4], [5], [6], [7], [8], [9]. Epidemiological models have been developed for water pollution, dispersion and treatment due to its need for life sustenance. Reaction-diffusion models are widely used for soluble pollutants that undergo chemical transformations. Crank [10] and Sposito [11] described how such models can capture the complex interplay of chemical reactions and diffusion processes, particularly for nutrients like nitrates and phosphates. However, these models typically exclude insoluble pollutants, which do not react in the same manner. Water is an element of the environment and a universal chemical solvent that can dissolve several substances. It is a primary component of cells, essential for biological processes and a meaningful part of living organisms. Also, it is functional to life survival and vital in different geological processes, weather patterns, and earth ecosystems [12], [13]. The water sources such as streams, lagoons, lakes, rivers and others are used for industrial, domestic, and irrigation purposes. Meanwhile, human activities often pollute and affect the quality and quantity of good water assessment, Shah et al. [14]. Thus, pollution describes susceptible substances that affect ecological systems, harm living resources, damage structures, and threaten human health, as stated by Sharma and Bhattacharya [15]. Water pollutants are caused by natural processes and human activities, which contaminate the water bodies and lead to water-borne diseases and environmental degradation. Water-borne diseases like respiratory infections, rashes, gastroenteritis, hepatitis, pink eyes, ear-arch and many more can be contracted from polluted water [16], [17]. Among the water pollutants are insecticides, pesticides, untreated sewage, domestic waste, industrial waste, oil spillage and many more, which are equally harmful to aquatic animals, Yeleliere et al. [18]. The water pollutants can be categorized into soluble and insoluble pollutants, as presented by Saravanan et al. [19]. Water pollutants such as soaps, chlorides, nitrates, and so on are grouped as soluble pollutants, while pollutants such as polythene bags, oils, plastics, and many more are grouped as insoluble pollutants [14], [19]. Many pollutants affect marine organisms, damage the human nervous system, and consequently signal serious threats to human life and ecosystems. Thus, contaminated water breeds mosquitoes, a malaria-causing agent and other agents like viruses, parasites and bacteria, Pandey et al. [20]. However, awareness of water pollution has increased recently, but water resource exploitation and sustainability need to be improved to prevent water pollution and water-borne diseases, Boelee et al. [21]. Several scholars have studied water pollution dynamics among is Agusto and Bamigbola [22], who applied a numerical Crank-Nicholson technique to a derivative water pollution and treatment model. The results show that pollutant concentration reduced greatly in each case as treatment was applied. Shah et al. [23] presented a mathematical formulation for the forest resource pollutants dispersion with control maximization strategies. It was revealed that forest resource pollutant density is reduced directly by the harvest from the wood-based factories, but the forest resources are affected directly by the non-wood-based factories. Parsaie and Haghiabi [24] conducted a computational artificial neural network to study river pollution transmission. A longitudinal pollution coefficient of transmission was predicted through the adopted technique. Pimpunchat et al. [25] studied river aeration pollution and its control via a mathematical model. The outcomes depicted that river pollutants insertion rate could be economically dead as the needed oxygen for fish survival is below standard. More earlier reports on water pollution and remedies can be obtained in [26], [27], [28]. Water pollution is a critical environmental issue that affects ecosystems, human health, and economies globally. Pollutants in water bodies can originate from industrial activities, agricultural runoff, urban waste, and natural processes. The harmful effects of water pollution range from habitat destruction, loss of biodiversity, and disruption of food chains to serious public health risks like waterborne diseases and the accumulation of toxic substances in humans through contaminated drinking water and food sources. The various reports on water pollution and the essentials of clear water for microorganisms, aquatic animals and human beings' survival motivated the research. Thus, this study aims to examine soluble and insoluble water pollutant dispersion and control optimization through a bio-mathematical model analysis. There is a clear research gap in developing integrated models and control strategies that simultaneously address soluble and insoluble pollutants in complex, dynamic environments. Moreover, the interaction between these pollutants, the long-term behavior of insoluble substances, and the integration of real-time data into control models remain under-explored. Addressing these gaps through hybrid modeling approaches and advanced optimization techniques will contribute significantly to improving water quality management. The significance of water in agricultural activities, domestic and engineering processes, and others cannot be overstressed. As such, the study tends to analyze the pollution stability equilibrium point at the pollutants-free equilibrium and endemic and optimize the pollutants control and treatment strategies in reducing pollution of water bodies. Thus, the study developed an integrated mathematical model that captures the dispersion dynamics of both soluble and insoluble pollutants, accounting for their interaction mechanisms, such as adsorption, desorption, and sedimentation. The descriptions model formulation Consider a compartmental mathematical model for the soluble and insoluble propagation of water pollutions with suitable treatment. A discrete cubic volume compartmental derivative model for the polluted water sources (Pw), susceptible-soluble water pollutants (Sw), infected-insoluble water pollutants (Iw) and treated-insoluble water sources (Tw) is formulated. In reality, many water bodies are contaminated with both types of pollutants, and they often interact in complex ways. As schematically demonstrated in Fig. 1, the dispersion of the water pollutants model is developed according to [14], [29], considering the assumptions that the water body is treated as a homogeneous medium, meaning that its properties are uniform throughout the domain of interest, the pollutants are assumed to be sufficiently diluted such that their interactions with the water body do not affect the flow properties or create significant nonlinear effects, the insoluble pollutants, biodegradation is often neglected in the modeling framework, assuming that the primary processes affecting these pollutants are physical, and the physical properties of the fluid, such as viscosity and density, are assumed to remain constant during the modeling period, irrespective of changes in pollutant concentration or flow conditions.Figure 1Model transmission configuration.Figure 1 Model transmission configuration. Appropriate dynamical parameters are considered in the development of the nonlinear coupled derivative model. The mechanisms for the mitigation of insoluble pollutants spread across the water regime are optimized for an effective control. Thus, the derivative dimensional model is given as:(1)dPwdt=α−βPwSw−γPwIw+ϕβIw−φPw,Pw0≥0,(2)dSwdt=βPwSw+λIw−(ϵ+φ)Sw,Sw0≥0,(3)dIwdt=γPwIw−ϕβIw−(λ+δ+φ)Iw,Iw0≥0,(4)dTwdt=ϵSw+δIw−φTw,Tw0≥0, where Pw(0)=Pw0,Sw(0)=Sw0,Iw(0)=Iw0,Tw(0)=Tw0 are the initial constraints. The embedded term α defines water pollutants rate, β and γ are soluble and insoluble water pollutants transmission rate separately. The parameter φ denotes compartmental water pollutants removable rate, λ describes insoluble water pollutants treatment rate, ϕ connotes water pollution rate of insoluble water pollutants, and ϵ and δ are the compartmental cubic volume treatment for soluble and insoluble water pollutants respectively. Analysis of the mathematical model The solution positivity and feasibility of the basic system of equations (1) to (4) are described in this section. Solution positivity For meaningful epidemiological translation, the water pollutants dispersion system of equations (1) to (4) are investigated for solution positivity. Thus, it is necessary to establish the non-negativity of all its stated variables for all t≥0. Theorem 2.1 The solution to the non-negative initial conditions of the coupled equations (1) to (4) of (Pw,Sw,Iw,Tw) are bounded uniformly on [0,∞) . The solution to the non-negative initial conditions of the coupled equations (1) to (4) of (Pw,Sw,Iw,Tw) are bounded uniformly on [0,∞) . ProofWith non-negative initial conditions, the solutions to (Pw,Sw,Iw,Tw) are assumed to exist and unique. Taking equation (1) from the derivative modeled equations and simplify to become:dPwdt=α+ϕβIw−(βSw+γIw+φ)Pw, removing the term α+ϕβIw since it is non-negative to have(5)dPwdt≥−(βSw+γIw+φ)Pw. Integrating both sides of equation (5) and apply separation of variables method, thus gives∫dPwPw≥−∫(βSw+γIw+φ)dt,ln|Pw|≥−(βSw+γIw+φ)t+K,Pw(t)≥e−(βSw+γIw+φ)t+K, apply the initial conditions at t=0; hence, resulted toPw(t)≥Pw0e−(βSw+γIw+φ)t≥0,since(βSw+γIw+φ)>0. Therefore, Pw≥0 for t≥0. Also, taking equation (2) from the derivative model,dSwdt=βPwSw+λIw−(ϵ+φ)Sw, due to non-negativity, the term βPwSw+λIw is removed, then the equation reduced to:(6)dSwdt≥−(ϵ+φ)Sw. Solving equation (6) by using variable separation method to have∫dSwSw≥−∫(ϵ+φ)dt,ln|Sw|≥−(ϵ+φ)t+K,Sw(t)≥e−(ϵ+φ)t+K, introducing the Sw(t) initial condition at t=0,Sw(t)≥Sw0e−(ϵ+φ)t≥0since(ϵ+φ)>0. Similarly, investigate the solution positivity of equation (3) given asdIwdt=γPwIw−(ϕγ+ϵ+φ+λ)Iw, remove the term γPwIw since it is non-negative, thus reduced to(7)dIwdt≥−(ϕγ+ϵ+φ+λ)Iw. Integrating both sides of equation (7) via variables separating method to have∫dIwIw≥−∫(ϕγ+ϵ+φ+λ)dt,ln|Iw|≥−(ϕγ+ϵ+φ+λ)t+K,Iw(t)≥e(ϕγ+ϵ+φ+λ)t+K, using the initial condition when t=0, this resulted toIw(t)≥Iw0e−(ϕγ+ϵ+φ+λ)t≥0,since(ϕγ+ϵ+φ+λ)>0. Finally, taking equation (4) from the modeled equations, such thatdTwdt=ϵSw+δIw−φTw, ignoring the non-negative term ϵSw+δIw, thus the equation gives(8)dTwdt≥−φTw By applying separation of variables via integrating both sides of equation (8) to have∫dTwdt≥−∫φdt,ln|Tw|≥−φt+K,Tw(t)≥e−φt+K, employ the initial condition at t=0, thus givesTw(t)≥Tw0eφt≥0,sinceφ>0. Therefore, at t>0, Pw,Sw,Iw,Tw are all non-negative i.e. all give positive values. Hence, the proof. With non-negative initial conditions, the solutions to (Pw,Sw,Iw,Tw) are assumed to exist and unique. Taking equation (1) from the derivative modeled equations and simplify to become:dPwdt=α+ϕβIw−(βSw+γIw+φ)Pw, removing the term α+ϕβIw since it is non-negative to have(5)dPwdt≥−(βSw+γIw+φ)Pw. Integrating both sides of equation (5) and apply separation of variables method, thus gives∫dPwPw≥−∫(βSw+γIw+φ)dt,ln|Pw|≥−(βSw+γIw+φ)t+K,Pw(t)≥e−(βSw+γIw+φ)t+K, apply the initial conditions at t=0; hence, resulted toPw(t)≥Pw0e−(βSw+γIw+φ)t≥0,since(βSw+γIw+φ)>0. Therefore, Pw≥0 for t≥0. Also, taking equation (2) from the derivative model,dSwdt=βPwSw+λIw−(ϵ+φ)Sw, due to non-negativity, the term βPwSw+λIw is removed, then the equation reduced to:(6)dSwdt≥−(ϵ+φ)Sw. Solving equation (6) by using variable separation method to have∫dSwSw≥−∫(ϵ+φ)dt,ln|Sw|≥−(ϵ+φ)t+K,Sw(t)≥e−(ϵ+φ)t+K, introducing the Sw(t) initial condition at t=0,Sw(t)≥Sw0e−(ϵ+φ)t≥0since(ϵ+φ)>0. Similarly, investigate the solution positivity of equation (3) given asdIwdt=γPwIw−(ϕγ+ϵ+φ+λ)Iw, remove the term γPwIw since it is non-negative, thus reduced to(7)dIwdt≥−(ϕγ+ϵ+φ+λ)Iw. Integrating both sides of equation (7) via variables separating method to have∫dIwIw≥−∫(ϕγ+ϵ+φ+λ)dt,ln|Iw|≥−(ϕγ+ϵ+φ+λ)t+K,Iw(t)≥e(ϕγ+ϵ+φ+λ)t+K, using the initial condition when t=0, this resulted toIw(t)≥Iw0e−(ϕγ+ϵ+φ+λ)t≥0,since(ϕγ+ϵ+φ+λ)>0. Finally, taking equation (4) from the modeled equations, such thatdTwdt=ϵSw+δIw−φTw, ignoring the non-negative term ϵSw+δIw, thus the equation gives(8)dTwdt≥−φTw By applying separation of variables via integrating both sides of equation (8) to have∫dTwdt≥−∫φdt,ln|Tw|≥−φt+K,Tw(t)≥e−φt+K, employ the initial condition at t=0, thus givesTw(t)≥Tw0eφt≥0,sinceφ>0. Therefore, at t>0, Pw,Sw,Iw,Tw are all non-negative i.e. all give positive values. Hence, the proof. Solution feasibility The solution feasibility set for the modeled derivative equations (1) to (4) can be obtained through its positively invariant set give asΛ={(Pw,Sw,Iw,Tw)∈R+4:A(t)≤αφ}. Lemma 2.1With respect to the dynamical system of the water pollutants transmission equations(1)to(4), the region Λ is positively invariant. With respect to the dynamical system of the water pollutants transmission equations(1)to(4), the region Λ is positively invariant. ProofTo demonstrate the positively invariant region, A(t)≤αφ as t→∞ shall be established. The total population from the model is A=Pw+Sw+Iw+Tw, thus differentiate A and the right hand side with respect to t to have an equation satisfying equations (1) to (4):A′=Pw′+Sw′+Iw′+Tw′,=α−βPwSw−γPwIw+ϕβIw−φPw+βPwSw+λIw−(ϵ+φ)Sw+γPwIw−ϕβIw−(δ+φ+δ)Iw+ϵSw+δIw−φTw,=α−φPw−φSw+φIw−φTw,=α−φA,sinceA=Pw+Sw+Iw+Tw, therefore, this givesdAdt=α−φA,(9)dAdt+φA=α. Apply the idea of integrating factor to equation (9), solving it gives(10)A(t)=αφ+Ce−φt, where C is a constant, introducing the initial condition A(0)=A0 on equation (10), this givesA(0)=αφ+Ce−φ(0)=A0,whereC=A0−αφ Hence, equation (10) becomes(11)A(t)=αφ(1−e−φt)+A0e−φt, taking the limit as t→∞, thus equation (11) gives(12)limt→∞A(t)≤αφ. Hence, equations (1) to (4) feasible region is obtained as equation (12), which can be described as Λ={(Pw,Sw,Iw,Tw)∈R+4:A=αφ} is positively invariant. Thus, the water pollutants transmission model is epidemiologically and mathematically posted. To demonstrate the positively invariant region, A(t)≤αφ as t→∞ shall be established. The total population from the model is A=Pw+Sw+Iw+Tw, thus differentiate A and the right hand side with respect to t to have an equation satisfying equations (1) to (4):A′=Pw′+Sw′+Iw′+Tw′,=α−βPwSw−γPwIw+ϕβIw−φPw+βPwSw+λIw−(ϵ+φ)Sw+γPwIw−ϕβIw−(δ+φ+δ)Iw+ϵSw+δIw−φTw,=α−φPw−φSw+φIw−φTw,=α−φA,sinceA=Pw+Sw+Iw+Tw, therefore, this givesdAdt=α−φA,(9)dAdt+φA=α. Apply the idea of integrating factor to equation (9), solving it gives(10)A(t)=αφ+Ce−φt, where C is a constant, introducing the initial condition A(0)=A0 on equation (10), this givesA(0)=αφ+Ce−φ(0)=A0,whereC=A0−αφ Hence, equation (10) becomes(11)A(t)=αφ(1−e−φt)+A0e−φt, taking the limit as t→∞, thus equation (11) gives(12)limt→∞A(t)≤αφ. Hence, equations (1) to (4) feasible region is obtained as equation (12), which can be described as Λ={(Pw,Sw,Iw,Tw)∈R+4:A=αφ} is positively invariant. Thus, the water pollutants transmission model is epidemiologically and mathematically posted. Pollutants free equilibrium [PFE] The concept of pollutants-free equilibrium is important in water pollution models as it represents the ideal target for water quality restoration efforts. While this equilibrium may be difficult to achieve in practice, mathematical models provide insights into the conditions necessary for the reduction of both soluble and insoluble pollutants, and the strategies required to move towards a cleaner, more sustainable state in aquatic ecosystems. The PFE can be determined from the derivative equations (1) to (4) by settingdPwdt=dSwdt=dIwdt=dTwdt=0. Taking from equation (3),Iw′=γPwIw−ϕβIw−(δ+φ−λ)Iw=0,(γPw−ϕβ−δ−φ−λ)Iw=0,Iw=0. From equation (2), the equilibrium point is obtained as followsSw′=βPwSw+λIw−(ϵ+φ)Sw=0,λIw−(ϵ+φ−βPw)Sw=0,sinceIw=0,Sw=0. Likewise, from equation (4), it can be gotten as followsTw′=ϵSw+δIw−φTw=0,φTw=0,sinceIw=Sw=0,Tw=0. Similarly, taking from equation (1), the equilibrium point is determinedPw′=α−βPwSw−γPwIw+ϕβIw−φpw=0,α−φPw=0,sinceIw=Sw=0,Pw=αφ.(13)(Pw0,Sw0,Iw0,Tw0)=(αφ,0,0,0). Therefore, the PFE is obtained as equation (13). The pollutants basic reproduction number (R0) This is the epidemiology models threshold quantity, which describes the secondary cases expected number that a pollution or infection produced with a single primary case in water body or host population such that everyone is susceptible as stated by Adeniyi et al. [30] and Oke et al. [31]. Thus, when R0>0 implies high water pollution or infection dispersion rate, and when R0<0 indicates that the pollution or infection will disappear or die overtime. Evaluation of R0 via next generation matrix The R0 can be determined using next generation matrix with distinct finitely several groups of pollutants are recognized. The next generation matrix method properly position terms in infected or polluted group to the vectors G and H. The compartmental new infection appearance rate is G and compartmental out and in individual transfer rate that need to be negated. As illustrated by Pandey et al. [32], differentiating G and H with respect to variables subset obtained from the free equilibrium pollutant gives Jacobian matrices. Thus, resulted in the matrices G and H separately. Let X′=(Pw,Sw,Iw), the system of equations can be express in matrix form as:dXdt=ur(X)−vr(X),r=Sw,Iw, whereX=(SwIw),ur(X)=(βPwSwβPwIw),vr(X)=(−λIw+(φ+ϵ)SwϕβIw+(λ+φ+δ)Iw). The pollutants propagation matrix G and transition matrix H denoted as:G=[∂ur(X)∂Xn],H=[∂vr(X)∂Xn],which resulted to,G=(βPw00γPw),H=(φ+ϵ−λ0βϕ+λ+φ+δ), Hence, the PFE Jacobian matrix for G and H is given as:G=(αβφ00αγφ),andH=(φ+ϵ−λ0βϕ+λ+φ+δ), Therefore, the next generation matrix is presented as GH−1, such that H givesH−1=(1φ+ϵλ(φ+ϵ)(βϕ+λ+φ+δ)01βϕ+λ+φ+δ), Thus,GH−1=(αββφ(φ+ϵ)αβλφ(φ+ϵ)(βϕ+λ+φ+δ)0αβφ(βϕ+λ+φ+δ)). Determine the eigenvalues of the characteristic equation to obtain R0,det(GH−1−χI)=0,(14)det(αββφ(φ+ϵ)−χαβλφ(φ+ϵ)(βϕ+λ+φ+δ)0αβφ(βϕ+λ+φ+δ)−χ)=0. Hence, equation (14) gives the eigenvalues of the PFE and gotten as(αβφ(φ+ϵ),αγφ(βϕ+λ+φ+δ)). The two distinct eigenvalues are:R1=αβφ(φ+ϵ)andR2=αγφ(βϕ+λ+φ+δ). As such, the basic reproduction ratio or number is described asR0=max(R1,R2). Existence of equilibrium point stability Stability analysis is the best tool for classifying equilibrium point of a dynamical system, be it the local and the global stability. Local stability of the PFE The PFE is asymptotically stable locally if R0<1 and unstable if otherwise. To show that the pollutants basic reproduction number R0 is asymptotically stable or not at the PFE P0. The Jacobian matrix is used to determine the stability of each equilibrium points.(15)τ=[−βSw−γIw−φ−βPw−γPw+βϕ0βSwβPw−φ−ϵλ0γIw0γPw−βϕ−λ−φ−δ00ϵδ−φ]. Evaluating the Jacobian matrix of equation (15) at the PFE P0(16)τP0=[φ−βαφ−γαφ+βϕ00βαφ−φ−ϵλ000γαφ−βϕ−λ−φ−δ00ϵδ−φ]. The eigenvalues of equation (16) give χ1=χ2=−φ<0. Reducing the matrix and finding the determinant to obtain the other eigen values.(17)τP0=[βαφ−φ−ϵλ0γαφ−βϕ−λ−φ−δ]. Hence, from equation (17), the remnant eigenvalues are:χ3=βαφ−φ−ϵχ4=γαφ−βϕ−λ−φ−δ. For PFE to be locally asymptotically stable, the eigenvalues χ3&χ4 must be negative. i.e., βαφ−φ−ϵ<0→βαφ<φ+ϵ This can be rewritten as: βαφ(φ+ϵ)<1 The PFE is locally asymptotically unstable if βαφ−φ−ϵ>0→βαφ>φ+ϵ This can be rewritten as: βαφ(φ+ϵ)>1 Therefore,(18)R1=βαφ(φ+ϵ) Also, for the χ4. It is stable if γαφ−βϕ−λ−φ−δ<0→γαφ<βϕ+λ+φ+δ. This can also be written as: γαφ(βϕ+λ+φ+δ)<1. It is asymptotically unstable if, γαφ−βϕ−λ−φ−δ>0→γαφ>βϕ+λ+φ+δ. This can also be written as: γαφ(βϕ+λ+φ+δ)>1. Therefore,(19)R2=γαφ(βϕ+λ+φ+δ) Thus, the combination of equations (18) and (19) gives R0, i.e. R1+R2=R0 Hence, the PFE is locally asymptotically stable if R0<1 or unstable if R0>1. It is vital to demonstrate that the PFE is globally-asymptotically stable in order to guarantee that the removal of pollutants is independent of the starting size of the infected water sources. Global stability of the PFE Castilla-Chavez et al. [33] demonstrated the global asymptotic stability of the PFE.(20)Γ′=Y(X,I)(21)Ξ′=Z(X,I),Z(X,0)=0 If the two requirements in equations (20) and (21) are met, then PFE becomes globally asymptotic stable. Also, X∈R2 represents the number of uncontaminated water sources and I∈R2 represents the number of contaminated water sources. P0=(Pw0,Sw0,Iw0,Tw0)=(X⁎,0,0,0) denotes PFE. To ensure the global stability of the PFE, two conditions must be met:(22)(ϒ1)forX′=Y(X,0),X⁎globally stable(23)(ϒ2)Z(X,I)=DI−Zˆ(X,I),Zˆ(X,I)≥0∀(X,I)∈B, where D=P1Z(X,0) is an M-matrix with non-negative off diagonal entries, and B are the systems of equation. The following theorem applies if the system of equations (1)-(4) meets the conditions in (22) and (23). Observe thatZˆ(X,I)=DI−I′ where I=(Sw,Iw) and I′=(Sw′,Iw′) given by equations (2) and (3) respectively. Theorem The PFE P0=(αφ,0,0,0) is globally asymptotically stable if R0<1 and the condition in (23) is met. The PFE P0=(αφ,0,0,0) is globally asymptotically stable if R0<1 and the condition in (23) is met. ProofThe systems of equations in equations (14) will be rewritten in the form of (20) and (21), and then X∈R2=(Pw,Tw) with X1=Pw and X2=Tw, I∈R2=(Sw,Iw),X′=Y(X,0)=(α−φPw0)X1′=α−φPw is a linear differential equation and its solution is given as:Pw(t)=αφ−(αφ−Pw0(t))e−ut Clearly, Pw(t)→αφast→∞, since e−∞=0, ∀Pw0. Hence X1⁎ is globally stable asymptotically.So thatZ(X,I)=(βPwSw+λIw−(ϵ+φ)SwγPwIw−(δ+λ+φ)Iw) and(24)D=(β−(ϵ+φ)λ0γ−(βϕ+δ+λ+φ)) Clearly, (24) is a non negative diagonal M-matrix. Therefore:(25)Zˆ(X,I)=(Z1ˆ(X,I)Z2ˆ(X,I))=(βSw(1−Pw)γIw(1−Pw)) (ϒ2) is satisfied whenever Pw<1 is satisfied, but (ϒ2) is unsatisfied when Pw<0.Hence from equation (25), the PFE P0 may be stable globally or not. The systems of equations in equations (14) will be rewritten in the form of (20) and (21), and then X∈R2=(Pw,Tw) with X1=Pw and X2=Tw, I∈R2=(Sw,Iw),X′=Y(X,0)=(α−φPw0)X1′=α−φPw is a linear differential equation and its solution is given as:Pw(t)=αφ−(αφ−Pw0(t))e−ut Clearly, Pw(t)→αφast→∞, since e−∞=0, ∀Pw0. Hence X1⁎ is globally stable asymptotically. So thatZ(X,I)=(βPwSw+λIw−(ϵ+φ)SwγPwIw−(δ+λ+φ)Iw) and(24)D=(β−(ϵ+φ)λ0γ−(βϕ+δ+λ+φ)) Clearly, (24) is a non negative diagonal M-matrix. Therefore:(25)Zˆ(X,I)=(Z1ˆ(X,I)Z2ˆ(X,I))=(βSw(1−Pw)γIw(1−Pw)) (ϒ2) is satisfied whenever Pw<1 is satisfied, but (ϒ2) is unsatisfied when Pw<0. Hence from equation (25), the PFE P0 may be stable globally or not. Existence of the pollutant endemic equilibrium points The endemic equilibrium points of the pollutant exist and can be denoted as: Pw⁎, Sw⁎, Iw⁎, Tw⁎ which is usually calculated by setting the equations (1)-(4) to zero and stating the states variables as a function of the transmission coefficients, according to [27], [30], [31]. The endemic equilibrium point here is divided into P1⁎, a situation where the insoluble pollutants in the water do not exist and P2⁎, where all types of pollutants exist in the water. The endemic equilibrium are given below:P1⁎=(φ+ϵβ,−φ2+ϵφ−αββ(φ+ϵ),0,−ϵ(φ2+ϵφ−αβ)βφ(φ+ϵ))P2⁎=(Pw⁎,Sw⁎,Iw⁎,Tw⁎) wherePw⁎=λ+φ+δ+γϕγ,Sw⁎=λ(λφ−αγ+φδ+φ2+γφϕa,Iw⁎=−(βλ+βφ−γφ+βδ−γϵ+βγϕ)γλSw⁎,Tw⁎=−(βδ2+βλδ−γλϵ+βφδ−γφδ−γϵδ+βγϕδ)φγλSw⁎, wherea=(βφ2+γφ2+βδ2+βλφ−γλφ+βλδ−γλϵ+2βφδ−γφϵ−γφδ−γϵδ+βγφϕ+βγϕδ) Local stability of the endemic equilibrium Theorem The endemic equilibrium point is said to be locally stable asymptotically if R0>1 and unstable if R0<1 . The endemic equilibrium point is said to be locally stable asymptotically if R0>1 and unstable if R0<1 . To show that R0 is stable or unstable at the endemic equilibrium P⁎. The local stability of the equilibrium points can be determined by evaluating the Jacobian matrix of the system in (1)-(4). Evaluating (15) at endemic equilibrium point P⁎, to give(26)τP⁎=[−λb1(β+λb2)a−φ−β(γϕ+λ+φ+δ)γ−(λ+φ+δ)0βλb1aβ(γϕ+λ+φ+δ)γ−φ−ϵλ0δ2b1b2a0000ϵδ−φ]. The eigenvalues are gotten from the characteristics equation |τP⁎−χI|=0 of (26). Clearly the first eigenvalue is: χ1=φ<0. The subsequent eigenvalues are gotten from the remaining 3×3 matrix given below:(27)τP⁎=[−λb1(β+λb2)a−φ−β(γϕ+λ+φ+δ)γ−(λ+φ+δ)βλb1aβ(γϕ+λ+φ+δ)γ−φ−ϵλδ2b1b2a00]. Equation (27) has the characteristics polynomial given as:a0χ3+a1χ2+a2χ+a3=0, wherea=(βφ2+γφ2+βδ2+βλφ−γλφ+βλδ−γλϵ+2βφδ−γφϵ−γφδ−γϵδ+βγφϕ+βγϕδ),b1=(−αγ+γφϕ+λφ+φ2+εδ),γφ+φϵ−βγϕ−βλ−βφ−βϵ,a0=1,a1=−β(γϕ+λ+φ+δ)γ+λb1(β+λb2)a+2φ+ϵ,a2=(−δb1(β+λb2)−φaa)(−β(γϕ+λ+φ+δ)γ),a3=(λ2b1b2−β(γϕ+λ+φ+δ)γ−φ−ϵa−βλ3b12b2a2)−(λ+φ+ϵ)−βλ3b1b2(γϕ+λ+φ+δ)aγ. From the characteristics equation above, the stability can be determined using the Routh-Hurwitz stability criterion which states that, if the conditions a1>0,a2>0,a1.a2−a3>0 are satisfied, then characteristic polynomial have negative real roots which shows that the endemic equilibrium point is asymptotically stable locally. By Routh Hurwitz criteria therefore, the endemic equilibrium point is stable when R0>1, otherwise unstable. Global stability of the endemic equilibrium Oludoun et al. [34] demonstrated the global asymptotic stability of the endemic equilibrium using the Lyapunouv method. Theorem The equations of the model have a positive distinctive endemic equilibrium whenever R0>1 , which is said to be globally asymptotically stable. The equations of the model have a positive distinctive endemic equilibrium whenever R0>1 , which is said to be globally asymptotically stable. ProofConsidering the Lyapunov function defined as:(28)L(Pw⁎,Sw⁎,Iw⁎,Tw⁎)=(Pw−Pw⁎ln(PwPw⁎))+(Sw−Sw⁎ln(SwSw⁎))+(Iw−Iw⁎ln(IwIw⁎))+(Tw−Tw⁎ln(TwTw⁎)), Taking the derivative of L in equation (28) along the system as:(29)dLdt=(1−Pw⁎Pw)dPwdt+(1−Sw⁎Sw)dSwdt+(1−Iw⁎Iw)dIwdt+(1−Tw⁎Tw)dTwdt, It would be interesting to point out that, at this point, the partial derivatives in equation (29) are replaced by their expressions following equations (1) to (4).(30)dLdt=(1−Pw⁎Pw)(α−βPwSw−γPwIw+ϕβIw−φPw)+(1−Sw⁎Sw)(βPwSw+λIw−(ϵ+φ)Sw)+(1−Iw⁎Iw)(γPwIw−ϕβIw−(λ+δ+φ)Iw)+(1−Tw⁎Tw)(ϵSw+δIw−φTw), At equilibrium, equation (30) gives:α=βPw⁎Sw⁎−γPw⁎Iw⁎+ϕβIw⁎−φPw⁎,ϵ+φ=βPw⁎Sw⁎+λIw⁎Sw⁎,ϕβ+λ+δ+φ=γPw⁎Iw⁎Iw⁎,φ=ϵSw⁎+δIw⁎Tw⁎,(31)dLdt=(1−Pw⁎Pw)(βPw⁎Sw⁎−γPw⁎Iw⁎+ϕβIw⁎−φPw⁎−βPwSw−γPwIw+ϕβIw−φPw)+(1−Sw⁎Sw)(βPwSw+λIw−(βPw⁎Sw⁎+λIw⁎Sw⁎)Sw)+(1−Iw⁎Iw)(γPwIw−γPw⁎Iw⁎Iw⁎Iw)+(1−Tw⁎Tw)(ϵSw+δIw−ϵSw⁎+δIw⁎Tw),dLdt=(1−Pw⁎Pw)[βPwSw(1−Pw⁎Sw⁎PwSw)−γPwIw(1−Pw⁎Iw⁎PwIw)+ϕβIw(1−Iw⁎Iw)−φPw(1−Pw⁎Pw)]+(1−Sw⁎Sw)[βPwSw+λIw−(βPw⁎Sw⁎Sw+λIw⁎SwSw⁎)]+(1−Iw⁎Iw)(γPwIw−γPw⁎Iw⁎IwIw⁎)+(1−Tw⁎Tw)((ϵSw+δIw)−ϵSw⁎+δIw⁎Tw),(32)dLdt=(1−Pw⁎Pw)[βPwSw(1−Pw⁎Sw⁎PwSw)−γPwIw(1−Pw⁎Iw⁎PwIw)+ϕβIw(1−Iw⁎Iw)−φPw(1−Pw⁎Pw)]+(1−Sw⁎Sw)[βPwSw(1−Pw⁎Sw⁎PwSw)+λIw(1−Iw⁎SwSw⁎)]+(1−Iw⁎Iw)[γPwIw(1−Pw⁎Iw⁎Iw⁎)]+(1−Tw⁎Tw)[ϵSw(1−Sw⁎SwTw⁎)+δIw−(1−Iw⁎IwTw⁎)],(33)dLdt=−φPw(1−Pw⁎Pw)2−βPwSw(1−Pw⁎Sw⁎PwSw)(1−Pw⁎Pw)−γPwIw(1−Pw⁎Iw⁎PwIw)(1−Pw⁎Pw)+ϕβIw(1−Iw⁎Iw)(1−Pw⁎Pw)+βPwSw(1−Pw⁎Sw⁎PwSw)(1−Sw⁎Sw)+λIw(1−Iw⁎SwSw⁎)(1−Sw⁎Sw)+γPwIw(1−Pw⁎Iw⁎Iw⁎)(1−Iw⁎Iw)+ϵSw(1−Sw⁎SwTw⁎)(1−Tw⁎Tw)−δIw−(1−Iw⁎IwTw⁎)(1−Tw⁎Tw), Simplifying equations (31)- (33) gives:=−φPw(1−Pw⁎Pw)2+P1(Pw,Sw,Iw,Tw)+P2(Pw,Sw,Iw,Tw). Where,P1(Pw,Sw,Iw,Tw)=ϕβIw(1−Iw⁎Iw)+βPwSw(1−Pw⁎Sw⁎PwSw)(1−Sw⁎Sw)+λIw(1−Iw⁎SwSw⁎)(1−Sw⁎Sw)+γPwIw(1−Pw⁎Iw⁎Iw⁎)(1−Iw⁎Iw)+ϵSw(1−Sw⁎SwTw⁎)(1−Tw⁎Tw),P2(Pw,Sw,Iw,Tw)=−βPwSw(1−Pw⁎Sw⁎PwSw)(1−Pw⁎Pw)−γPwIw(1−Pw⁎Iw⁎PwIw)(1−Pw⁎Pw)−δIw−(1−Iw⁎IwTw⁎)(1−Tw⁎Tw), P1≤0, whenever,(34)Iw≥Iw⁎,PwSw≥Pw⁎Sw⁎,Sw≥Sw⁎,IwSw≥Iw⁎Sw⁎, P2≤0, whenever,(35)Pw≥Pw⁎,PwSw≥Pw⁎Sw⁎,Tw≥Tw⁎,IwTw≥Iw⁎Tw⁎. Thus, dLdt≤0 if the conditions in (34) and (35) hold.Therefore, by Oludoun et al. [34], the positive equilibrium state is globally asymptotically stable in the positive region R+4. Considering the Lyapunov function defined as:(28)L(Pw⁎,Sw⁎,Iw⁎,Tw⁎)=(Pw−Pw⁎ln(PwPw⁎))+(Sw−Sw⁎ln(SwSw⁎))+(Iw−Iw⁎ln(IwIw⁎))+(Tw−Tw⁎ln(TwTw⁎)), Taking the derivative of L in equation (28) along the system as:(29)dLdt=(1−Pw⁎Pw)dPwdt+(1−Sw⁎Sw)dSwdt+(1−Iw⁎Iw)dIwdt+(1−Tw⁎Tw)dTwdt, It would be interesting to point out that, at this point, the partial derivatives in equation (29) are replaced by their expressions following equations (1) to (4).(30)dLdt=(1−Pw⁎Pw)(α−βPwSw−γPwIw+ϕβIw−φPw)+(1−Sw⁎Sw)(βPwSw+λIw−(ϵ+φ)Sw)+(1−Iw⁎Iw)(γPwIw−ϕβIw−(λ+δ+φ)Iw)+(1−Tw⁎Tw)(ϵSw+δIw−φTw), At equilibrium, equation (30) gives:α=βPw⁎Sw⁎−γPw⁎Iw⁎+ϕβIw⁎−φPw⁎,ϵ+φ=βPw⁎Sw⁎+λIw⁎Sw⁎,ϕβ+λ+δ+φ=γPw⁎Iw⁎Iw⁎,φ=ϵSw⁎+δIw⁎Tw⁎,(31)dLdt=(1−Pw⁎Pw)(βPw⁎Sw⁎−γPw⁎Iw⁎+ϕβIw⁎−φPw⁎−βPwSw−γPwIw+ϕβIw−φPw)+(1−Sw⁎Sw)(βPwSw+λIw−(βPw⁎Sw⁎+λIw⁎Sw⁎)Sw)+(1−Iw⁎Iw)(γPwIw−γPw⁎Iw⁎Iw⁎Iw)+(1−Tw⁎Tw)(ϵSw+δIw−ϵSw⁎+δIw⁎Tw),dLdt=(1−Pw⁎Pw)[βPwSw(1−Pw⁎Sw⁎PwSw)−γPwIw(1−Pw⁎Iw⁎PwIw)+ϕβIw(1−Iw⁎Iw)−φPw(1−Pw⁎Pw)]+(1−Sw⁎Sw)[βPwSw+λIw−(βPw⁎Sw⁎Sw+λIw⁎SwSw⁎)]+(1−Iw⁎Iw)(γPwIw−γPw⁎Iw⁎IwIw⁎)+(1−Tw⁎Tw)((ϵSw+δIw)−ϵSw⁎+δIw⁎Tw),(32)dLdt=(1−Pw⁎Pw)[βPwSw(1−Pw⁎Sw⁎PwSw)−γPwIw(1−Pw⁎Iw⁎PwIw)+ϕβIw(1−Iw⁎Iw)−φPw(1−Pw⁎Pw)]+(1−Sw⁎Sw)[βPwSw(1−Pw⁎Sw⁎PwSw)+λIw(1−Iw⁎SwSw⁎)]+(1−Iw⁎Iw)[γPwIw(1−Pw⁎Iw⁎Iw⁎)]+(1−Tw⁎Tw)[ϵSw(1−Sw⁎SwTw⁎)+δIw−(1−Iw⁎IwTw⁎)],(33)dLdt=−φPw(1−Pw⁎Pw)2−βPwSw(1−Pw⁎Sw⁎PwSw)(1−Pw⁎Pw)−γPwIw(1−Pw⁎Iw⁎PwIw)(1−Pw⁎Pw)+ϕβIw(1−Iw⁎Iw)(1−Pw⁎Pw)+βPwSw(1−Pw⁎Sw⁎PwSw)(1−Sw⁎Sw)+λIw(1−Iw⁎SwSw⁎)(1−Sw⁎Sw)+γPwIw(1−Pw⁎Iw⁎Iw⁎)(1−Iw⁎Iw)+ϵSw(1−Sw⁎SwTw⁎)(1−Tw⁎Tw)−δIw−(1−Iw⁎IwTw⁎)(1−Tw⁎Tw), Simplifying equations (31)- (33) gives:=−φPw(1−Pw⁎Pw)2+P1(Pw,Sw,Iw,Tw)+P2(Pw,Sw,Iw,Tw). Where,P1(Pw,Sw,Iw,Tw)=ϕβIw(1−Iw⁎Iw)+βPwSw(1−Pw⁎Sw⁎PwSw)(1−Sw⁎Sw)+λIw(1−Iw⁎SwSw⁎)(1−Sw⁎Sw)+γPwIw(1−Pw⁎Iw⁎Iw⁎)(1−Iw⁎Iw)+ϵSw(1−Sw⁎SwTw⁎)(1−Tw⁎Tw),P2(Pw,Sw,Iw,Tw)=−βPwSw(1−Pw⁎Sw⁎PwSw)(1−Pw⁎Pw)−γPwIw(1−Pw⁎Iw⁎PwIw)(1−Pw⁎Pw)−δIw−(1−Iw⁎IwTw⁎)(1−Tw⁎Tw), P1≤0, whenever,(34)Iw≥Iw⁎,PwSw≥Pw⁎Sw⁎,Sw≥Sw⁎,IwSw≥Iw⁎Sw⁎, P2≤0, whenever,(35)Pw≥Pw⁎,PwSw≥Pw⁎Sw⁎,Tw≥Tw⁎,IwTw≥Iw⁎Tw⁎. Thus, dLdt≤0 if the conditions in (34) and (35) hold. Therefore, by Oludoun et al. [34], the positive equilibrium state is globally asymptotically stable in the positive region R+4. Sensitivity analysis This section outlines the method Oludoun et al. [34] used to identify the parameter values that significantly affect the fundamental reproduction number. Sensitivity analysis is important not only for exploratory research design only but also for understanding how important each parameter is to the spread of disease. Sensitivity analysis is mostly used to assess how well model predictions hold up to various parameter values. This will establish a direct link between the prevalence of water pollutants and the endemic equilibrium point, as well as between the early transmission of water pollutants and basic reproduction numbers. This aids in our understanding of the relative importance of each characteristic for the prevalence and transmission of contaminants in water. Extremely sensitive parameters need to be evaluated carefully because even a small deviation in one of these parameters can have a significant quantitative impact on the quantity of interest and, in turn, affect the qualitative outcomes. Conversely, determining the sensitivity index of an insensitive parameter doesn't need a lot of effort because even a small divergence from the parameter won't significantly alter the number of interests. To reduce the amount of water pollutants spreading into water sources, these factors ought to be the main consideration while figuring out how to regulate water pollution. Explanation of sensitivity indices Sensitivity indices are estimated to determine how a state variable varies in response to parameter changes. A normalized forward can be used to determine R0 sensitivity index to its dependent parameter. The normalized forward sensitivity index is the ratio of a variable's relative change to that of a parameter. Partial derivatives can be used to define the sensitivity index for variables that are differentiable functions of the parameter. DefinitionThe normalized forward sensitivity index of a differentiable variable R0 is defined as follows:ħℓR0=∂R0∂ℓℓR0. The sensitivity indices of R0 with reference to α;β;φ and ϵ denoted by ħαR0;ħβR0;ħφR0 and ħϵR0 respectively, are shown below. These indices denote the usefulness of each parameter in determining pollutant water transmission and incidence.ħαR0=∂R0∂α×αR0=1,ħβR0=∂R0∂β×βR0=1,ħφR0=∂R0∂φ×φR0=φ2(φ+ϵ)(−αβφ(φ+ϵ)2−αβφ2(φ+ϵ)),ħϵR0=∂R0∂ϵ×ϵR0=−ϵφ+ϵ, Table 1 described the sensitivity index of R0, which measures how changes in various parameters affect the basic reproduction number R0. A positive sensitivity index means that an increase in the parameter will increase R0, while a negative index indicates that an increase in parameter will decrease R0. Sensitivity analysis helps prioritize which parameters have the most significant influence on controlling the spread of an infection, guiding public health interventions.Table 1Sensitivity Index of R0 for some estimated parameters.Table 1ParametersSensitivity indexα1β1φ-0.66667ϵ-0.33333 The normalized forward sensitivity index of a differentiable variable R0 is defined as follows:ħℓR0=∂R0∂ℓℓR0. The sensitivity indices of R0 with reference to α;β;φ and ϵ denoted by ħαR0;ħβR0;ħφR0 and ħϵR0 respectively, are shown below. These indices denote the usefulness of each parameter in determining pollutant water transmission and incidence.ħαR0=∂R0∂α×αR0=1,ħβR0=∂R0∂β×βR0=1,ħφR0=∂R0∂φ×φR0=φ2(φ+ϵ)(−αβφ(φ+ϵ)2−αβφ2(φ+ϵ)),ħϵR0=∂R0∂ϵ×ϵR0=−ϵφ+ϵ, Table 1 described the sensitivity index of R0, which measures how changes in various parameters affect the basic reproduction number R0. A positive sensitivity index means that an increase in the parameter will increase R0, while a negative index indicates that an increase in parameter will decrease R0. Sensitivity analysis helps prioritize which parameters have the most significant influence on controlling the spread of an infection, guiding public health interventions.Table 1Sensitivity Index of R0 for some estimated parameters.Table 1ParametersSensitivity indexα1β1φ-0.66667ϵ-0.33333 Table 1Sensitivity Index of R0 for some estimated parameters.Table 1ParametersSensitivity indexα1β1φ-0.66667ϵ-0.33333 Sensitivity Index of R0 for some estimated parameters. Numerical simulation The following parameter values α=0.8, β=0.18, γ=0.02, ϕ=0.25, φ=0.4, λ=0.3, ϵ=0.2, and δ=0.5 as taken from Shah et al. [14] while some where assumed will be used for the numerical analysis and simulation to investigate the dynamics of the system of equations (1)-(4). The maple software is used for the simulation at the initial conditions Pw0=500, Sw0=400, Iw0=100, and Tw0=0. Fig. 2 illustrates the transmission dynamics of polluted water. It can be seen from the figure that the increasing water pollutants do not have any significant effect as the days go by, this signifies that polluted water sources keeps increasing if measures are not put in place. An increasing water pollutants do not show significant immediate effects over time due to factors like pathogen decay, population immunity, or natural dilution, the underlying long-term risks remain significant. If left unchecked, increasing pollution could lead to chronic health issues, environmental degradation, and even delayed disease outbreaks under the right conditions. Fig. 3 shows the transmission dynamics of the susceptible soluble water, the plot indicates a decreasing soluble water pollutants as the time increases which may be as a result of the availability of treatment present in the soluble water pollutants. An initial rise in pollutant levels as contamination enters the water system (e.g., from runoff, industrial waste, or improper waste disposal). However, as time progresses, treatment processes reduce the concentration of soluble pollutants, leading to a gradual decline in the amount of contamination, which suggests that treatment or other mitigation efforts are effectively reducing contamination over time. Also a similar dynamics can be noticed in Fig. 4, i.e. a decreasing insoluble water pollutant was seen. This could be as a result of effective treatment and awareness on not dumping insoluble waste to the water ways which could be the best management practice or insoluble water pollutants. Likewise, some insoluble organic pollutants may be broken down by microorganisms over time. However, this process is slower and more dependent on environmental conditions such as temperature, oxygen levels, and microbial presence. The key challenge is ensuring these pollutants are not just removed from the water column but also effectively managed in the long term to avoid environmental buildup. It is displayed from Fig. 5 an upward spike in the plot at the point 0≤t≤2 that which may be as a result of positive adherence to treatment but a continuous decrease in the plot as the time increases was shown which could be as a result of negligence to treatment. Meanwhile, treatment negligence of water pollutants, whether soluble or insoluble, can lead to severe long-term consequences for public health, the environment, and even the economy. If treatment efforts are neglected, especially in managing waterborne pollutants, the impacts could include increasing levels of contaminants, a resurgence of waterborne diseases, and irreversible damage to ecosystems.Figure 2Transmission dynamics for Pw.Figure 2Figure 3Propagation dynamics for Sw.Figure 3Figure 4Proliferation dynamics for Iw.Figure 4Figure 5Transmission dynamics for Tw.Figure 5 Transmission dynamics for Pw. Propagation dynamics for Sw. Proliferation dynamics for Iw. Transmission dynamics for Tw. Fig. 6 depicts the effect of water pollution transmission β on the soluble water compartments. It is shown from figure that as the rate of soluble water pollutants β increases, the soluble water pollutants decrease. The increasing rate of β increases the viability of soluble water pollutants in the water sources which poses threat to the water bodies and makes the water harmful for consumption and domestic uses. Fig. 7 reveals the effect of water pollution transmission γ on the insoluble water compartments, it can be seen that an increase in the rate of insoluble water pollutants increases the rate of the insoluble water pollutants γ. As γ increases, the water ways becomes dirty as a result of an increase in the insoluble water pollutants and this causes environmental degradation and consequential puts the lives of both humans and animals at risk of germs and probably extinction of some aquatic animals.Figure 6Soluble pollutant spread for β.Figure 6Figure 7Impact of γ on insoluble pollutant.Figure 7 Soluble pollutant spread for β. Impact of γ on insoluble pollutant. Fig. 8 illustrates the effect of treatments and controls on the soluble pollutants. It can be seen from plot that pollutants take longer to vanish without control unlike the control plot which treatment eradicates the pollutants just after 4 months. Also, Fig. 9 revealed the effect of treatments and controls on the insoluble water pollutants, the figure depicts that insoluble water pollutants diminished and got eradicated at 6 months but the control plot the pollutants were eradicated shortly after 2 months. This suggests that improving water quality should be encouraged in order to lower the prevalence of water pollution and consequential reduce the emerging of water borne diseases.Figure 8Optimal control effect on Sw.Figure 8Figure 9Insoluble pollutant Iw control.Figure 9 Optimal control effect on Sw. Insoluble pollutant Iw control. Fig. 10 shows the effect of varying treatment of soluble water pollutants ϵ on the soluble water pollutants compartment. It can be seen clearly at as the treatment rate of soluble water pollutant is increased, it decreases the water pollutants and this will eventually fizzle out over time. Also Fig. 11 depicts the effect of varying treatment of insoluble water pollutants δ on the insoluble water compartment. The plots illustrate that as the treatment rate of insoluble water pollutant is increased, the pollutants decrease over time. This shows that improving treatment will help improves the water from both soluble and insoluble water pollutants and this eventually gives safe water for drinking and aquatic animals will thrive in safe water.Figure 10Impact of ϵ on soluble pollutant.Figure 10Figure 11Rising effect of δ on insoluble.Figure 11 Impact of ϵ on soluble pollutant. Rising effect of δ on insoluble. Conclusion The study focuses on the causative factors of water pollution which have been identified in this study as the soluble water pollutants and the insoluble water pollutants arising from natural occurrences due to anthropogenic that resulted in eutrophication needs to be guided. Indiscriminate harmful waste discharge and dispersion into the water bodies leads to water pollution. As such, constant management and monitoring of water resources for an enhanced quality and quantity is essential. Therefore, this study employed a coupled derivative mathematical model to investigate soluble and insoluble water pollutants. The insoluble pollutants are converted to soluble pollutants through an applied control or treatment. To optimize the water quality, parametric sensitivities in relation to reproduction number are examined. From the computational analysis carried out, it was revealed that treatment still remains the only remedy for safe water. This means that, in order to minimize pollution, treatment of water pollutants before it is discharged into water bodies should be strictly adhered to. Sensitivity analysis conducted on the advection-diffusion parameters indicated that the dispersion of soluble pollutants was most sensitive to changes in flow velocity, while insoluble pollutants were highly affected by variations in sedimentation rates and particle size distribution. Future research could expand on this study by developing more sophisticated multiphase flow models that capture a wider range of pollutant behaviors, including gas-phase pollutants or pollutants with more complex chemical reactions in water. This could further improve the prediction and control of pollutant dispersion in diverse environments, such as industrial waterways or areas affected by chemical spills. CRediT authorship contribution statement O.Y. Oludoun: Writing – original draft. S.O. Salawu: Conceptualization. S.O. Adesanya: Writing – review & editing. O.E. Abiodun: Data curation. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.