ed

Mathematical Modeling of Vehicle Carbon Dioxide

Emissions

iew
Pita Donalda,b,∗, Maranya Mayengoa , Aristide G. Lamburac
a School of Computational and Communication Science and Engineering, Nelson Mandela
African Institution of Science and Technology (NM-AIST), P.O. Box 447, Arusha,
Tanzania
b Department of Mathematics, Humanities and Social Sciences, National Institute of

ev
Transport(NIT), [Link] 7O5, Dar es Salaam, Tanzania
c Department of Computer Systems and Mathematics, Ardhi University (ARU), [Link]

35176 , Dar es Salaam, Tanzania

r
Abstract

er
In this study, we present a nonlinear mathematical model designed to inves-
tigate the dynamics of atmospheric carbon dioxide (CO2 ) concentration, with a
pe
specific focus on the influence of emissions from transportation activities. The
substantial role of greenhouse gas emissions, particularly CO2 , in driving cli-
mate change has been widely recognized. The model integrates both natural
and anthropogenic sources of atmospheric CO2 , allowing us to capture the intri-
cate dynamics involved. We consider various factors such as human population,
ot

economic activities, forest biomass, and vehicle production, recognizing their

interconnectedness and impact on atmospheric CO2 levels. Through rigorous
tn

stability analysis and numerical simulations, we gain insights into the long-term
behavior of the model. Our findings reveal that increased rates of vehicle produc-
tion and the presence of vehicles with high emission rates contribute to elevated
atmospheric CO2 concentrations. This relationship underscores the critical role
rin

of the transportation sector in CO2 emissions and subsequent climate change.

Keywords: Mathematical Model, Vehicular emission, Greenhouse gas,
Atmospheric Carbon dioxide, Climate change, Sustainable transportation.
ep

∗ Correspondingauthor
Email address: donaldp@[Link] (Pita Donald)
Pr

Preprint submitted to Results in Applied Mathematics June 13, 2023

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
1. Introduction

iew
Greenhouse gases, among which carbon dioxide (CO2 ) is a prominent ex-
ample, play a significant role in the greenhouse effect, which is a key driver of
climate change. These gases are emitted by diverse economic sectors, encom-
passing energy generation, industrial processes, building operations, transporta-
tion activities, and agricultural practices, as well as forestry and other land use

ev
activities (AFOLU). Among these sectors, transportation plays a substantial
role in CO2 emissions due to the burning of fossil fuels [10, 11, 22]. The con-
centration of CO2 in the atmosphere has been increasing primarily due to the

r
consumption of fossil fuels. As a result, climate-related issues such as ozone
layer depletion, droughts, and global warming have become more pronounced.

er
It is worth noting that while all sectors contribute to CO2 emissions, trans-
portation is a significant contributor [10, 11]. The expanding global population
pe
and the corresponding economic development have generated a heightened need
for energy, predominantly derived from fossil fuel sources. This heightened en-
ergy demand has resulted in amplified carbon dioxide (CO2 ) emissions, further
aggravating the challenges posed by climate change [16, 6]. As the population
continues to grow, the use of fossil fuels and natural gas for energy production
ot

also increases, resulting in higher CO2 emissions. In summary, the increase

in CO2 emissions from human activities, especially in the transportation sec-
tn

tor, has a significant impact on the Earth’s climate. The correlation between
population growth, economic development, and CO2 emissions underscores the
importance of transitioning to cleaner and more sustainable energy sources,
reducing reliance on fossil fuels, and adopting environmentally friendly trans-
rin

portation practices. Transportation is indeed a significant consumer of fossil

fuels as an energy source, and urban transport alone accounted for 23% of total
energy-related emissions in 2010 [22]. According to IEA [9], the global trans-
ep

portation sector contributes approximately 25% of total CO2 emissions from

fossil fuel burning, which increases the global volume of atmospheric CO2 [6].
Despite advancements in vehicle technology and fuel economy, CO2 emissions
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
from the urban transportation sector are projected to double by 2050 [9].

iew
While mathematical models have been developed to study the dynamics of
atmospheric CO2 , previous models have primarily focused on CO2 emissions
from the energy sector. For instance, Verma et al. [21] employed a nonlin-
ear mathematical model to analyze CO2 dynamics from the energy sector and
emphasized the importance of using low-carbon energy sources to minimize mit-

ev
igation costs. Other studies have explored various strategies such as genetically
engineered plants and education or incentives for energy conservation to reduce
CO2 emissions [13, 8]. However, there is currently a gap in the literature re-

r
garding mathematical models specifically exploring the dynamics of atmospheric
CO2 due to vehicular emissions. Hence, our study proposes a mathematical

er
model to investigate the impact of vehicle production on the dynamics of atmo-
spheric CO2 concentration. The structure of this paper is well-organized, with
sections dedicated to formulating the mathematical model, performing qualita-
pe
tive analysis, parameter estimation and numerical simulations, and concluding
the findings.

2. Model Formulation
ot

The system of non-linear ordinary differential equations used to describe the

dynamics of vehicular carbon dioxide emission. The model is made up of five
tn

distinct compartments, such as atmospheric CO2 concentration in relation to

human population (N), human economic activities (H), forest area (B), and ve-
hicle production (V) in the region of consideration at a time t. It is assumed that
the human population grows logistically and depletes due to the adverse impact
rin

of increased atmospheric CO2 concentration [14, 12]. Also, it is assumed that

human population increases due to consumption of forest resources [14]. Human
economic activities are expected to expand in response to population growth, as
ep

it drives increased demand for goods and services, and depletes naturally. The
growth of human activities is assumed to be influenced by the production of
vehicles. As vehicle production increases, it stimulates various economic sectors
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
and promotes overall economic growth. This, in turn, leads to an expansion of

iew
human activities and their impact on the environment. Human economic ac-
tivities, such as industrial and agricultural activities, contribute to the increase
in atmospheric CO2 concentration. Additionally, we assume that the growth of
human economic activities is quantified by the Gross Domestic Product (GDP).
The GDP serves as an indicator of economic performance and reflects the over-

ev
all value of goods and services produced within a country. As GDP increases,
it signifies a higher level of economic activity, including industrial, commercial,
and agricultural sectors.

r
The concentration of CO2 in the atmosphere is assumed to increase due
to the growing number of vehicles and human activities. Additionally, CO2

er
concentration in the atmosphere is naturally depleted through processes such
as ocean sinking and forest biomass, which occurs through the photosynthesis
process [18]. The growth rate of atmospheric CO2 resulting from natural pro-
pe
cesses, such as respiration of living organisms, volcanic eruptions, and changes
in ocean circulation, is assumed to be constant [15]. According to Pal et al. [17],
plants consume CO2 during the photosynthesis process, and increasing forest
cover leads to a depletion of atmospheric CO2 concentration. Forest biomass
ot

is assumed to follow logistic growth and depletes due to population growth. In

the region of consideration, the vehicle population is assumed to have a growth
rate of β and naturally depletes, which is represented by ξ. Consequently, we
tn

formulate the following system of equations of non linear differential equation

as presented in Eq.(1).
rin

  
N
Ṅ = rN 1 − − αCN + ωψN B





 L


Ḣ = sN + τ V − ηH






Ċ = C0 + γV + δH − ρBC − πC (1)
ep


  

 B
 Ḃ = θB 1 − K − ψN B







 V̇ = βN − ξV

Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
where N (0) ≥ 0, H(0) ≥ 0, C(0) > 0, B(0) ≥ 0, and V (0) ≥ 0 and all model

iew
parameters (presented in Table 1 are assumed to be positive.

Lemma 1. The region of attraction is given by Ω = {(N, H, C, B, V ) ∈ R5 :

0 ≤ N ≤ Nm , 0 ≤ H ≤ Hm , 0 < C ≤ Cm , 0 ≤ B ≤ Bm , and 0 ≤ V ≤
Q L(s + βτ )
Vm , N + H + C + B + V ≤ }, where Nm = L, Hm = , Cm =
χ ηξ
C0 ξη + ηγβL + δξL(s + τ ) βL

ev
, Bm = K , Vm = , Q = rN + ωψN B + sN +
ξηρK + π ξ 
r θ
τ V + C0 + γV + σH + θB + βN and χ = min , α, η, ρ, π, . The solution
L K
of system Eq.(1) is bounded within Ω.

r
The proof of lemma 1 is in Appendix D.

Parameter
r
Description er
Table 1: Model parameters and description

Intrinsic human population growth rate

pe
L Human population carrying capacity
α Mortality rate coefficient due to enhanced CO2
s Economic activities growth rate due to population
τ Human economic activities growth rate due to vehicle population
ot

η Human economic activities depletion rate constant

C0 Emission rate of CO2 from natural and respiration processes
γ CO2 vehicle emission rate constant
tn

σ Human economic activities CO2 emission rate

ρ CO2 depletion rate coefficient due to forest biomass
π Natural CO2 depletion rate coefficient
rin

θ Forest biomass intrinsic growth rate coefficient

K Forest area carrying capacity
ψ Deforestation rate coefficient
β Vehicle’s population growth rate due to demand for mobility
ep

ξ Vehicle depletion rate coefficient

ω Human population growth rate coefficient due forest harvesting
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
3. Model analysis

iew
The long-term behavior of the system Eq.(1) was established by using sta-
bility theory. The feasible equilibrium points are obtained and their stability
properties are established.
 The model system Eq.(1)  has two non-negative triv-
C0 θC0
ial solutions E1 = 0, 0, , 0, 0 and E2 = 0, 0, , K, 0 and two
π Kρ + π
non-trivial solutions E3 = (N3 , H3 , C3 , 0, V3 ) and E ∗ = (N ∗ , H ∗ , C ∗ , B ∗ , V ∗ ).

ev
The proofs for existence of equilibria E3 and E ∗ are provided in Appendix A
and Appendix B respectively, existence of both E3 and E ∗ depend on the
satisfaction of condition 2.

r
3.1. Stability Analysis
r
α
−
C0
π
er
>0 (2)
pe
In this section the behaviors of different equilibrium points of the model
system are analyzed. From the system Eq.(1) the Jacobian matrix is given by
   
N rN
r 1 − L − L − αC + ωψB 0 −αN ωξN 0
 
−η
 
s 0 0 τ 
ot


 
J = σ −ρB − π −ρC γ .
 
0
   
 B θB 
 −ψB 0 0 θ 1− − − ψN 0
K K
tn

 
 
β 0 0 0 −ξ
 
C0
At the equilibrium point E1 , 0, 0 the corresponding Jacobian ma-
0, 0,
π
rin

trix JE1 , resulted from evaluating matrix J through E1 is given by

  
r C0

α α − π 0 0 0 0
 
s −η 0 0 τ 
 

ep

−ρC0
 
JE1 = 
 0 σ −π γ .


 π 


 0 0 0 θ 0
β 0 0 0 −ξ
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
The eigenvalues of a matrix JE1 are λ1 = −ξ, λ2 = θ, λ3 = −η, λ4 = −π and
r C0

iew
λ5 = − > 0. The equilibrium point E1 is unstable since it have at least
α π
one positive eigenvalues, that is λ2 > 0 and λ5 > 0.
Similarly, evaluating matrix J at E2 , the following matrix is established.
   
2 r C0
 K ωψρ + Kπψω + Krρ + απ α − π 
 0 0 0 0

ev
ρK + π
 
 
 
 s −η 0 0 τ 
J E2 =  .
 
−C 0ρ
 0 σ −Kρ − π γ

 ρK + π 

−ψK 0 0 −θ 0
 

r

 
β 0 0 0 −ξ

and λ5 =
2
K ωψρ + Kπψω + Krρ + απ

ρK + π
er
The eigenvalues of JE2 are λ1 = −ξ, λ2 = −θ, λ
r
α
−
C0
π
3 = −η, λ4 = −(Kρ + π),

. Clearly, upon utilizing

pe
condition 2, λ5 > 0 making E2 unstable.
At the equilibrium point E3 , we establish a matrix JE3 , such that

rN3
 
− 0 −αN3 ωψN3 0
 L 
 
−η
ot

 s 0 0 τ 
 
JE3 = 0 .
 
 σ −π −ρC3 γ 
 
 0 0 0 − (ψN3 − θ) 0 
tn

 
β 0 0 0 −ξ

θ
N3 > (3)
ψ
rin

Upon satisfaction of condition 3 all the entries at the main diagonal will be
negative, following Devi and Gupta [8], we utilize the Gershgorin’s theorem to
deduce the local stability of E3 . As such, E3 is proved to be asymptotically
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
stable upon satisfaction of conditions in Eq.(4).

iew
1 r


 (s + β) < (2N3 − 1) + C3


 α α




 σ < η

 π

 N3 <
α (4)
 

 θ − ρC 3 θ + ρC3
 N3 > max
 ,
ψ(1 − ω) ψ(1 + ω)

ev







τ <ξ




r
Theorem 2. The equilibrium point E3 , if exists, is locally asymptotically stable
whenever conditions (4) are satisfied.

er
At the interior equilibrium point E ∗ , the Jacobian matrix J, was evaluated
to get a matrix JE ∗ , such that
pe
rN ∗
 
− L 0 −αN ∗ ωψN ∗ 0
 
 s
 −η 0 0 τ 
 
JE ∗ =
 0 σ −ρB ∗ − π −ρC ∗ γ .

 −ψB ∗ θB ∗ 
−
ot

0 0 0
K
 
 
β 0 0 0 −ξ
Clearly, all the entries in the main diagonal of JE ∗ are negative. Following
tn

Devi and Gupta [8], we use the Gershgorin’s theorem to deduce the local stability
of E ∗ . As such, E ∗ is proved to be locally asymptotically stable upon satisfaction
of the following set of conditions
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
L(β + s + r − αC ∗ − ψB ∗ (1 − ω)) L(−β − s + r − αC ∗ + ψB ∗ (1 + ω))
  
∗
N > max ,


2r 2r





σ<η






ρB ∗ + π


N∗ <

 α
ρKC ∗ + θK − 2θB ∗ ρKC ∗ − θK + 2θB ∗


∗
<N <




 Kψ(ω + 1) Kψ(ω − 1)

ev







 τ +γ <ξ

(5)

Theorem 3. The interior equilibrium point E ∗ , if exists, is locally asymptoti-

r
cally stable whenever conditions (5) are satisfied.

er
Theorem 4. The interior equilibrium point E ∗ , if exists, is globally asymptot-
ically stable in Ω if conditions 6 are satisfied.
pe
τ2 γ2
  
k4 > max ,



ηξ ξ (ρBm + π) 




2
ρ


k1 >



η (ρBm + π)  
ot

 2 2
 (6)
β L s L 
r > max , 

ξ η 




2
α L


tn

k2 >



r (ρBm + π)  

The proof of theorem 4 is presented in Appendix C.

rin

3.2. Ecological interpretation of stability analysis

Equilibrium points E1 and E2 represent different situations regarding the ex-

tinction of the human population and the presence or absence of forest biomass.
The point E1 signifies the extinction of forest biomass, while E2 represents a
ep

situation where forest biomass exists. The absence of human population would
lead to the cessation of human-related activities, resulting in a lack of stabiliza-
tion in atmospheric CO2 concentration. Only natural processes such as volcanic
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
eruptions will contribute to CO2 emissions, resulting in a natural depletion of

iew
atmospheric CO2 concentration. However, the analysis reveals further that the
forest will rapidly grow up to its carrying capacity, acting as a sink for CO2
absorption. The atmospheric CO2 concentration will solely depend on natural
processes and will naturally deplete.
The interior equilibrium point E ∗ considers the coexistence of human pop-

ev
ulation, economic activities, atmospheric CO2 concentration, forest area, and
vehicle production. For this equilibrium to occur, the intrinsic growth rate of the
human population must exceed the mortality rate caused by natural CO2 emis-

r
sions. Additionally, certain conditions must be met to stabilize the atmospheric
CO2 concentration. When these conditions are satisfied, the forest biomass

er
absorbs the available atmospheric CO2 , preventing harm to the environment,
while the remaining CO2 naturally depletes. These ecological interpretations
emphasize the role of human activities, natural processes, and the importance
pe
of maintaining conditions for the stability of atmospheric CO2 concentration
and environmental balance.

4. Parameter estimation, numerical simulation and Discussion

ot

The parameter estimation process involves adjusting the model parameters

to best fit the available annual data for the world population [3], atmospheric
tn

CO2 concentration [2], forest area [1], number of vehicles produced [5], and
Gross Domestic Product (GDP) [4]. The specific values of the parameters were
obtained through the least squares method, which minimizes the sum of squared
differences between the model predictions and the actual data. In addition to the
rin

data-driven parameter estimation, some of the parameters were obtained from

literature sources as cited in Table 2. These sources provide information on the
values and units of the parameters, allowing for their inclusion in the model. For
ep

the intrinsic human population growth rate (r), data from [3] was considered,
where a range of 0.02 to 0.03 per year was reported by Verma and Verma
[20]. This range provides an estimate for the value of r that is consistent with
Pr

10

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
the available data and existing literature. Overall, the parameter estimation

iew
process combines data analysis and literature review to determine the most
appropriate values for the model parameters, ensuring that the model aligns
well with the observed trends and dynamics of the real-world data. The initial
time is set to 1994, and the starting values are N (0) = 5.66315 billion people,
H(0) = 2, 787.67662 billion US$, C(0) = 358.96 ppm, B(0) = 4, 099.2026 million

ev
hectares, and V (0) = 49.658288 million vehicles.
Real and model data on human population, economic activity, atmospheric
CO2 concentration, forest area, and vehicle manufacturing were analyzed, re-

r
sulting in R-squared values of 0.9988, 0.9467, 0.9862, 0.9103, and 0.9270, re-
spectively. These high R-squared values indicate a strong correlation between

er
the actual data and the model predictions. The comparison between the real
and model data is shown in Figure 1. The close alignment between the model
projections based on system Eq.1 and the actual data suggests that the model
pe
accurately represents the reality. To validate the feasibility of our model anal-
ysis, we utilized MATLAB R2021a and Maple 2015 software. The chosen set
of parameters from Table 2 satisfied the condition (2) for the existence of the
interior equilibrium point E ∗ . The components of the interior equilibrium point
ot

E ∗ were determined to be N = 11.06 billion people, H = 95.148 billion US$,

C = 88.516 ppm, B = 7.498 × 105 million hectares, and V ∗ = 75.528 million
vehicles.
tn

The eigenvalues of the Jacobian matrix corresponding to the equilibrium

point E ∗ were calculated to be −0.0319, −0.013, −0.0196, −0.00000181, and
−0.00004061, all of which are negative. This indicates that the equilibrium
rin

point E ∗ is locally asymptotically stable. Furthermore, the solution trajectories

of the model Eq.(1) were plotted in Figure (3) with different starting values.
It can be observed that all trajectories initiated inside the region Ω converge
towards the interior equilibrium point E ∗ . This demonstrates the nonlinear sta-
ep

bility behavior of the point E ∗ in the C(t) − B(t) − N (t) and B(t) − C(t) − H(t)
space, as depicted in Figure (3).
To examine the impact of increased vehicular CO2 emissions on the human
Pr

11

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
population, we can vary the parameter γ in the model. Figure 2 illustrates

iew
the assumption that human population declines, due to the adverse impacts
of increased CO2 emissions from vehicles. This indicates that the increased
emissions have detrimental effects on the human population, leading to a decline
in population growth over time. Similarly, Figure (4(a)) shows the impact
of increased vehicle manufacturing on the time evolution of atmospheric CO2

ev
concentration, as depicted in Figure (4(b)).
Additionally, according to Figure 5, when the emission rate from economic
activities is doubled, for instance when t = 60 years, it results in an increase in

r
atmospheric CO2 concentration from 391.541 ppm to 452.91 ppm. This increase
occurs while the forest biomass is growing towards its carrying capacity. How-

er
ever, once the forest biomass reaches its carrying capacity, the atmospheric CO2
concentration starts to stabilize. This observation indicates that doubling the
emission rate from economic activities has a direct impact on the atmospheric
pe
CO2 levels, especially during the growth phase of forest biomass. However, as
the forest biomass reaches its limit, its capacity to absorb CO2 becomes satu-
rated, leading to a stabilization of atmospheric CO2 concentration.
It is evident that the atmospheric CO2 concentration increases with higher
ot

levels of vehicle production and economic activities. These figures demonstrate

the sensitivity of the model to changes in vehicular CO2 emissions and highlight
the potential consequences for both the human population and atmospheric CO2
tn

concentration.
rin
ep
Pr

12

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
ev
Table 2: Estimated model parameter values, units and their respective source
Parameter Value Units Source
r 0.026 per year [20]
α 0.000001 per year [19]

r
s 0.035 billion $ per year [7]
τ 0.0000394 billion $ per million vehicles

η 0.0041
er
per year
billion $ per year
Estimated
Estimated
pe
C0 1.68 ppm per year [19]
γ 4.9337 × 10−7 ppm per year Estimated
L 11 billion persons [20]
σ 0.0054 × 10−3 ppm per year Estimated
−9
ρ 4.8 × 10 million Ha per year [19]
ot

π 0.016 ppm per year [19]

θ 0.0523 Million Ha per year Estimated
tn

β 0.0729 million vehicles per year Estimated

K 750000 million Ha [19]
ψ 0.71 × 10−6 million Ha per year Estimated
−6
ξ 1.0676 × 10 million vehicles per year Estimated
rin

ω 0.001 billion person per year Estimated

ep
Pr

13

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
7
6.5
6

ev
1994 1999 2004 2009 2014 2019

8000

r
6000

4000

1994 1999
er2004 2009 2014 2019
pe
400

350
1994 1999 2004 2009 2014 2019
ot

4100

4050
tn

4000
1994 1999 2004 2009 2014 2019
rin

100

50
ep

1994 1999 2004 2009 2014 2019

Model fit Real data

Figure 1: Comparison between actual and model data for 22 years (from 1994 to 2016)
Pr

14

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
r ev
er
pe

Figure 2: Impact of increased vehicular CO2 emission

ot
tn
rin
ep

Figure 3: Graphical representation of non-linear stability of E ∗

Pr

15

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
10 5
7 480

6 460

5 440

4 420

ev
3 400

2 380

1 360

r
0 340
0 50 100 150 200 0 50 100 150 200

er
Figure 4: Effect of change in vehicle production rate on the time evolution of CO2 , figure 4
(a) while other parameters remains the same as listed in table 2
pe
12 800

11
700

10
600
ot

500
8

7 400

6
300
tn

0 200 400 600 800 1000 0 200 400 600 800 1000

10 5
8 3100

7
3050
6
3000
5
rin

4 2950

3 2900

2
2850
1
2800
0
0 200 400 600 800 1000 0 200 400 600 800 1000
ep

Figure 5: Variation of economic activities emission rate (σ), while other parameters remains
the same as listed in table 2
Pr

16

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
5. Conclusion

iew
In summary, this study developed a mathematical model to understand and
analyze the dynamics of vehicular CO2 emissions and their impact on atmo-
spheric CO2 concentration. The model considered both natural and anthro-
pogenic sources of CO2 and explored the existence and stability of equilibrium
points. The results indicated that for sustainable economic development, the

ev
intrinsic human population growth rate should be lower than the mortality rate
due to natural CO2 emissions. The study also investigated the influence of
various factors, such as vehicle production rate, vehicle emission rate, and emis-

r
sions from other human activities, on atmospheric CO2 dynamics. Numerical
simulations revealed that increasing vehicle production rates led to higher at-

er
mospheric CO2 concentrations, which in turn had adverse effects on the human
population. Without controlling vehicular CO2 emissions, the study projected
pe
a decline in the world population due to the negative impacts of increased CO2 ,
including global warming and associated climate-related problems. To mitigate
vehicular CO2 emissions, the study proposed several measures. These included
promoting clean energy sources for vehicles, reducing vehicle ownership through
the introduction of shared mobility systems, and addressing deforestation to en-
ot

hance the capacity of forest biomass to store CO2 . The study also suggested
areas for future research, such as considering the impact of electric vehicles and
tn

alternative transportation modes on atmospheric CO2 dynamics. Addition-

ally, incorporating other greenhouse gases and their sources, such as methane
emissions, and examining feedback loops between atmospheric CO2 and envi-
ronmental factors like ocean acidification and climate patterns would enhance
rin

the model’s comprehensiveness.

Overall, the study underscored the importance of tackling vehicular CO2
emissions and implementing sustainable transportation and land-use policies to
ep

mitigate the adverse effects of CO2 on the environment. Further research and
action are necessary to address this issue and promote a more sustainable and
resilient future.
Pr

17

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
Appendix A. Existence of E3

iew
To prove the existence of E3 we solved the system Eq. A.1 to obtain alge-
braic equations for N3 , H3 , C3 and V3 by means of intermediate value theorem.
Consider the system Eq. A.1 below,
  
N3
r 1− − αC3 = 0


L



ev



sN3 + τ V3 − ηH3 = 0

. (A.1)
C0 + γV3 + δH3 − πC3 = 0 





βN − ξV = 0 

3 3

r
From the fourth equation of system Eq. A.1 we have
βN3
V3 =

er ξ
. (A.2)

Similarly, using the third equation and appropriate substitution of Eq. A.2 we
get,
pe
ηξC0 + (γβη + σ (sξ + τ β)) N3
C3 = = f (N3 ). (A.3)
πξη
Using second equation and appropriate substitutions, the value of H3 is
established such that,
ot

sξN3 + τ βN3
H3 = . (A.4)
ηξ
On the other hand, the first equation provides g(N3 ) such that
tn

 
N3
g(N3 ) = r 1 − − αf (N3 ). (A.5)
L
By using equation Eq. A.5, we analyze g(N3 ) in three different scenarios as
rin

follows:

(a) When N3 = L we have the following

 
ηξC0 + γβηL + σ (sξ + τ β) L
g(L) = −α < 0.
πηξ
ep

(b) At N3 = 0, a function g(N3 ) is given by

 
r C0
g(0) = α − .
α π
Pr

18

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
Since g(L) < 0, utilizing the Intermediate Value Theorem, there exist a

iew
positive root N3 in the interval of 0 < N3 < L if and only if g(0) > 0.
Clearly, this requirement is met upon satisfaction of condition Eq. 2.

(c) For uniqueness of N = N3 , we need to prove that g ′ (N3 ) < 0. Considering

equation Eq. A.5 and some algebraic simplifications, it can be easily shown
that

ev
 
r γβη + σ (sξ + τ β)
g ′ (N3 ) = − +α < 0.
L πξη

Thus, a unique positive root, say N = N3 of the equation Eq. A.5 lies on

r
interval 0 < N3 < L. Substituting N = N3 into equations Eq. A.2–A.4 we
get the appropriate values of V = V3 , C = C3 and H = H3 . Hence, it can be

er
concluded that an equilibrium point E3 exists, whenever the condition Eq. 2
holds.
pe
Appendix B. Existence of interior equilibrium point E ∗

Interior equilibrium point E ∗ of the system 1 can be obtained by solving

the system Eq. B.1 and its existence is shown by using Intermediate Value
Theorem.
ot

N∗
  
∗ ∗
r 1− − αC + ωψB =0 

L





sN ∗ + τ V ∗ − ηH ∗

=0 
tn




C0 + γV ∗ + δH ∗ − ρB ∗ C ∗ − πC ∗ =0 (B.1)
B∗
  


θ 1− − ψN ∗ =0 

K





βN ∗ − ξV ∗

=0 
rin

The solution of the system Eq. B.1 establishes the following results

βN ∗
V∗ = (B.2)
ξ
ep

(sξ + τ β)N ∗
H∗ = (B.3)
ξη
Pr

19

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
iew
K
B∗ = (θ − ψN ∗ ) (B.4)
θ

ηC0 ξ + ηγβN ∗ + σN ∗ (sξ + τ β)

C∗ = (B.5)
ηξ(ρB ∗ + π)

ev
N∗
 
∗ ∗ α
h(N , B ) = r 1 − − (ξηC0 + ξγβN ∗ + σN ∗ (sξ + τ β)) + ωψB ∗
L ηξπ
(B.6)

r
From equation (B.6) when no forest resources (B ∗ = 0) we get the following
function of N ∗

∗

f (N ) = r 1 −
N∗
L

−
α
ηξπ er
(ξηC0 + ξγβN ∗ + σN ∗ (sξ + τ β)) .

By using equation Eq. B.7 f (N ∗ ) can be analyzed as follows,

(B.7)
pe
(a) We evaluate a function f (N ∗ ) at N ∗ = 0 to get the following
 
r C0
f (0) = α − >0
α π

(b) When N ∗ = L we have the following

ot

α
f (L) = − (ξηC0 + ξγβL + σL(sξ + τ β)) < 0
ηξπ
Utilizing the Intermediate Value Theorem, there exists a positive root
tn

N = N ∗ of equation B.7 in the interval (0, L) under condition (2).

(c) For uniqueness of N = N ∗ , we need f ′ (N ∗ ) < 0, that is

 
r α
f ′ (N ∗ ) = − + (ξγβ + σ(sξ + τ β)) < 0
rin

L ηξπ

Thus, a unique positive root, say N ∗ of the equation B.7 lies on interval 0 <
N ∗ < L. Substituting the obtained value of N ∗ into equations Eqs. B.2, B.3
and B.4 and performing algebraic simplifications we get the appropriate values
ep

of V ∗ , B ∗ and H ∗ . Since the values of N ∗ and B ∗ are known, then to obtain

the value of C ∗ we substitute N ∗ and B ∗ into equation B.5 and we conclude
that an interior point E ∗ exists under condition 2.
Pr

20

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
Appendix C. Global stability of E ∗

iew
To analyze the global stability of the interior equilibrium E ∗ , we use Lya-
punov stability theory. Consider the following scalar valued positive definite
function
N∗
 
1 1
U = N − N ∗ − N ∗ ln + k1 (H − H ∗ ) + k2 (C − C ∗ )
N 2 2
(C.1)

ev
∗
 
1 B 1
+ k3 B − B ∗ − B ∗ ln + k4 (V − V ∗ )
2 B 2
Where k1 , k2 , k3 and k4 are suitable positive constants to be determined later.
Differentiating U with respect to t;

r
dU r 2
= − (N − N ∗ ) − α (C − C ∗ ) (N − N ∗ ) + ψω (B − B ∗ ) (N − N ∗ )
dt L

2
−k2 (ρB + π) (C − C ∗ ) −
K
er
+sk1 (H − H ∗ ) (N − N ∗ ) + τ k1 (V − V ∗ ) (H − H ∗ ) − ηk1 (H − H ∗ )
+γk2 (V − V ∗ ) (C − C ∗ ) + σk2 (H − H ∗ ) (C − C ∗ )
k3 θ 2
(B − B ∗ ) − k3 ψ (N − N ∗ ) (B − B ∗ )
2
pe
2
+βk4 (N − N ∗ ) (V − V ∗ ) − ξk4 (V − V ∗ )
(C.2)

Performing algebraic operations we get the following;

dU r 2 2 2 k3 θ 2
= − (N − N ∗ ) − ηk1 (H − H ∗ ) − k2 (ρB + π) (C − C ∗ ) − (B − B ∗ )
ot

dt L K
2
−ξk4 (V − V ∗ ) + (ω − k3 ) (B − B ∗ ) (N − N ∗ ) − α (C − C ∗ ) (N − N ∗ )
+sk1 (H − H ∗ ) (N − N ∗ ) + τ k1 (V − V ∗ ) (H − H ∗ ) + γk2 (V − V ∗ ) (C − C ∗ )
+σk2 (H − H ∗ ) (C − C ∗ ) + βk4 (N − N ∗ ) (V − V ∗ )
tn

(C.3)

Choosing k3 = ω we have the following;

dU r 2 2 2 k3 θ 2
− (N − N ∗ ) − ηk1 (H − H ∗ ) − k2 (ρB + π) (C − C ∗ ) − (B − B ∗ )
rin

=
dt L K
2
−ξk4 (V − V ∗ ) − α (C − C ∗ ) (N − N ∗ ) + sk1 (H − H ∗ ) (N − N ∗ )
+τ k1 (V − V ∗ ) (H − H ∗ ) + γk2 (V − V ∗ ) (C − C ∗ ) + σk2 (H − H ∗ ) (C − C ∗ )
+βk4 (N − N ∗ ) (V − V ∗ )
ep

(C.4)

dU
Thus is a negative definite inside the region of attraction ’Ω’, provided that
dt
conditions 6 holds.
Pr

21

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
Appendix D. Proof of Lemma 1

iew
Adding equations of the system Eq. 1 we have the following

dN dH dC dB dV N


+ + + + = rN 1 − L− αCN + ωψN B + sN + τ V − ηH
dt dt dt dt dt
B


+C0 + γV + δH − ρBC − πC + θB 1 − K

ev
−ψN B + βN − ξV
≤ rN + ωψN B + sN + τ V + C0 + γV + σH + θB
−rN 2
+βN − − αCN − ηH − ρBC − πC
2
L
θB
− − ψN B − ξV

r
K
(D.1)

Let χ = min
r
L
σH + θB + βN .

, α, η, ρ, π,
θ
K


er
and Q = rN + ωψN B + sN + τ V + C0 + γV +
pe
This implies that

Q
lim sup {N (t) + H(t) + C(t) + B(t) + V (t)} ≤
t→∞ χ

Acknowledgments
ot

Pita Donald gratefully acknowledges The Nelson Mandela African Institu-

tion of Science and Technology (NM-AIST) and National Institute of Transport
tn

(NIT).

Declaration of competing interest

rin

The authors assert that they do not have any identifiable financial conflicts
of interest or personal connections that could have potentially influenced the
findings presented in this research paper.
ep

Funding

No any fund received in accomplishing this study.

Pr

22

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
References

iew
[1] Food and Agriculture Organization (FAO). Forest area (sq. km),Food and
Agriculture Organization, electronic files and web site. [Link]
[Link]/indicator/[Link].K2. Accessed: 2022-12-16.

[2] National Oceanic and Atmospheric Administration (NOAA). Trends in

ev
Atmospheric Carbon Dioxide, Mauna Loa CO2 annual mean data. https:
//[Link]/ccgg/trends/[Link]. Accessed: 2022-12-16.

[3] World Bank Open Data (WBOD). population, total. [Link]

r
[Link]/indicator/[Link], . Accessed: 2022-12-16.

[4] World Bank Open Data (WBOD). GDP (current US dOllar), World Bank

er
national accounts data, and OECD National Accounts data files. https:
//[Link]/indicator/[Link], . Accessed: 2022-
pe
12-16.

[5] Beureau of Transportation Statistics. World Motor Ve-

hicle Production. [Link]
world-motor-vehicle-production-selected-countries. Accessed:
ot

2022-12-16.

[6] M. Appiah, F. Li, and B. Korankye. Modeling the linkages among

CO2 emission, energy consumption, and industrialization in sub-Saharan
tn

African (SSA) countries. Environmental Science and Pollution Research,

2021.

[7] Marco Antonio Leonel Caetano, Douglas Francisco Marcolino Gherardi,

rin

and Takashi Yoneyama. Optimal resource management control for CO2

emission and reduction of the greenhouse effect. ecological modelling, 213
(1):119–126, 2008.
ep

[8] Sapna Devi and Nivedita Gupta. Dynamics of carbon dioxide gas (CO2):
Effects of varying capability of plants to absorb CO2. Natural Resource
Modeling, 32(1):e12174, 2019.
Pr

23

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
[9] IEA. CO2 emissions from fuel combustion 2019 edition. IEA CO2 emissions

iew
from fuel combustion database. Available at: [Link] iea. org, 2019.

[10] William F. Lamb, Thomas Wiedmann, Julia Pongratz, Robbie Andrew,

Monica Crippa, Jos G. J. Olivier, Dominik Wiedenhofer, Giulio Mattioli,
Alaa Al Khourdajie, Jo House, et al. A review of trends and drivers of
greenhouse gas emissions by sector from 1990 to 2018. Environmental re-

ev
search letters, 16(7):073005, 2021.

[11] Yi Li, Jiahui Shang, Chi Zhang, Wenlong Zhang, Lihua Niu, Longfei Wang,

r
and Huanjun Zhang. The role of freshwater eutrophication in greenhouse
gas emissions: A review. Science of The Total Environment, 768:144582,
2021.
er
[12] Anthony J. McMichael, Rosalie E. Woodruff, and Simon Hales. Climate
change and human health: present and future risks. The Lancet, 367(9513):
pe
859–869, 2006.

[13] A. K. Misra and Kusum Lata. A mathematical model to achieve sustainable

forest management. International Journal of Modeling, Simulation, and
ot

Scientific Computing, 6(04):1550040, 2015.

[14] A. K. Misra and Maitri Verma. A mathematical model to study the dy-
namics of carbon dioxide gas in the atmosphere. Applied Mathematics and
tn

Computation, 219(16):8595–8609, 2013.

[15] A. K. Misra, Maitri Verma, and Ezio Venturino. Modeling the control
of atmospheric carbon dioxide through reforestation: effect of time delay.
rin

Modeling Earth Systems and Environment, 1(3):1–17, 2015.

[16] Ahmad Mohammadi, Ali Ahmad Burhan, and Rahmatullah Mangal. Im-
pact of Population and Economic Growth on CO2 Emission (Case of
ep

Afghanistan). Journal of Emerging Technologies and Innovative Research

(JETIR), 2020.
Pr

24

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]
 ed
[17] Saheb Pal and Indrajit Ghosh. Dynamics of a coupled socio-environmental

iew
model: An application to global CO2 emissions. Ecological Modelling, 478:
110279, 2023.

[18] Stephen A. Roosa and Arun G. Jhaveri. Carbon Reduction Technologies.

In Energy Management Handbook, pages 561–581. River Publishers, 2020.

ev
[19] Maitri Verma and A. K. Misra. Optimal control of anthropogenic carbon
dioxide emissions through technological options: a modeling study. Com-
putational and Applied Mathematics, 37(1):605–626, 2018.

r
[20] Maitri Verma and Alok Kumar Verma. Effect of plantation of genetically
modified trees on the control of atmospheric carbon dioxide: A modeling

er
study. Natural Resource Modeling, 34(2):e12300, 2021.

[21] Maitri Verma, Alok Kumar Verma, and A. K. Misra. Mathematical mod-
pe
eling and optimal control of carbon dioxide emissions from energy sector.
Environment, Development and Sustainability, 23(9):13919–13944, 2021.

[22] Xiao Zhao, Anjali Mahendra, Nick Godfrey, Holger Dalkmann, Philipp
Rode, and Graham Floater. Unlocking the power of urban transport sys-
ot

tems for better growth and a better climate. The New Climate Economy,
2016.
tn
rin
ep
Pr

25

This preprint research paper has not been peer reviewed. Electronic copy available at: [Link]