The Significance of Energy Factors, Green Economic Indicators, Blue Economic Aspects towards Carbon Intensity: A Study of Saudi Vision 2030

Abstract

The current study explores the nexus between energy factors, blue factors, green factors, and carbon intensity in Saudi Arabia. The non-linear ARDL technique is applied to data from 1991 to 2020. The results suggest that the overall impact of energy factors on carbon emission is insignificant, except the adverse shocks in energy intensity, which increase carbon intensity in the long run. Green factors are also irrelevant for carbon emissions in the post-Vision 2030 period. Nevertheless, blue factors are significant for minimizing carbon intensity for post-Vision 2030. Policymakers should invest in efforts to concentrate on energy and blue factors. Investment in the renewable energy and marine sectors is also essential to cater to carbon-related environmental issues.

1. Introduction

In the current era, global economies are heading towards sustainable economic growth that is primarily subject to the importance of low carbon emissions. The Gulf is considered to be the most polluted region, as it is heavily dependent on oil and gas [1]. The reason for the global surroundings’ concern about carbon emissions is briefly described by [2], who states that the culprit behind hazardous gases is the generation of electricity by fossil fuels, which later contributes to factors such as smog, acid rain, and climate change. Climate change itself produces happenings such as forest fires, diseases, and scarcities. According to statistics, in the time between 2008 and 2013, air pollution rose by 8% worldwide, and air quality is still deteriorating [3]. As with other factors, marine pollution is also considerably impacted. According to the International Maritime Organization (IMO), marine transportation emits 940 million tons of carbon annually and is a major contribution to environmental hazards. Ref. [4] questioned greenhouse gas (GHG) emissions through shipping and the considerations undertaken, but a lack of regulations and policies among the shipping industry require steps for implementation.

The relevance of this study is due to emissions that are causing significant effects on the world climate due to fossil fuel burning. As an individual contributor, Saudi Arabia alone emitted 614 million metric tons of carbon dioxide from fossil fuel and manufacturing purposes in 2019 [5]. This is reported to be the biggest emission of carbon production for the country. According to [6], 85% of greenhouse gases are released from energy and electricity generation, and likewise, energy consumption by households and industries in the Middle East and North African (MENA) region due to it being the largest oil exporter globally. Among greenhouse gases, carbon emissions are the chief greenhouse factor from vessels and consume a universal heating potential of 98%. Saudi Arabian government-owned national petroleum and gas hulk Saudi Aramco has been the largest participant in worldwide carbon dioxide production since the 1960s. Saudi Aramco has secreted approximately 60 billion metric tons of carbon dioxide since 1965 [6]. Saudi Arabia, being a Paris agreement member, favors controlling world environmental hazards and making the biosphere human friendly. Moreover, fuel price reforms have escalated oil prices and reduced domestic consumption, supporting environmental friendly and pollutant-preventing factors.

1.1. Contribution

Energy economics literature is notably focused on energy utilization, economic growth, and environmental degradation. Few of studies have documented that energy consumption and generation policies will lower the GHG emissions from the MENA region. Despite the fact that former studies have extensively scrutinized the energy factors, few studies have investigated the importance of green economic factors such as research and development (R&D), innovation, and technological advancement. It has been argued that benefits of innovation are projected for the future, but earlier funding and nurturing of the strategy costs is needed in the initial phases of climate policies [7]. According to [8], Saudi Arabia is willing to transform its economic structure, such as an energy mix from nonrenewable to renewable sources, developing a technologically innovative society, etc., all of which will help with the reduction of carbon emissions. However, the first contribution of this study is to investigate the role of green economic indicators on carbon intensity. In recent years, blue economic indicators have been widely discussed among policymakers and considered an imperative instrument to achieve sustainable economic growth. The impact of blue economic indicators is assumed to be positive to curb environmentally degrading activities [9]. On the contrary, Ref. [10] argued that fisheries are major participants in the emission of greenhouse gases worldwide, as boating emits large amounts of carbon. Similarly, marine trade and marine tourism report contradicting remarks: [11,12] document that marine transport is responsible for 2.5% of global greenhouse gas emissions. Ref. [13] affirmed a negative association between marine trade and carbon emissions. However, the environmental consequences of blue economic activities are still inconclusive. Due to these disparate arguments, an empirical analysis is proposed that depicts the role of blue economic indicators in the control of carbon emissions. So far, the second contribution of the existing study is the inclusion of blue economic indicators. Marine eco-system pollutants have been generated by many activities, such as fishing, marine tourism, and trading. According to [14], fish farming feed includes nitrogen, which is a hazard to the marine environment, and [15] considered aquatic production activity from raw material to final consumption an impurity for the ecosystem. Therefore, [16] concluded that climate action should be addressed, such as renewable energy from the ocean, marine transport, and food production (such as fishing and aquaculture). In terms of marine transport, decarbonizing transport will help to minimize greenhouse gas emissions and water pollution. By using low-carbon fuels, the carbon emissions caused by marine transport and fishing can be minimized. The expansion of greenhouse gas emissions can be controlled by promoting offshore renewable energy, which has been adopted by a number of countries. Another significant measure is the low-carbon food system, which involves introducing low-tropic and organic aquaculture, which leads to low-impact aquaculture.

To provide in-depth analyses, the study focused on a single country rather than panel data. This study selected Saudi Arabia as the country, as it contains seashores on two sides, the eastern seashore and western seashore. Moreover, a lack of research into the context of Saudi Arabia makes us pioneers in learning the role of blue economic factors in carbon intensity. In 2016, Saudi Arabia introduced Vision 2030, which proposed a number of structural reforms that lead to a sustainable environment. To account for the results of Vision 2030, we looked into the turnout of fresh policies. The third contribution is based on Vison 2030: Is Vision 2030 constructive in restraining carbon intensity?

1.2. Objective

There are four objectives of the study based on the contributions: firstly, how energy indicators influence carbon intensity in the context of Saudi Arabia. Energy indicators include energy intensity, renewable energy consumption, and nonrenewable energy consumption. Secondly, the study empirically examines the role of green economic factors in carbon intensity, where green factors include research and development, innovation, and technology. Thirdly, to the study determines whether blue economic factors play a significant role in minimizing carbon intensity, where blue economic factors mainly incorporate fisheries, marine trade, and marine tourism. Lastly, pre-Vision 2030 and post-Vision 2030 analysis is used to evaluate the outcomes of Saudi Arabia’s Vision 2030 to control carbon intensity.

This study is important because it monitors the initiatives taken by government towards sustainability. Three noteworthy initiatives, e.g., promoting clean and affordable energy consumption, climate protection, and reducing marine pollution, are identified, and their results can aid to further implement monitoring and amended policies. Several projects under these categories by the KSA government have been started that support the Sustainable Development Goals and 2030 Agenda. This research mainly provides some variables and the action on these goals, and it notifies how these goals can better support growth and development in a sustainable way.

Furthermore, this article shines a light on the literature, previous studies, and their findings across the world. In a later section, the data and methodology are used to briefly describe the model of the study, as well as the basic methodology and its importance. Results and discussion follow, and the findings in the research are briefly described. The last section concludes with recommendations and an overview of the research work.

2. Literature Review

The literature contributes to environmental degradation by focusing on energy intensity, green economic factors, and blue economic factors. This paper divides the study into four sections: energy and carbon, green indicators and carbon, blue indicators and carbon, and existing gaps in earlier literature.

Besides many other factors, energy consumption is considered a significant source of carbon emissions, adding to carbon intensity. Previous researchers tried to explore the nexus between energy consumption and carbon intensity. In this regard, Ref. [17] used Chinese data to check whether energy consumption is the main reason behind carbon emissions. Through generalized method of moments (GMM) techniques, they found evidence that energy consumption is the main factor behind high carbon emissions and intensity. Similarly, Ref. [18] explored data from 22 countries from 1990 to 2015 and employed ARDL techniques. They also found a causal relationship between energy consumption and carbon emissions. In the same line, Ref. [19] explored data from 15 Asian countries from 1990 to 2013 to check whether this nexus is present in these countries as well. Their analysis was based on the ARDL technique. Results suggest that fossil fuel consumption is the main culprit behind high levels of carbon emissions and environmental degradation at the regional level. Additionally, Ref. [20] argued that it is a fact that energy consumption is essential for economic development. However, this negatively affects the environment through rising carbon emissions. They used Pakistani data from the years 1990 to 2017 and applied the vector error correction model to extract these results.

Ref. [21] explored the determinants of carbon emissions in OECD countries. They used data from 1980 to 2011 and applied the STRIPAT model to find that non-renewable energy consumption adds to increased carbon emissions. However, carbon emissions are reduced through the use of renewable energy. Ref. [22] tried to check the factors that determine the consumption through FMOLS and DOLS techniques. They found that carbon emissions are a major factor behind renewable energy consumption, because renewable energy is an effective way to reduce carbon intensity by reducing overall carbon emissions. Similarly, [23] made an effort to check the role of renewable energy consumption in reducing carbon emissions in Nordic countries. His analysis also proved a unidirectional causality between renewable energy consumption and carbon emissions. Another study by [24] was conducted on data from the United States from 1985 to 2017. Ref. [25] reported a negative relationship between energy and environment. Ref. [26] studied the relationship for the case of China and concluded that there is a significant relationship between energy and carbon emissions. Ref. [27] reported a negative relationship between renewable energy and carbon emissions for the case of Pakistan. Ref. [28] confirmed the negative and significant impact of renewable energy and carbon emissions for belt and road countries. Refs. [29,30,31] reported on the significance of energy and environment. Their analysis shows that high energy intensity is the reason behind environmental degradation due to carbon emissions. In addition, in OECD countries, renewable energy consumption helps mitigate environmental degradation by reducing carbon emissions [32].

Technology and innovation play a significant role in reducing environmental degradation, and previous studies examined this relationship in different contexts. Ref. [33] used the RECIPE technique to check whether low-carbon technologies are essential to reducing environmental degradation. They proved that technical enhancements play a vital role in reducing carbon emissions. Additionally, Ref.[34] used Chinese data to check the role of technical factors in the path of low-carbon economic growth. They proved that technical factors play an essential role in achieving a low-carbon economy. They also suggested that technical development plays a role in reducing carbon emissions at the acceleration stage of development. Hence, economic progress can be achieved without affecting the environment through technological developments. Another study by [35] used data from 15 European Union countries and data from China and the United States from 1990 to 2013. Their econometric estimations showed that investment in research and development is a significant factor in reducing carbon emissions. However, this impact is prominent in developed countries. Hence, it is recommended that policymakers promote research and development to reduce environmental degradation.

Ref. [36] used data from China to explore the nexus between innovations in energy technology and carbon emissions. Their analysis showed that technical innovations in renewable energy help abate carbon emissions. However, technical innovations in fossil energy do not help reduce carbon emissions. Ref. [37] used Korean data to explore factors that affect carbon emissions. This study showed that innovation activities significantly and negatively affect carbon emissions at a 5% significance level. Ref. [38] used data from 96 countries from 1996 to 2018 to understand the nexus between technical innovations and carbon emissions. They found a spatial correlation across nations regarding carbon emissions and research and development through spatial econometric techniques. However, technological innovations do not play any role in mitigating carbon emissions globally. Still, group-based analysis showed that in high-income and high-technology countries, carbon emissions can be significantly reduced in neighboring countries. They also found that globalization plays a vital role in reducing carbon emissions through technical innovations. Refs. [39,40] reported on the importance of innovation to control carbon emissions.

The latest data prove that marine factors also play a role in reducing carbon emissions. Although some amount of carbon is emitted through fisheries, overall, it has a lower carbon footprint due to less livestock care [41]. Hence, fisheries can play a vital role in reducing carbon intensity. Marine ranching is also a significant factor that helps absorb carbon from the atmosphere [42,43]. Besides these factors, fisheries can help reduce carbon intensity by cultivating seaweed [44]. In China, this technique is used to solve the issue of carbon emissions, and significant results have been achieved. Besides, marine fishery trade is also essential for studying carbon intensity because almost 80% of the world’s trade is conducted through the oceans. Hence, the importance of this factor for the economy of any country cannot be ignored. However, this also shows that environmental factors related to marine trade are also important, so proper regulations should be made [45]. Another study was conducted [46] to check whether marine trade emits less carbon compared to other types of trades. Their study proved that carbon emissions are significantly lower due to energy efficiency. Another study by [47] also proved this argument, favoring marine trade as the most energy-efficient and least carbon-emitting trade method. However, Ref. [48] conducted a study in Tunisia that found that bidirectional causality runs in marine trade for carbon emissions. Hence, they suggested that carbon emissions are increasing due to marine trade.

These studies did not directly check the nexus between marine factors and carbon intensity. However, checking this nexus is vital because, as an essential factor in boosting the economy, marine trade should be investigated to see whether it can help reduce carbon intensity. Hence, filling this research gap is essential by using empirical data to check whether marine factors, including fisheries, marine trade, and marine tourism, can help in the reduction of carbon intensity.

Some studies were conducted on data from Saudi Arabia and the whole region overall. In this regard, Ref. [49] conducted a study on MENA countries to explore whether energy consumption contributes to carbon emissions. Their analysis proved that energy consumption is significantly and positively related to carbon emissions in the long run. They suggested that energy conservation can reduce the harmful impacts of energy consumption without interrupting economic growth. Similarly, Ref. [50] used data from oil-rich countries of the MENA region to explore the same nexus from 1975 to 2019. Granger causality analysis found unidirectional causality between energy factors and carbon emissions in the short run. Likewise, Ref. [51] explored Saudi Arabian data from 1970 to 2014 to check whether the oil sector or fossil fuels are the reason behind excessive carbon emissions in the country. They found that a significant amount of carbon in the country is due to the oil sector. It was also suggested that Saudi Arabia focus on the development and consumption of renewable energy, because even though the oil sector is an important economic sector, its carbon footprint is high. Ref. [52] also found a positive impact of energy consumption on carbon emissions in the MENA region.

Based on the previous literature, this paper inspects the effects of CO₂ intensity by generating the following hypothesis:

Hypothesis 1 (H1).

There is an indirect association between CO₂ intensity and renewable energy in Saudi Arabia.

Hypothesis 2 (H2).

There is a direct association between CO₂ intensity and non-renewable energy in Saudi Arabia.

Hypothesis 3 (H3).

Green factors have a negative relationship with carbon intensity in Saudi Arabia.

Hypothesis 4 (H4).

Blue economic factors have a negative relationship with carbon intensity in Saudi Arabia.

3. Data and Methodology

3.1. Data

This article relies on data from World Development Indicators (WDI), the Food and Agricultural Organization (FAO) of the UN, the World Tourism Organization (WTO) and the International Renewable Energy Agency (IRENA). Our data were selected from the years 1991 to 2020 for Saudi Arabia to inquire about three main factors impacting carbon intensity. In our research, the three major factors were subdivided as energy factors, green factors, and blue factors. Energy factors incorporate three variables: energy intensity, calculated as gross inland energy consumption (GIEC) divided by gross domestic product (GDP) at constant 2010 prices; nonrenewable energy, measured as fossil fuel energy consumption (% of total); and renewable energy consumption, measured as total renewable energy (GWh). Secondly, green factors were designated as research and development expenditure (% of GDP), total patents (calculated as patent applications, nonresidents + patent applications, residents), and medium- and high-tech industry (including construction) (% manufacturing value added) to analyze its impact on carbon intensity. Lastly, blue factors were demonstrated via capture fishery production (metric tons), commodity trade, and production and water arrivals (thousands). The models of the study are detailed below:

$$\mathit{Model}1—CI=f \left(EI, NRE, RE\right)$$

$$\mathit{Model}2—CI=f \left(RD, PAT, HT\right)$$

$$\mathit{Model}3—CI=f \left(AQUA, MTRADE, MTOURISM\right)$$

where CI represents the carbon intensity; $EI$, $NRE$, and $RE$ are energy intensity, nonrenewable energy, and renewable energy, respectively; $RD$, $PAT$, and $HT$ reflect the green indicators of research and development, patent, and high technological development, respectively; and $AQUA$, $MTRADE$, and $MTOURISM$ represent aquatic production, marine trade, and marine tourism, respectively. The linear equations studied are as follows:

$$CI={\gamma }_0+{\gamma }_1 EI+{\gamma }_2 NRE+{\gamma }_3 RE+\epsilon$$

$$CI={\delta }_0+{\delta }_1 RD+{\delta }_2 PAT+{\delta }_3 HT+\epsilon$$

$$CI={\sigma }_0+{\sigma }_1 AQUA+{\sigma }_2 MTRADE+{\sigma }_3 MTOURISM+\epsilon$$

Moreover, to investigate the implications of Saudi Vision 2030, we split the data into two subsets: before Vision and after Vision. This assisted in evaluating the impacts of plans that were formulated under the umbrella of Vision 2030, which helps to determine the future plans and directions that lead to a reduction in carbon intensity.

3.2. Methodology

3.2.1. KSS Unit Root Test

In order to check the nonlinear adjustment characteristics, the KSS unit root test was used. The following equation shows the KSS ESTAR specifications:

$$\mathsf{\Delta }{y}_t=\gamma y_{t-1}\left. [1-\exp (-\theta y_{t-1}^2)\right]+{\epsilon }_t<\left(\theta \ge 0\right)$$

(1)

where

$y_t$ = detrended time series of interest;

$\gamma$ = the unknown perimeter;

${\epsilon }_t$ = error term;

$\left. [1-\exp (-\theta y_{t-1}^2)\right]$ = exponential transitional to present the nonlinear adjustments.

In the KSS unit root test, $H_0:\theta =0$ is the null hypothesis and $H_1:\theta >0$ is the alternative hypothesis. In addition, in the case of a null as well as alternative hypothesis, a nonlinear unit root along with the stationary ESTAR process is followed by $y_t$. However, to avoid the indirect null hypothesis testing issue, parametrization of Equation (1) was necessary, and auxiliary regression through first-order Tylor series approximation was obtained through the following Equation:

$$\mathsf{\Delta }{y}_t=\delta y_{t-1}^3+{\epsilon }_t$$

(2)

However, to avoid the issue of serial correlation in the error term, an extension of Equation (2) was used as follows:

$$\mathsf{\Delta }{y}_t={{\displaystyle \sum }}_{j=1}^p{\rho }_j\mathsf{\Delta }{y}_{t-j}+\delta y_{t-1}^3+{\epsilon }_t$$

(3)

It was also necessary to correct the serial correlation error; hence, $\rho$ was used. For lag determination, AR models were used to check the null and alternative hypotheses, i.e., $H_0:\delta =0$ and $H_1:\delta <0$ for Equations (2) and (3) are used, respectively. To present the asymptotic critical value of t NL, Equation (4) was used.

$$t_{NL}=\delta / s.e \left(\widehat{\delta }\right)$$

(4)

where

$\widehat{\delta }$ = OLS estimate of $\delta$;

$s.e \left(\widehat{\delta }\right)$ = standard error of $\widehat{\delta }$.

3.2.2. Autoregressive Distributed Lag (ARDL)

For empirical estimations, this study adopts the autoregressive distributed lag (ARDL) technique that applies in case of mix evidences of stationarity in studied variables.

$$\begin{array}{c}Carbon Intensity\\ \overbrace{CN}\end{array}=\begin{array}{c}Energy Factors\\ \overbrace{\begin{array}{ccc}EI& NRE& RE\end{array}}\end{array},\begin{array}{c}Green Factors\\ \overbrace{\begin{array}{ccc}RD& PAT& HT\end{array}}\end{array} ,\begin{array}{c}Blue Factors\\ \overbrace{\begin{array}{ccc}AQUA& MTRADE& MTOURISM\end{array}}\end{array}$$

$$lnC{I}_t={\alpha }_0+{\beta }_1 lnE{I}_t+{\beta }_2 lnNR{E}_t+{\beta }_3 lnR{E}_t+{\epsilon }_t$$

(5)

$$lnC{I}_t={\alpha }_0+{\beta }_1 lnR{D}_t+{\beta }_2 lnPA{T}_t+{\beta }_{3}lnH{T}_t+{\epsilon }_t$$

(6)

$$lnC{I}_t={\alpha }_0+{\beta }_1 lnAQU{A}_t+{\beta }_2 lnMTRAD{E}_t+{\beta }_3 lnMTOURIS{M}_t+{\epsilon }_t$$

(7)

In the above equations, $CI$ is carbon intensity, $EI$ is energy intensity, $NRE$ is nonrenewable energy, $RE$ is renewable energy, $RD$ is research and development, $PAT$ is patent, $HT$ is high tech, $AQUA$ is aquatic production, $MTRADE$ is marine trade, $MTOURISM$ is marine tourism, ${\epsilon }_t$ is the error term, and ${\beta }_1$, ${\beta }_2$, and ${\beta }_3$ are elasticity coefficients. In terms of the cointegration test, the above equations were used to check the short-run elasticities and long-run cointegration.

The ARDL form of above equations is as follows:

$$\begin{array}{c}\mathsf{\Delta }lnC{I}_t={\alpha }_0+{{\displaystyle \sum }}_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_2\mathsf{\Delta }lnE{I}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_3\mathsf{\Delta }lnNR{E}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_4\mathsf{\Delta }lnR{E}_{t-i}+{\gamma }_{0}lnC{N}_{t-1}+\\ {\gamma }_{1}lnE{I}_{t-1}+{\gamma }_{2}lnNR{E}_{t-1}+{\gamma }_{3}lnR{E}_{t-1}+{\omega }_t\end{array}$$

(8)

$$\begin{array}{c}\mathsf{\Delta }lnC{I}_t={\alpha }_0+{{\displaystyle \sum }}_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_2\mathsf{\Delta }lnR{D}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_3\mathsf{\Delta }lnPA{T}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_4\mathsf{\Delta }lnH{T}_{t-i}+{\gamma }_{0}lnC{I}_{t-1}+\\ {\gamma }_{1}lnR{D}_{t-1}+{\gamma }_{2}lnPA{T}_{t-1}+{\gamma }_{3}lnH{T}_{t-1}+{\omega }_t\end{array}$$

(9)

$$\begin{array}{c}\mathsf{\Delta }lnC{I}_t={\alpha }_0+{{\displaystyle \sum }}_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_2\mathsf{\Delta }lnAQU{A}_{t-i}+{{\displaystyle \sum }}_{i=0}^n{\mu }_3\mathsf{\Delta }lnMTRAD{E}_{t-i}+\\ {{\displaystyle \sum }}_{i=0}^n{\mu }_4\mathsf{\Delta }lnMTOURIS{M}_{t-i}+{\gamma }_{0}lnC{I}_{t-1}+{\gamma }_{1}lnAQU{A}_{t-1}+{\gamma }_{2}lnMTRAD{E}_{t-1}+{\gamma }_{3}lnMTOURIS{M}_{t-1}+{\omega }_t\end{array}$$

(10)

The matrix form of Equations (8)–(10) were formed through Equations (11)–(13). However, there was a chance that cointegration existed in the long run as well as the short run. Hence, to measure long-run cointegration, $\left. [H_0:{\gamma }_{11} to {\gamma }_{43}=0\right]$ and $\left. [H_0:{\gamma }_{11} to {\gamma }_{43}\ne 0\right]$ were formulated as the null hypothesis and alternative hypothesis, respectively. Likewise, $\left. [H0:{\mu }_{11} to {\mu }_{43}=0\right]$ and $\left. [H0:{\mu }_{11} to {\mu }_{43}\ne 0\right]$ were the short-run null and alternative hypotheses, respectively.

$$\left(1-B\right)\left. [\begin{array}{c}\begin{array}{c}lnCI\\ lnEI\\ lnNRE\\ lnRE\end{array}\end{array}\right]=\left. [\begin{array}{c}{\alpha }_{01}\\ {\alpha }_{02}\\ {\alpha }_{03}\\ {\alpha }_{04}\end{array}\right]+{{\displaystyle \sum }}_{i=1}^{k}1-B{\left. [\begin{array}{c}lnCI\\ lnEI\\ lnNRE\\ lnRE\end{array}\right]}_{t-i}\times \left. [\begin{array}{ccc}{\mu }_{11}& {\mu }_{12}& {\mu }_{13}\\ {\mu }_{21}& {\mu }_{22}& {\mu }_{23}\\ {\mu }_{31}& {\mu }_{32}& {\mu }_{33}\\ {\mu }_{41}& {\mu }_{42}& {\mu }_{43}\end{array}\right]+{\left. [\begin{array}{c}lnCN\\ lnEI\\ lnNRE\\ lnRE\end{array}\right]}_{t-1}\times \left. [\begin{array}{ccc}{\gamma }_{11}& {\gamma }_{12}& {\gamma }_{13}\\ {\gamma }_{21}& {\gamma }_{22}& {\gamma }_{23}\\ {\gamma }_{31}& {\gamma }_{32}& {\gamma }_{33}\\ {\gamma }_{41}& {\gamma }_{42}& {\gamma }_{43}\end{array}\right]+{\left. [\begin{array}{c}\omega \\ \omega \\ \omega \\ \omega \end{array}\right]}_t$$

(11)

$$\left(1-B\right)\left. [\begin{array}{c}\begin{array}{c}lnCN\\ lnRD\\ lnPAT\\ lnHI\end{array}\end{array}\right]=\left. [\begin{array}{c}{\alpha }_{01}\\ {\alpha }_{02}\\ {\alpha }_{03}\\ {\alpha }_{04}\end{array}\right]+{{\displaystyle \sum }}_{i=1}^{k}1-B{\left. [\begin{array}{c}lnCN\\ lnRD\\ lnPAT\\ lnHI\end{array}\right]}_{t-i}\times \left. [\begin{array}{ccc}{\mu }_{11}& {\mu }_{12}& {\mu }_{13}\\ {\mu }_{21}& {\mu }_{22}& {\mu }_{23}\\ {\mu }_{31}& {\mu }_{32}& {\mu }_{33}\\ {\mu }_{41}& {\mu }_{42}& {\mu }_{43}\end{array}\right]+{\left. [\begin{array}{c}lnCN\\ lnRD\\ lnPAT\\ lnHI\end{array}\right]}_{t-1}\times \left. [\begin{array}{ccc}{\gamma }_{11}& {\gamma }_{12}& {\gamma }_{13}\\ {\gamma }_{21}& {\gamma }_{22}& {\gamma }_{23}\\ {\gamma }_{31}& {\gamma }_{32}& {\gamma }_{33}\\ {\gamma }_{41}& {\gamma }_{42}& {\gamma }_{43}\end{array}\right]+{\left. [\begin{array}{c}\omega \\ \omega \\ \omega \\ \omega \end{array}\right]}_t$$

(12)

$$\left(1-B\right)\left. [\begin{array}{c}\begin{array}{c}lnCI\\ lnAQUA\\ lnMTRADE\\ lnMTOURISM\end{array}\end{array}\right]=\left. [\begin{array}{c}{\alpha }_{01}\\ {\alpha }_{02}\\ {\alpha }_{03}\\ {\alpha }_{04}\end{array}\right]+{{\displaystyle \sum }}_{i=1}^{k}1-B{\left. [\begin{array}{c}lnCI\\ lnAQUA\\ lnMTRADE\\ lnMTOURISM\end{array}\right]}_{t-i}\times \left. [\begin{array}{ccc}{\mu }_{11}& {\mu }_{12}& {\mu }_{13}\\ {\mu }_{21}& {\mu }_{22}& {\mu }_{23}\\ {\mu }_{31}& {\mu }_{32}& {\mu }_{33}\\ {\mu }_{41}& {\mu }_{42}& {\mu }_{43}\end{array}\right]+{\left. [\begin{array}{c}lnCI\\ lnAQUA\\ lnMTRADE\\ lnMTOURISM\end{array}\right]}_{t-1}\times \left. [\begin{array}{ccc}{\gamma }_{11}& {\gamma }_{12}& {\gamma }_{13}\\ {\gamma }_{21}& {\gamma }_{22}& {\gamma }_{23}\\ {\gamma }_{31}& {\gamma }_{32}& {\gamma }_{33}\\ {\gamma }_{41}& {\gamma }_{42}& {\gamma }_{43}\end{array}\right]+{\left. [\begin{array}{c}\omega \\ \omega \\ \omega \\ \omega \end{array}\right]}_t$$

(13)

where $\mathsf{\Delta }$ is the first difference operator, short-run elasticity operators are presented by ${\mu }_1$ to ${\mu }_4$, long-run elasticity operators are presented by ${\gamma }_1$ to ${\gamma }_4$, ${\alpha }_0$ is the constant, and ${\omega }_t$ is the noise.

The decision regarding the acceptance or rejection of the hypothesis was based on F-statistics along with critical values. Refs. [53,54] presented the critical values for this purpose, and the same was used in this study as well.

3.2.3. Nonlinear ARDL

Although the linearity of the relationships can be checked through cointegration tests, to determine whether this association is positive or negative, [55] advised a nonlinear ARDL approach. The current paper used this approach to determine the direction of the association.

Following [56,57] Equations (11)–(13) were formed. According to them, two sets of series should be formed by decomposing them so that positive and negative changes in independent variables can be catered to.

$$\left\{\begin{array}{c}POS{\left(EF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnE{F}_L^{+}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnE{F}_L,0\right)\\ NEG{\left(EF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnE{F}_k^{-}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnE{F}_L,0\right)\end{array}\right.$$

(14)

$$\left\{\begin{array}{c}POS{\left(GF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnG{F}_L^{+}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnG{F}_L,0\right)\\ NEG{\left(GF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnG{F}_k^{-}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnG{F}_L,0\right)\end{array}\right.$$

(15)

$$\left\{\begin{array}{c}POS{\left(BF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnB{F}_L^{+}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnB{F}_L,0\right)\\ NEG{\left(BF\right)}_t={\displaystyle \sum }_{L=1}^{t}lnB{F}_k^{-}={\displaystyle \sum }_{L=1}^{T}MAX\left(\mathsf{\Delta }lnB{F}_L,0\right)\end{array}\right.$$

(16)

The ARDL form of Equations (8)–(10) was transformed into Equations (17)–(19) after incorporating the positive and negative changes.

$$\begin{gathered} \mathsf{\Delta }lnC{I}_t={\alpha }_0+{\displaystyle \sum }_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{+}\mathsf{\Delta }lnPOS{\left(EF\right)}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{-}\mathsf{\Delta }lnNEG{\left(EF\right)}_{t-i}+{\gamma }_{0}lnC{I}_{t-1} \\ +{\gamma }_1^{+}lnPOS{\left(EF\right)}_{t-1}+{\gamma }_1^{-}lnNEG{\left(EF\right)}_{t-1}+{\omega }_t \end{gathered}$$

(17)

$$\begin{gathered} \mathsf{\Delta }lnC{I}_t={\alpha }_0+{\displaystyle \sum }_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{+}\mathsf{\Delta }lnPOS{\left(GF\right)}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{-}\mathsf{\Delta }lnNEG{\left(GF\right)}_{t-i}+{\gamma }_{0}lnC{I}_{t-1} \\ +{\gamma }_1^{+}lnPOS{\left(GF\right)}_{t-1}+{\gamma }_1^{-}lnNEG{\left(GF\right)}_{t-1}+{\omega }_t \end{gathered}$$

(18)

$$\begin{gathered} \mathsf{\Delta }lnC{I}_t={\alpha }_0+{\displaystyle \sum }_{i=1}^n{\mu }_1\mathsf{\Delta }lnC{I}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{+}\mathsf{\Delta }lnPOS{\left(BF\right)}_{t-i}+{\displaystyle \sum }_{i=0}^n{\mu }_2^{-}\mathsf{\Delta }lnNEG{\left(BF\right)}_{t-i}+{\gamma }_{0}lnC{I}_{t-1} \\ +{\gamma }_1^{+}lnPOS{\left(BF\right)}_{t-1}+{\gamma }_1^{-}lnNEG{\left(BF\right)}_{t-1}+{\omega }_t \end{gathered}$$

(19)

where the coefficients of short-run elasticity are presented by ${\mu }_1$ and ${\mu }_2$, the coefficients of long-run elasticity are presented by ${\gamma }_0$ and ${\gamma }_1$, carbon intensity is presented by $C{I}_t$, energy factors are presented by $EF$, green factors are presented by $GF$, and blue factors are presented by $BF$.

It was also vital to measure the asymmetries in the short and long run; hence, the Wald test was applied for this purpose. Additionally, the Akaike information criterion for optimal lag determination was applied. The bound test, introduced by [58], was used to investigate the cointegration in the long run, where the null hypothesis takes the form ${\gamma }_0={\gamma }_1^{+}={\gamma }_1^{-}=0$.

4. Results and Discussions

Table 1. Descriptive statistic.

Variable	Obs.	Mean	Std. Dev.	Min	Max
CI	30	−2.148	0.065	−2.267	−2.013
EI	30	−1.787	0.200	−2.209	−1.442
NRE	30	14.862	0.432	14.150	15.405
RE	30	5.566	0.279	5.088	6.285
RD	30	−1.704	1.073	−3.163	−0.108
PAT	30	6.925	0.593	6.165	8.203
HT	30	3.409	0.266	3.065	3.743
AQUA	30	10.958	0.176	10.604	11.184
MTRADE	30	7.756	0.792	7.007	9.540
MTOURISM	30	5.005	1.412	0.000	6.928

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM refer to aquatic production, marine trade, and marine tourism, respectively.

depicts the descriptive statistics, which show that the total observations taken were 30 for all variables. The highest standard deviation was found for $RD$ and $MTourism$, with 1.073 and 1.412, respectively. The lowest standard deviation in our data was for carbon intensity ($CI$). The highest mean value was found to be in the non-renewable energy variable. We applied the Chow structural break test to measure the constant value across the sample.

Table 2. Structural break test.

Chow Structural Break
F-Statistics	0.734	Critical Value (5%)	0.683

Note: The null hypothesis represents that the coefficients are constant across the sample.

shows that F-statistics were less than critical value at 5%.

Before the estimation of the models, unit root tests were applied to determine the stationarity of the data, as reported in

Table 3. Unit root test.

	DF-GLS				KSUR
DF-GLS	Level		Diff		Level		Diff
Variable	Stat		Stat		p-Value		p-Value
CI	−2.51	Unit root	−5.69 ***	Stationary	0.519	Unit root	0.000 ***	Stationary
EI	−1.485	Unit root	−7.085 ***	Stationary	0.072 *	Stationary	0.028 **	Stationary
NRE	−0.873	Unit root	−6.201 ***	Stationary	0.927	Unit root	0.001 ***	Stationary
RE	−2.926	Unit root	−5.277 ***	Stationary	0.223	Unit root	0.000 ***	Stationary
RD	−1.583	Unit root	−4.284 ***	Stationary	0.824	Unit root	0.000 ***	Stationary
PAT	−2.121	Unit root	−3.276 *	Stationary	0.421	Unit root	0.024 **	Stationary
HT	−2.241	Unit root	−6.415 ***	Stationary	0.301	Unit root	0.000 ***	Stationary
AQUA	−2.79	Unit root	−6.405 ***	Stationary	0.564	Unit root	0.001 ***	Stationary
MTRADE	−1.641	Unit root	−6.185 ***	Stationary	0.942	Unit root	0.008 ***	Stationary
MTOURISM	−2.447	Unit root	−4.47 ***	Stationary	0.07 *	Stationary	0.098 *	Stationary

Note: GF-GLS does not assume stationary nonlinearity, whereas the KSUR unit root test presented by Kapetanios and Shin (2008) considers stationary nonlinearity. CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. ***, **, and * represent the level of significance at 1%, 5%, and 10%, respectively.

. DF-GLS and KSUR tests were performed, where DF-GLS was found to be stationary at the first difference for all the variables and KSUR was considered nonlinear stationary at the first difference. Performing the NARDL technique required a bounds cointegration test, and all models showed the existence of cointegration. CI = f (EI, NRE, RE) is cointegrated at 1%, CI = f (RD, PAT, HT) at 10%, and CI = f (AQUA, MTRADE, MTOURISM) at 5%, which is presented in

Table 4. ARDL bound cointegration test.

ARDL Bounds Cointegration Test	F-Stat	CI
CI = f(EI, NRE, RE)	9.534 ***	Exist
CI = f(RD, PAT, HT)	3.914 *	Exist
CI = f(AQUA, MTRADE, MTOURISM)	4.619 **	Exist
Lower-bound critical value at 1%		4.29
Upper-bound critical value at 1%		5.61
Lower-bound critical value at 5%		3.23
Upper-bound critical value at 5%		4.35
Lower-bound critical value at 10%		2.72
Upper-bound critical value at 10%		3.77

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. ***, **, and * represent the levels of significance at 1%, 5%, and 10%, respectively.

. Therefore, nonlinear ARDL models could be applied in our models.

Table 5. Nonlinear ARDL estimation.

Long Run	Model 1	Model 2	Model 3
CIt−1	−0.405	−1.682 ***	−1.036 *
EI+t−1	−0.068
EI−t−1	0.145 *
NRE+t−1	0.033
NRE−t−1	14.582
RE+t−1	0.046
RE−t−1	−0.422
RD+t−1		−0.001
RD−t−1		−0.427 **
PAT+t−1		0.004
PAT−t−1		−0.441 *
HT+t−1		−0.025
HT−t−1		0.538
AQUA+t−1			0.169
AQUA−t−1			0.576 **
MTRADE+t−1			0.034 **
MTRADE−t−1			0.001
MTOURISM+t−1			−0.037
MTOURISM−t−1			−0.039
Short Run	Model 1	Model 2	Model 3
ΔCIt−1	−0.278	0.368	−0.090
ΔEI+t−1	−0.117
ΔEI−t−1	−0.177 **
ΔNRE+t−1	−0.212
ΔNRE−t−1	−0.555
ΔRE+t−1	−0.143
ΔRE−t−1	0.253
ΔRD+t−1		0.043
ΔRD−t−1		0.333 *
ΔPAT+t−1		−0.038
ΔPAT−t−1		0.369
ΔHT+t−1		−0.174
ΔHT−t−1		0.440 **
ΔAQUA+t−1			−0.212
ΔAQUA−t−1			−0.106
ΔMTRADE+t−1			−0.008
ΔMTRADE−t−1			−0.027
ΔMTOURISM+t−1			0.017
ΔMTOURISM−t−1			0.037
Constant	−0.948 *	−3.408 ***	−2.148 *

Notes: CI stands for carbon intensity. EI, NRE, and RE represents the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, **, and * represent the level of significance at 1%, 5%, and 10%, respectively.

demonstrates empirically that a negative change in energy intensity had a direct and significant relationship with carbon intensity. In the long run, a unit negative change in energy intensity had a negative effect of 0.145 of carbon released by the country. It seemed to be important, but no other significant results proved the impact on carbon intensity in the case of Saudi Arabia in this era of research data. Refs. [59,60,61] depicted in conclusion a direct relationship between energy depletion and CO₂ emissions. We justified this, as [62] stated that oil and its depletion have a positive impact on CO₂ emissions for oil-exporting nations because its ecological impacts imply a higher degree of CO₂ emissions in the country and last for an additional lengthy period. In view of the short run, similar results show that negative modifications had some effects on carbon intensity, as they increased the carbon hazard release effect. Hazards had a long-term effect. A reduction in energy usage also had major effects, such as climate change and other emissions that are hazardous for the environment. Change in renewable and non-renewable energy had no significant impact on carbon releases or intensity in this case. Refs. [63,64] stated that it does not contribute to reducing CO₂ emissions. Ref. [65] concluded that there was no significant impact on carbon emissions and demonstrated that the usage of renewable energy has not yet reached a particular amount that makes any change in reducing the carbon.

In the case of long-term effects of green economic factors, we saw research and development with a negative shock effect resulting in an indirect impact on carbon emissions. This means that the decrease in research and development is accelerating carbon emissions, which was proven to be true by many earlier researchers, such as [35,37]. Ref. [36] reflected on research and development as an imperative feature in the backing of sustainable development with lower carbon emissions. Similarly, patents have an indirect impact on carbon intensity as a negative shock effect. Refs. [66,67,68] stated the vital role of patents and their reduction of carbon intensity. Ref. [69] performed research on KSA and confirmed that patents reduce carbon intensity. Ref. [70] are of the view that it technically enhances productivity of the product with less energy usage. Short-term empirical data show that R&D with a negative change is decreasing carbon emissions, and similar results have been demonstrated by the negative shock of high technology. In the short run, a decrease in carbon intensity may lead to less support in terms of sustainability, as the effect of carbon emissions is a hazard for the longer term. Nations should focus more on long-term impacts, as these lead to real hazards. Our marine factors were only seen to have a significant impact in the long run: Aquatic life received a negative shock, which had a direct and significant impact on carbon intensity and marine trading, whereas a positive change increased carbon intensity. These relationships were proved by [41,48]. We can justify that urbanization could be leading the negative changes in fisheries, as [71] are of the view that effect of urbanization on carbon emissions is cut in fisheries and other factors. Ref. [72] studied seafood’s carbon footprint and found it to present a hurdle in attaining sustainability. Similarly, as Saudi Arabia is an exporter of fuel and also imports its major consumer goods, marine trade releases carbon and creates pollutants.

According to

Table 6. Asymmetric and model diagnostics.

	Long Run (+)	Long Run (−)	Long Run Asymmetry (p-Value)	Short Run Asymmetric (p-Value)
CI = f(EI, NRE, RE)
EI	−0.167	−0.359	0.376	0.159
NRE	0.081	−36.024	0.452	0.501
RE	0.113	1.043	0.453	0.297
Cointegration				−1.800
Portmanteau test				0.225
Heteroskedasticity				0.159
Ramsey test				0.232
J−B test				0.450
CI = f(RD, PAT, HT)
RD	0.074	−0.254 ***	0.012 **	0.089 *
PAT	0.003	−0.262 **	0.015 **	0.170
HT	−0.015	−0.320	0.285	0.033 **
Cointegration				−4.736
Portmanteau test				0.473
Heteroskedasticity				0.537
Ramsey test				0.122
J−B test				0.356
CI = f(AQUA, MTRADE, MTOURISM)
AQUA	0.163	0.556 *	0.012 **	0.854
MTRADE	0.032 **	−0.001	0.782	0.354
MTOURISM	−0.036	0.038	0.936	0.631
Cointegration				−2.079
Portmanteau test				0.147
Heteroskedasticity				0.304
Ramsey test				0.602
J−B test				0.864

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, **, and * represent the level of significance at 1%, 5%, and 10%, respectively.

, our results show that none of the energy factors in the long run changed the impacts of carbon intensity. Neither long-run asymmetry nor short-run asymmetry were found to have an empirical impact. Oil exports contribute majorly to GDP, which itself does not support the carbon-to-GDP ratio within the country. Research and development with a 1% long-term negative change has the ability to increase carbon intensity by 0.25%, and a 1% negative change in patents has the ability to increase carbon emissions to GDP by 0.26%. As mentioned above, researchers have shown the relationship between a country’s research and its effect on carbon emissions. R&D and patents were found to have long-run asymmetry with carbon intensity, whereas research and development and high technology had short-run asymmetry with the carbon intensity of Saudi Arabia. Saudi Arabia is working on both sectors in order to minimize carbon release in the long term. In long-term empirical statistics, a 1% change in negative shock to aquatic life had a positive effect on carbon intensity by 0.55% due to overfishing and less innovative fishing vessels, which have a higher carbon footprint. Our findings demonstrate that a long-term positive change of 1% in marine trading increases carbon intensity by 0.03 percent. Hence, aquatic life is considered more significant in its impact on carbon intensity and only long-term asymmetry was seen with carbon intensity. Tourism is an important factor for sustainable growth and creates fewer carbon-releasing activities.

In the case of the pre-Vision 2030 era,

Table 7. Nonlinear ARDL estimation (pre-Vision 2030).

Long Run	Model 1	Model 2	Model 3
CIt−1	−0.024 ***	−0.008 ***	−0.017 ***
EI+t−1	0.000
EI−t−1	0.004 ***
NRE+t−1	0.003 ***
NRE−t−1	0.022 ***
RE+t−1	0.027 *
RE−t−1	0.003
RD+t−1		−0.059 **
RD−t−1		0.041
PAT+t−1		0.013
PAT−t−1		0.090
HT+t−1		−0.001 **
HT−t−1		0.003 **
AQUA+t−1			−0.002 ***
AQUA−t−1			0.004 ***
MTRADE+t−1			0.008
MTRADE−t−1			0.001 ***
MTOURISM+t−1			0.015
MTOURISM−t−1			0.083 **
Short Run	Model 1	Model 2	Model 3
ΔCI t−1	0.946 ***	0.977 ***	0.963 ***
ΔEI+t−1	0.051 *
ΔEI−t−1	0.102 ***
ΔNRE+t−1	0.081 **
ΔNRE−t−1	0.118
ΔRE+t−1	0.005
ΔRE−t−1	0.002
ΔRD+t−1		0.011
ΔRD−t−1		−0.016
ΔPAT+t−1		−0.019 ***
ΔPAT−t−1		−0.054 ***
ΔHT+t−1		0.028 ***
ΔHT−t−1		0.087 ***
ΔAQUA+t−1			0.045 **
ΔAQUA−t−1			−0.032
ΔMTRADE+t−1			0.001 *
ΔMTRADE−t−1			−0.026 ***
ΔMTOURISM+t−1			−0.001
ΔMTOURISM−t−1			0.002
Constant	−0.049 ***	−0.015 ***	−0.035 ***

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, **, and * represent the level of significance at 1%, 5%, and 10%, respectively.

demonstrates the estimations that negative change in energy intensity directly affected the environment by lessening the CO₂ emissions for Saudi Arabia. Our results are of the view that, for sustainable growth, a significant negative change in energy intensity is favorable for Vision 2030. Similarly, the long-term direct impact of non-renewable energy usage was identified in our results, which supports that fossil fuels are contributing to carbon intensity. We can justify this through the cost of domestic energy: The cost of domestic energy was lower [73]. This raised the usage of energy to a vast level and, hence, carbon intensity and hazardous releases boomed. A positive change in renewable energy variables gave a sense of interest from the government in carbon hazards for the nation and worldwide. It was also evidenced that renewable energy consumption has a direct relationship with carbon intensity. Refs. [21,22] empirically showed a direct relationship between non-renewable energy consumption and carbon emissions. Due to less reliance on renewable energy in the pre-Vision 2030 era, long-term carbon intensity has not seen a reduction and a few years are needed to overcome the intense hazard created by the country’s economic policies. Renewable energy was found to be significant at 10% in the long-term effect and insignificant in the short-term effect. Hence, this factor needs to be accounted for by the government. Overall, the energy intensity showed that the least coefficients and a unit change are not contributing majorly to carbon intensity in the case of Saudi Arabia. In regard to the long term, a positive change in R&D is decreasing the carbon intensity for the country. Similarly, another aspect is important and significant among green economic factors: High technology had a direct relationship with carbon intensity. Carbon intensity is decreasing with the help of high technological steps, and a negative change in technology demonstrates an increasing trend in carbon releases. Ref. [36] supported R&D and [74,75] favored technological development for environmentally friendly actions. Innovation was found to be statically insignificant towards carbon intensity, which resembles the results of [76]. According to marine factors, positive and negative changes in aquaculture significantly reduced the carbon intensity. Ref. [51] are of the view that urbanization reduces fisheries and hence increases CO₂ emissions; Saudi Arabia’s urban development has gained attention and led towards hazardous emissions. Another reason identified, according to the FAO of the United States, is that aquaculture amplified progressively from 6000 tons in 2000 to nearby 26,000 tons in 2009 and 2010. However, there was a decrease in production to 16,000 tons in 2011 due to the eruption of a virus-related white spot infection in marine shrimp, the most farmed species in the nation. The fishery sector needs to focus on the sustainable developmental goals of the nation. A negative change in marine trade and tourism provides an image of lessening carbon intensity for the case of Saudi Arabia. Saudi Arabia is not only a major exporter of oil for their GDP but also a heavy importer of consumer goods for domestic customers. Trading and tourism have a direct relationship with energy usage [77]. In the MENA region a direct impact of energy consumption and trade was found on the CO₂ emissions. Ref. [78] considered tourism to be the cause of the carbon footprint in Barcelona, with the reason for emissions being energy consumption by transportation.

Our short-term results demonstrate that energy intensity had a direct relationship with carbon intensity in the pre-Vision 2030 era. A positive change in energy consumption gave a boost to carbon intensity and hence had more effect in the short run. Hazard emissions should be more of a concern for short-term in comparison to long-term effects. A positive change in non-renewable energy significantly effected carbon intensity, as supported by [21]. In the short-run, patents showed an indirect relationship with carbon intensity, which means a positive change in innovative activities supports a significant reduction in carbon emissions. Ref. [79] conducted research on Middle Eastern countries and showed that environmental innovations and management support the reduction of carbon emissions. Similarly, high technology had a direct relationship with carbon intensity. Ref. [80] concluded, based on endogenous growth theories, that technological advancement is significant to growth and developmental activities. In marine indicators, trading was found to be significant, with a positive change having a direct relationship and a negative change having an indirect relationship with carbon intensity. Saudi Arabia’s trading sector cannot be depleted due to its severe need for the economy. Lessening it still raised the carbon intensity for the country. A positive change in marine trade proved to be insignificant in the long-term effect and significant at 10% in the short-term effect. Thus, marine trade’s significance proved that the carbon intensity effect is not reliable.

The pre-Vision asymmetric results show that energy factors significantly contributed to carbon intensity. We found that a 1% long-run negative change in energy intensity increased carbon intensity by 0.18%. However, a focus point of this era is that a positive long-term change in non-renewable energy change directly affected carbon intensity by 0.141%. However, when non-renewable energy consumption was reduced by 1%, it could increase carbon intensity to 0.88%. The Saudi economy is heavily reliant on fossil fuels, which might affect its GDP growth, leading to a hurdle in their sustainability goals. Energy intensity and non-renewable energy both had long-term asymmetry with carbon intensity, whereas renewable energy was found to have short-term asymmetry. According to

Table 8. Asymmetric and model diagnostics (pre-Vison 2030).

Pre Vision 2030	Long Run (+)	Long Run (−)	Long-Run Asymmetry (p-value)	Short-Run Asymmetric (p-Value)
CI = f(EI, NRE, RE)
EI	−0.004	0.180 ***	0.000 ***	0.223
NRE	0.141 ***	−0.892 ***	0.000 ***	0.134
RE	0.020 *	−0.014 *	0.650	0.000 ***
Cointegration				−12.842
Portmanteau test				0.330
Heteroskedasticity				0.801
Ramsey test				0.165
J−B test				0.227
CI = f(RD, PAT, HT)
RD	−0.031 *	−0.130	0.123	0.075 *
PAT	0.002	0.006	0.863	0.904
HT	−0.107 **	0.381 *	0.158	0.165
Cointegration				−7.051
Portmanteau test				0.824
Heteroskedasticity				0.737
Ramsey test				0.414
J−B test				0.630
CI = f(AQUA, MTRADE, MTOURISM)
AQUA	0.111 ***	−0.209 ***	0.00 ***	0.001 ***
MTRADE	0.018 *	−0.034 ***	0.007 *	0.968
MTOURISM	0.005	0.004 **	0.007 ***	0.658
Cointegration				−11.977
Portmanteau test				0.535
Heteroskedasticity				0.142
Ramsey test				0.293
J−B test				0.988

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, **, and * represent the level of significance at 1%, 5%, and 10%, respectively.

, long-term positive and negative changes in renewable energy had the very least effect on carbon intensity augmentation. We are of the view that the Saudi economy has not been relying on generating energy through renewable sources.

Our models depict post-Vision 2030 results in

Table 9. Nonlinear ARDL estimation (post-Vision 2030).

Long Run	Model 1	Model 2	Model 3
CI t−1	−0.015 ***	−0.060 ***	−0.030 ***
EI+t−1	−0.008
EI−t−1	0.0288 ***
NRE+t−1	4.779 ***
NRE−t−1	0.018
RE+t−1	0.004
RE−t−1	0.287 ***
RD+t−1		−0.011
RD−t−1		0.000
PAT+t−1		0.015
PAT−t−1		0.003
HT+t−1		−0.048
HT−t−1		−0.001
AQUA+t−1			0.061 ***
AQUA−t−1			−0.031 ***
MTRADE+t−1			0.006 ***
MTRADE−t−1			0.001
MTOURISM+t−1			0.010 ***
MTOURISM−t−1			0.005 ***
Short Run	Model 1	Model 2	Model 3
ΔCI t−1	0.910 ***	0.923 ***	1.174 ***
ΔEI+t−1	−0.051
ΔEI−t−1	−0.368 ***
ΔNRE+t−1	−8.973
ΔNRE−t−1	0.000
ΔRE+t−1	0.000
ΔRE−t−1	−76.048
ΔRD+t−1		8.652
ΔRD−t−1		0.006 *
ΔPAT+t−1		−0.145
ΔPAT−t−1		−0.013
ΔHT+t−1		3.924
ΔHT−t−1		−0.303 ***
ΔAQUA+t−1			−1.287 ***
ΔAQUA−t−1			−1.307 ***
ΔMTRADE+t−1			−0.025 ***
ΔMTRADE−t−1			−0.441 ***
ΔMTOURISM+t−1			0.046 ***
ΔMTOURISM−t−1			−0.058 ***
Constant	577.716 ***	−0.018	−0.037 ***

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflects the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, and * represent the level of significance at 1%, and 10%, respectively.

. A negative change in energy intensity and renewable energy consumption significantly and directly affected carbon intensity reduction in the long-term. Similarly, a positive change in non-renewable energy still contributed to carbon intensity. Ref. [60] proved that energy directly impacts carbon intensity. Urbanization in the country is a major factor of energy usage, but technological innovation is somewhat helpful in reducing non-renewable energy hazard emissions. However, being an oil-based country, this requires time. In the short term, a negative change in energy intensity still contributed to carbon intensity, as carbon hazards are long term, so even energy reduction played a significant role. Our model 2 was identified to have insignificant results, meaning that we can justify the assumptions made by [81]: Investment in R&D does not straightaway convert into innovation and progressive technologies, as time is taken and people hesitate before adapting because old habits and savings are not easy to leave. Translation takes time and contributes to hysteresis. On the other hand, the adoption of new technologies also takes some time. Moreover, long-term results require several years to demonstrate outcomes. In the short term, negative changes in high technology increase carbon intensity because, as discussed, technological adaption needs ample time, whereas negative changes in research and development significantly decrease the carbon intensity because economic growth decreases. Our blue factors proved significant at 1%, showing their importance in carbon intensity for the Saudi economy. Long-term positive and negative changes in aquaculture create a positive long-term effect on carbon emissions. There is a need to focus on technologies, transportation, and other usages for activities raising carbon intensity. According to the FAO of the United Nations, after disease outbreak, the whole production further improved to 30,000 tons in 2015. Shrimp farming is the greatest success in the country, with a manufacture of 21,000 tons. However, energy usage has raised CO₂ emissions due to feed production and proper environment handling. Saudi Arabia modestly exports shrimp, as it is not renowned for this among other nations. This increase in trade saw a direct relationship with carbon intensity for the country at the 1% significant level. Tourism had positive significant results, which means that negative and positive changes accordingly affect carbon intensity, as tourism creates transport activities in the country and transport is the main source of energy usage. This justifies the results of [82]. As for the long term, blue factors showed the importance of the short-term effect on carbon intensity, as positive changes in aquaculture decreased carbon intensity and negative changes increased carbon intensity. This indirect relationship demonstrates that it supports economic activities but at a cost of environmental degradation. An indirect relationship with trade was seen at the 1% significance level for carbon intensity, but marine tourism in both positive and negative shocks showed a carbon intensity augmentation. [83] are of the view that trade openness somewhat reduces the carbon emissions if renewable energy is being used, so it was identified in our short-term results because a proper shift to a renewable energy source needs to be given years. Otherwise, trade openness may cause severe environmental problems.

Table 10. Asymmetric and model diagnostics (post-Vison 2030).

Post Vision 2030	Long Run (+)	Long Run (−)	Long-Run Asymmetry (p−Value)	Short-Run Asymmetric (p−Value)
CI = f(EI, NRE, RE)
EI	−0.549	1.888 **	0.170	0.005 ***
NRE	0.909 ***	0.000	0.002 ***	0.697
RE	0.000	−0.880	0.302	0.589
Cointegration				−4.546
Portmanteau test				0.000
Heteroskedasticity				0.842
Ramsey test				0.000
J−B test				0.618
CI = f(RD, PAT, HT)
RD	−0.188	0.007	0.365	0.000 ***
PAT	0.259	−0.042	0.722	0.696
HT	−0.796	0.014	0.925	0.513
Cointegration				−5.069
Portmanteau test				0.643
Heteroskedasticity				0.978
Ramsey test				0.189
J−B test				0.167
CI = f(AQUA, MTRADE, MTOURISM)
AQUA	2.019 ***	1.043 ***	0.000 ***	0.000 ***
MTRADE	0.209 ***	−0.028	0.170	0.116
MTOURISM	0.346 ***	−0.172 ***	0.000 ***	0.502
Cointegration				−12.658
Portmanteau test				0.974
Heteroskedasticity				0.811
Ramsey test				0.206
J−B test				0.350

Notes: CI stands for carbon intensity. EI, NRE, and RE represent the energy intensity, non-renewable energy, and renewable energy consumption, respectively, which reflect the energy indicators. RD, PAT, and HT are research and development, patent, and high technology, respectively, which highlight the green economic indicators. AQUA, MTRADE, and MTOURISM represent aquatic production, marine trade, and marine tourism, respectively. (+) and (−) indicate the positive and negative shocks in the respective variables. ***, and ** represent the level of significance at 1% and 5%, respectively.

demonstrates positive and negative changes in variables in the long run. Thus, a 1% long-run negative change in energy intensity decreased carbon intensity by 1.8% which was as expected; energy is a big source of carbon emissions and short-term asymmetry was found among energy intensity and carbon emissions. A long-run positive change in nonrenewable energy means that depending on fossil fuels creates an environment of hazardous emissions, with results showing a 0.9% effect on carbon intensity with a 1% change. Long-term asymmetry was seen among these variables. In this era, short-term asymmetry was seen with R&D and carbon intensity. A major long-term positive and negative change in aquaculture increased carbon intensity by 2.8% and reduced it by 1.04%, respectively. This variable is important to notice because the marine industry is utilizing energy and creating economic growth for the economy. However, this needs to be eco-friendly and sustainable. Marine trade’s and marine tourism’s long-term changes increase carbon intensity. News (2021) stated that “Saudi Arabia is surrounded by vast coasts, the Arabian Gulf and the Red Sea, through which 13% of the global trade in this giant industry passes.” The transportation sector therefore needs to be focused on the usage of ecofriendly fuel. Short-term asymmetry was identified among aquaculture and carbon, whereas only long-term asymmetry between tourism and carbon intensity was found in our results. We justified this as per the results of [77], who presented a positive relationship between carbon intensity and trade and energy consumption.

5. Conclusions

This research states the effect on carbon intensity via positive or negative shocks in energy and green and blue economic indicators. Primarily, the study expected that a direct relationship would be found with renewable energy and indirect with nonrenewable energy consumption in the case of Saudi Arabia, as per the previous literature by [19,21,23]. However, the results found that energy factors are not significantly subsidizing carbon intensity. It was seen that pre-Vision 2030, non-renewable energy impacted carbon intensity but with a minor coefficient showing an almost minor impact on carbon intensity. However, important asymmetric results show that positive shocks in non-renewable energy, for the long-term, increased carbon intensity. Refs. [50,77] found in the MENA region a direct relationship with carbon. Ref. [51] found Saudi Arabia to be heavily oil depend for its GDP. To counter this issue, the country needs to minimize its dependence on oil, which is not suitable because of its heavy reliance on oil. However, the reduction in oil would negatively affect economic growth. We conclude that Saudi Arabia needs a heavy investment of time, money, and policies in order to shift their GDP towards non-oil and to be independent of hazardous emissions. Renewable energy was not seen to be as significant as expected in post-Vision 2030, meaning its contribution is needed in terms of policies and procedures for sustainable growth and carbon emissions reduction. Marine factors had a significant impact, as per the results in post-Vision 2030. Tourism had a direct relationship with carbon emissions, indicating that tourism is a source of carbon emissions, which is a shift from oil based-income, similar to UAE [84]. Saudi Arabia is contributing to this sector and creating a pathway for a better environment and climate.

The research concluded by rejecting the hypothesis, as no significant importance was seen as a result of green factors in the post-Vision 2030 era. Long-term changes in high technology participated in the reduction of carbon intensity in pre-Vision 2030, whereas patents and R&D had less significance in our case. The paper suggests that Saudi Arabia’s policymakers concentrate more greatly on these variables, as this can help a lot in creating sources of income other than oil, leading towards a sustainable environment and sustainable economy. Investments can be generated for national projects, which can create employment opportunities, educational opportunities, high living standards, and income for the betterment of the public and suitably attain the sustainable goals of Vision 2030.

This study suggest that Saudi Arabia should boost the private region to capitalize on production and practice renewable energy. Furthermore, struggles should be addressed to amend the complete energy mix. To improve renewable energy usage, the government should offer energy subsidies and encourage small renewable energy producers to cope with its cost and provide easy access. At the governmental level, the usage of renewable energy appliances and vehicles should be encouraged in order to reduce carbon intensity. Efficient fuel shipping methods should be introduced and further reduction of over exploitation of fishing should be considered. The government of Saudi Arabia needs strict actions on no spill-over fuel in trading to gain sustainable blue economy goals. For this purpose, local fisheries and trading partners should be encouraged to avoid mishandling in trade.

Further research can be done to compare other MENA countries and the identification of more varied policies can be carried out. Different countries have dissimilar policies, and this can help to pinpoint the factors each country feels are important according to their economy. Another gap has been fulfilled in the way that, as the years pass, changes occur in policies, and rapid attention to Vision 2030 might change our results or other important factors for the country. In later years, additional coming years can be included in the work. Lastly, more environmental hazards other than carbon footprints can be included and their impacts can be studied separately by each factor.