As part of the 2015 Paris Climate Agreement, the Kingdom of Saudi Arabia (KSA) pledged to implement policies aimed at reducing greenhouse gas emissions. The 2030 Vision, which KSA unveiled with the goal of building a more sustainable and diverse economy, led to a number of projects, such as the national renewable energy program, the Saudi green initiative, and the circular carbon economy. Additionally, KSA just revealed a bold plan to achieve net-zero status by 2060. The Kingdom pledged to cut its carbon emissions by 278 million tons of CO2eq (equivalent) yearly by 2030 in its revised nationally determined contribution (NDC). When compared to the previously stated goal of 130 million tons of CO2eq, this aim more than doubles. In this paper, a predictive model is intended to precisely anticipate future CO2 emission levels by combining gradient boosting regression using XGBoost, industrial activity indicators, and temporal aspects. The model is able to capture intricate linkages and patterns influencing emissions by utilizing these various data components. Initial findings show this approach's potential, and feature engineering and optimization techniques are used to further improve it. The results demonstrate how well domain-specific predictors and cutting-edge machine learning techniques can be used to solve environmental issues and provide a trustworthy and understandable framework for CO2 emission forecasting in practical applications.
Albalawi et al. (Fri,) studied this question.