Data-Driven Approach to Climate Change Mitigation and Adaptation: The Role of Data Science and Predictive Analytics

Abstract

Climate change remains one of the most pressing challenges to sustainable development around the world, particularly in the global south where environmental, agricultural, and infrastructural systems are highly vulnerable. Data science and predictive analytics have emerged as valuable tools for discovering actionable insights and supporting informed decision-making as a result of increasing demand by governments and institutions for effective strategies to climate mitigation and adaptation. This paper aims to presents a critical review of recent studies on the application of data science and predictive analytics in addressing climate change. The study focuses on peer-reviewed journal articles from Web-of-Science, Springer, IEE Xplore, ScienceDirect, Scopus and Google Scholar published between 2021 to 2025. The paper analyses how machine learning, time series forecasting, and geospatial analytics are employed in key areas such as flood prediction, drought monitoring, early warning systems, and climate-resilient agriculture. The main focus is on how these methods are being applied in the global south, especially across Africa, where there is need for urgent climate solutions. This review identifies the potentials and the limitations of these technologies in supporting climate action. Although risk assessment and planning have been enhanced using data-driven approaches in several contexts, challenges such as data scarcity, limited technical expertise, and weak institutional capacity remain significant barriers to implementation in many developing regions. The paper concludes by recommending strategic actions to strengthen the capacity of Nigerian government and institutions for climate resilience through investments in data infrastructure, institutional and interdisciplinary collaboration, and open access environmental data initiatives.

Keywords: Climate Change, Data Science, Data-Driven, Predictive Analytics, Machine Learning