Forecasting of climate change by using big data and machine learning with python program

Climate change encompasses substantial and enduring modifications in the Earth's overarching climatic patterns. Moreover, the term "big data" pertains to a diverse array of data types, characterized by continuous expansion in volume and rapid influx, often distinguished by the quintessential five Vs. This study's principal aim revolves around the examination of carbon emissions, greenhouse gas emissions, and environmental trade data concerning the countris of Türkiye the United States, Germany, Iraq, and China. Additionally, the research endeavors to establish associations with various factors, such as air temperature, surface temperature, and weather-related calamities within these respective countries. The data utilized in this investigation were procured from the official website of the International Monetary Fund. To accomplish this research objective, the study employs the Shapley Additive Explanation methodology and a nonlinear external input autoregressive network, executed through a Python-based program. These techniques are harnessed to prognosticate climate change utilizing the amalgamation of big data and machine learning methodologies. The outcomes of this study reveal that a unitary increment in the United States' trade balance related to low-carbon technology products results in a 0.042-unit elevation in the mean global surface temperature variable. Furthermore, a one-unit augmentation in the United States' exports of environmental goods as a proportion of total exports induces a 0.028-unit surge in the mean global surface temperature variable. In the context of Germany, a one-unit increase in the total trade in environmental goods variable corresponds to a noteworthy 8.745-unit reduction in the Land Cover Woody Crops variable. Additionally, a one-unit increase in the Germany (Expenditure on Waste Water Management) variable leads to a substantial 193,482-unit escalation in the Land Cover (Shrub-covered Areas) variable. All graphical representations within this investigation consistently evince that the mean global surface temperature variable exhibits the closest alignment with the NARX model in comparison to all other variables, signifying the model's reliability in generating results.