How COVID-19 Impacted CO2 Emissions Based on Electricity Usage: A Machine Learning Approach

Authors

  • Qiancheng Sun Department of Mechanical and Aerospace Engineering University of Dayton, Dayton, OH
  • Andrea Zela-Koort Department of Mechanical and Aerospace Engineering University of Dayton, Dayton, OH
  • Ava Stokes Department of Mechanical and Aerospace Engineering University of Dayton, Dayton, OH
  • Salahaldin Alshatshati Department of Mechanical and Aerospace Engineering University of Dayton, Dayton, OH

Abstract

The goal of this study is to determine the difference in CO2 emissions between 2019-2020 and 2020-2021, more specifically during lockdown periods during the COVID-19 pandemic. In the beginning of the pandemic, most countries were forced into lockdowns, and a countless number of people had to continue their daily work from home in isolation. Previously, people would go to an office or to school and leave their houses empty for eight hours, without having lights or any electronics on. Because of this, there should be a direct correlation between electricity usage before and during lockdowns, as a private residence should have higher electricity consumption during 2020-2021, when they are at home. Using machine learning, we will investigate to see if COVID-19 affected CO2 emissions as a result of more electricity usage in private residences. A model will be made to predict what the CO2 emissions would be for 2019-2020, based on electricity usage data from 2020-2021. Then, the real CO2 emissions from 2019-2020 will be compared with the model’s predicted values, and the difference will indicate if COVID-19 caused an inconsistency between actual and predicted CO2 emissions. Factors that were taken into account when making a model were independent variables relating to outdoor conditions, the number of people living in the house, and the temperature that the thermostat is set at, making the response variable CO2 emissions

Published

2021-07-01