Comparative Analysis of Traditional and Artificial Intelligence-Based Emission Inventory Models

Abstract

Emission inventories are crucial to quantify the emission sources of various anthropogenic and natural sources to the air. They aid with environmental planning, air quality assessment and air quality policy-making. The traditional approach to the emission inventory models concentrates on the emission factors, the activity data and statistical estimation methods. While these strategies are established, challenges with data uncertainty, evolving over time, and computational issues can limit application. Recently Artificial Intelligence (AI) technologies and Machine Learning (ML)/Deep Learning (DL) have been introduced as a new technique to perform the various estimation tasks by the use of data-driven models in order to achieve efficient and accurate estimation of the emissions. Emissions inventory models, which are supported by artificial intelligence, are able to process huge amounts of data, identify the non-linear relationship between the data and create predictions, all at the same time. The paper compares the traditional model to the AI based emission inventory model with respect to their methodology, requirement of data, computation time, prediction, scalability and applications. The study accentuates the significance of gaining these features and its necessity to be supplemented by the conventional methods for developing more precise, reliable and intelligent emission inventory system to ensure the sustainability of the environment.