Title: Measuring carbon emission of deep learning: a systematic literature review

Authors: Abdul Kasiru Zakaria; Patrick Kwabena Mensah; Faiza Umar Bawah; Mighty Abra Ayidzoe; Emmanuel Botchway

Addresses: University of Energy and Natural Resources, P.O. Box 214, Sunyani, Bono Region, Ghana ' University of Energy and Natural Resources, P.O. Box 214, Sunyani, Bono Region, Ghana ' University of Energy and Natural Resources, P.O. Box 214, Sunyani, Bono Region, Ghana; University of Ghana, Legon Boundary, Accra, Ghana ' University of Energy and Natural Resources, P.O. Box 214, Sunyani, Bono Region, Ghana ' University of Energy and Natural Resources, P.O. Box 214, Sunyani, Bono Region, Ghana

Abstract: This series of studies explores carbon emission measurement and management, focusing on deep learning models. It meticulously reviews literature, identifying significant trends in using convolutional neural networks (CNN), recurrent neural networks (RNN), and long short-term memory (LSTM) architectures for carbon emission assessment. The research highlights global interest and collaboration across continents and countries, examining publication trends in major databases. Key findings include the need to update outdated guidelines, address potential biases in search and selection criteria, and broaden data sources and methodologies. The studies contribute valuable insights into understanding carbon emissions and propose future work to integrate emerging technologies, standardise methodologies, assess socio-economic impacts, and explore innovative financing mechanisms. These efforts aim to enhance our knowledge and develop effective strategies for mitigating climate change globally.

Keywords: deep learning; DL; carbon emissions; carbon footprint; measurement; convolutional neural network; CNN.

DOI: 10.1504/IJAISC.2025.151136

International Journal of Artificial Intelligence and Soft Computing, 2025 Vol.9 No.3, pp.325 - 341

Received: 02 Dec 2024
Accepted: 23 Oct 2025
Published online: 14 Jan 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article