ARCHIVES
VOL. 11, ISSUE 3 (2026)
Technique for anomaly based wild fire outbreak prediction in forest environment
Authors
Erhovwosere Dafiewhare, Abel Efetobor Edje, Chukwuemeka Augustine Obidike, Umokoro Gift
Abstract
One of the most important ecological resources
in the world is forests. However, forest fires (FFs), which harm the ecosystem
and have an effect on species and economy, pose a serious threat to them. More
precise prediction techniques are desperately needed due to the rising
incidence of FFs worldwide. The accuracy and scalability of traditional FF
forecast methods, which depend on meteorological data and human knowledge, are
frequently constrained. This research review various current techniques for the
prediction of forest fire, with detailed analysis of their processes,
challenges, effectiveness and weaknesses. Covering research published between
2018 to 2025, 25 out of 140 papers were selected for in-depth investigations
analysis. These findings show that machine learning is mostly adopted for the
prediction of FF, followed by recurrent convolutional neural network
algorithms. Models or processes such as neural network clustering were also
deployed. Furthermore, python programming language and pytorch simulation tool
were mainly adopted for the implementation and experimentation of the existing
algorithms. The research outcome also shows that misclassification, high false
positive rate and spatio-temporal differences were the common weaknesses of the
algorithms, which may lead to further research directions. Also, integrating
human activity data is still understudied. Closing this gap could improve Deep
Learning and machine learning models' usefulness for the prediction of wild
Fire in Forest environment.
Download
Pages:5-18
How to cite this article:
Erhovwosere Dafiewhare, Abel Efetobor Edje, Chukwuemeka Augustine Obidike, Umokoro Gift "Technique for anomaly based wild fire outbreak prediction in forest environment". International Journal of Advanced Science and Research, Vol 11, Issue 3, 2026, Pages 5-18
Download Author Certificate
Please enter the email address corresponding to this article submission to download your certificate.