CATENA, Journal Year: 2025, Volume and Issue: 250, P. 108801 - 108801
Published: Feb. 6, 2025
Language: Английский
CATENA, Journal Year: 2025, Volume and Issue: 250, P. 108801 - 108801
Published: Feb. 6, 2025
Language: Английский
Forests, Journal Year: 2025, Volume and Issue: 16(1), P. 122 - 122
Published: Jan. 10, 2025
Forest fires pose a significant ecological threat, particularly in the Diamer District, Gilgit-Baltistan, Pakistan, where climatic factors combined with human activities have resulted severe fire incidents. The present study sought to investigate correlation between incidence of forest and critical meteorological elements, including temperature, humidity, precipitation, wind speed, over period 25 years, from 1998 2023. We analyzed 169 recorded events, collectively burning approximately 109,400 hectares land. Employing sophisticated machine learning algorithms, Random (RF), Gradient Boosting Machine (GBM) revealed that temperature relative humidity during season, which spans May through July, are key influencing activity. Conversely, speed was found negligible impact. RF model demonstrated superior predictive accuracy compared GBM model, achieving an RMSE 5803.69 accounting for 49.47% variance burned area. This presents novel methodology risk modeling under climate change scenarios region, offering insights into management strategies. Our results underscore necessity real-time early warning systems adaptive strategies mitigate frequency intensity escalating driven by change.
Language: Английский
Citations
1CATENA, Journal Year: 2025, Volume and Issue: 250, P. 108801 - 108801
Published: Feb. 6, 2025
Language: Английский
Citations
0