Advanced Wildfire Prediction in Morocco: Developing a Deep Learning Dataset from Multisource Observations
Ayoub Jadouli, Chaker El Amrani
TL;DR
This work addresses the need for accurate wildfire prediction in Morocco by creating a novel, publicly available multisource dataset that merges satellite observations (NDVI, soil moisture, FIRMS) with ground-based meteorology and human activity indicators. A comprehensive methodology combines data collection, preprocessing, feature augmentation, and a GPU-accelerated evaluation of traditional ML models and DL architectures, with a grid search guiding deep learning configurations. The results show that deep learning models, alongside strong gradient-boosting baselines like LightGBM, achieve high predictive performance (up to ~0.90 accuracy and strong AUC-PR), aided by permutation feature importance analyses that highlight latitude, NDVI, and sea-distance as key drivers. The dataset and code are publicly available to promote replication, benchmarking, and adaptation to other regions, enabling rapid advancement in localized wildfire risk management and proactive mitigation strategies.
Abstract
Wildfires pose significant threats to ecosystems, economies, and communities worldwide, necessitating advanced predictive methods for effective mitigation. This study introduces a novel and comprehensive dataset specifically designed for wildfire prediction in Morocco, addressing its unique geographical and climatic challenges. By integrating satellite observations and ground station data, we compile essential environmental indicators such as vegetation health (NDVI), population density, soil moisture levels, and meteorological data aimed at predicting next-day wildfire occurrences with high accuracy. Our methodology incorporates state-of-the-art machine learning and deep learning algorithms, demonstrating superior performance in capturing wildfire dynamics compared to traditional models. Preliminary results show that models using this dataset achieve an accuracy of up to 90%, significantly improving prediction capabilities. The public availability of this dataset fosters scientific collaboration, aiming to refine predictive models and develop innovative wildfire management strategies. Our work not only advances the technical field of dataset creation but also emphasizes the necessity for localized research in underrepresented regions, providing a scalable model for other areas facing similar environmental challenges.