AI for Natural Disasters Detection, Prediction and Modeling

Journal Name	Impact Factor	CiteScore	Launched Year	First Decision (median)	APC
AI ai	3.1	7.2	2020	18.9 Days	CHF 1600	Submit
Big Data and Cognitive Computing BDCC	3.7	7.1	2017	25.3 Days	CHF 1800	Submit
Fire fire	3.0	3.1	2018	16.5 Days	CHF 2400	Submit
GeoHazards geohazards	-	2.6	2020	19 Days	CHF 1000	Submit
Remote Sensing remotesensing	4.2	8.3	2009	23.9 Days	CHF 2700	Submit

29 pages, 13792 KiB

Open AccessArticle

Improving Fire and Smoke Detection with You Only Look Once 11 and Multi-Scale Convolutional Attention

by Yuxuan Li, Lisha Nie, Fangrong Zhou, Yun Liu, Haoyu Fu, Nan Chen, Qinling Dai and Leiguang Wang

Fire 2025, 8(5), 165; https://doi.org/10.3390/fire8050165 - 22 Apr 2025

Viewed by 259

Abstract

Fires pose significant threats to human safety, health, and property. Traditional methods, with their inefficient use of features, struggle to meet the demands of fire detection. You Only Look Once (YOLO), as an efficient deep learning object detection framework, can rapidly locate and [...] Read more.

Fires pose significant threats to human safety, health, and property. Traditional methods, with their inefficient use of features, struggle to meet the demands of fire detection. You Only Look Once (YOLO), as an efficient deep learning object detection framework, can rapidly locate and identify fire and smoke objects in visual images. However, research utilizing the latest YOLO11 for fire and smoke detection remains sparse, and addressing the scale variability of fire and smoke objects as well as the practicality of detection models continues to be a research focus. This study first compares YOLO11 with classic models in the YOLO series to analyze its advantages in fire and smoke detection tasks. Then, to tackle the challenges of scale variability and model practicality, we propose a Multi-Scale Convolutional Attention (MSCA) mechanism, integrating it into YOLO11 to create YOLO11s-MSCA. Experimental results show that YOLO11 outperforms other YOLO models by balancing accuracy, speed, and practicality. The YOLO11s-MSCA model performs exceptionally well on the D-Fire dataset, improving overall detection accuracy by 2.6% and smoke recognition accuracy by 2.8%. The model demonstrates a stronger ability to identify small fire and smoke objects. Although challenges remain in handling occluded targets and complex backgrounds, the model exhibits strong robustness and generalization capabilities, maintaining efficient detection performance in complicated environments. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

25 pages, 167605 KiB

Open AccessArticle

Development and Application of Self-Supervised Machine Learning for Smoke Plume and Active Fire Identification from the Fire Influence on Regional to Global Environments and Air Quality Datasets

by Nicholas LaHaye, Anastasija Easley, Kyongsik Yun, Hugo Lee, Erik Linstead, Michael J. Garay and Olga V. Kalashnikova

Remote Sens. 2025, 17(7), 1267; https://doi.org/10.3390/rs17071267 - 2 Apr 2025

Viewed by 449

Abstract

Fire Influence on Regional to Global Environments and Air Quality (FIREX-AQ) was a field campaign aimed at better understanding the impact of wildfires and agricultural fires on air quality and climate. The FIREX-AQ campaign took place in August 2019 and involved two aircraft [...] Read more.

Fire Influence on Regional to Global Environments and Air Quality (FIREX-AQ) was a field campaign aimed at better understanding the impact of wildfires and agricultural fires on air quality and climate. The FIREX-AQ campaign took place in August 2019 and involved two aircraft and multiple coordinated satellite observations. This study applied and evaluated a self-supervised machine learning (ML) method for the active fire and smoke plume identification and tracking in the satellite and sub-orbital remote sensing datasets collected during the campaign. Our unique methodology combines remote sensing observations with different spatial and spectral resolutions. With as much as a 10% increase in agreement between our produced masks and high-certainty hand-labeled pixels, relative to evaluated operational products, the demonstrated approach successfully differentiates active fire pixels and smoke plumes from background imagery. This enables the generation of a per-instrument smoke and active fire mask product, as well as smoke and fire masks created from the fusion of selected data from independent instruments. This ML approach has the potential to enhance operational active wildfire monitoring systems and improve decision-making in air quality management through fast smoke plume identification and tracking and could improve climate impact studies through fusion data from independent instruments. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

19 pages, 5855 KiB

Open AccessArticle

Predicting Railway Slope Failure Under Heavy Rainfall Using the Soil Moisture Extended Cohesive Damage Element Method

by Sudath Loku-Pathirage, Jiye Chen and Min Fu

GeoHazards 2025, 6(1), 14; https://doi.org/10.3390/geohazards6010014 - 13 Mar 2025

Viewed by 489

Abstract

Slope failure, as a natural disaster, can cause extensive human suffering and financial losses worldwide. This paper introduces a new soil moisture extended cohesive damage element (SMECDE) method to predict railway slope failure under heavy rainfall. A correlation between rainfall intensity and soil [...] Read more.

Slope failure, as a natural disaster, can cause extensive human suffering and financial losses worldwide. This paper introduces a new soil moisture extended cohesive damage element (SMECDE) method to predict railway slope failure under heavy rainfall. A correlation between rainfall intensity and soil moisture content is first established to create an equivalence between the two. Considering slope failure mechanisms dominated by the loss of soil or the cohesion of slope materials due to heavy rainfall infiltration, the soil moisture decohesion model (SMDM) is developed using previous experimental data to express how soil cohesion varies with different soil moistures and depths. The SMDM is incorporated into the extended cohesive damage element (ECDE) method to fundamentally study slope failure mechanisms under varying soil moisture levels and depths. The proposed SMECDE approach is used to predict the failure propagation of a selected railway embankment slope at the critical soil moisture or rainfall intensity. This SMECDE failure prediction is validated using relevant data from previous fieldwork and meteorological reports on the critical rainfall intensity at the site. Additionally, the corresponding slope damage scale prediction is validated with a large plastic deformation analysis using the commercial FEM package ABAQUS. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

16 pages, 37656 KiB

Open AccessArticle

Smoke and Fire-You Only Look Once: A Lightweight Deep Learning Model for Video Smoke and Flame Detection in Natural Scenes

by Chenmeng Zhao, Like Zhao, Ka Zhang, Yinghua Ren, Hui Chen and Yehua Sheng

Fire 2025, 8(3), 104; https://doi.org/10.3390/fire8030104 - 4 Mar 2025

Viewed by 925

Abstract

Owing to the demand for smoke and flame detection in natural scenes, this paper proposes a lightweight deep learning model, SF-YOLO (Smoke and Fire-YOLO), for video smoke and flame detection in such environments. Firstly, YOLOv11 is employed as the backbone network, combined with [...] Read more.

Owing to the demand for smoke and flame detection in natural scenes, this paper proposes a lightweight deep learning model, SF-YOLO (Smoke and Fire-YOLO), for video smoke and flame detection in such environments. Firstly, YOLOv11 is employed as the backbone network, combined with the C3k2 module based on a two-path residual attention mechanism, and a target detection head frame with an embedded attention mechanism. This combination enhances the response of the unobscured regions to compensate for the feature loss in occluded regions, thereby addressing the occlusion problem in dynamic backgrounds. Then, a two-channel loss function (W-SIoU) based on dynamic tuning and intelligent focusing is designed to enhance loss computation in the boundary regions, thus improving the YOLOv11 model’s ability to recognize targets with ambiguous boundaries. Finally, the algorithms proposed in this paper are experimentally validated using the self-generated dataset S-Firedata and the public smoke and flame virtual dataset M⁴SFWD. These datasets are derived from internet smoke and flame video frame extraction images and open-source smoke and flame dataset images, respectively. The experimental results demonstrate, compared with deep learning models such as YOLOv8, Gold-YOLO, and Faster-RCNN, the SF-YOLO model proposed in this paper is more lightweight and exhibits higher detection accuracy and robustness. The metrics mAP50 and mAP50-95 are improved by 2.5% and 2.4%, respectively, in the self-made dataset S-Firedata, and by 0.7% and 1.4%, respectively, in the publicly available dataset M⁴SFWD. The research presented in this paper provides practical methods for the automatic detection of smoke and flame in natural scenes, which can further enhance the effectiveness of fire monitoring systems. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

24 pages, 6622 KiB

Open AccessArticle

DBSANet: A Dual-Branch Semantic Aggregation Network Integrating CNNs and Transformers for Landslide Detection in Remote Sensing Images

by Yankui Li, Wu Zhu, Jing Wu, Ruixuan Zhang, Xueyong Xu and Ye Zhou

Remote Sens. 2025, 17(5), 807; https://doi.org/10.3390/rs17050807 - 25 Feb 2025

Viewed by 400

Abstract

Deep learning-based semantic segmentation algorithms have proven effective in landslide detection. For the past decade, convolutional neural networks (CNNs) have been the prevailing approach for semantic segmentation. Nevertheless, the intrinsic limitations of convolutional operations hinder the acquisition of global contextual information. Recently, Transformers [...] Read more.

Deep learning-based semantic segmentation algorithms have proven effective in landslide detection. For the past decade, convolutional neural networks (CNNs) have been the prevailing approach for semantic segmentation. Nevertheless, the intrinsic limitations of convolutional operations hinder the acquisition of global contextual information. Recently, Transformers have garnered attention for their exceptional global modeling capabilities. This study proposes a dual-branch semantic aggregation network (DBSANet) by integrating ResNet and a Swin Transformer. A Feature Fusion Module (FFM) is designed to effectively integrate semantic information extracted from the ResNet and Swin Transformer branches. Considering the significant semantic gap between the encoder and decoder, a Spatial Gate Attention Module (SGAM) is used to suppress the noise from the decoder feature maps during decoding and guides the encoder feature maps based on its output, thereby reducing the semantic gap during the fusion of low-level and high-level semantic information. The DBSANet model demonstrated superior performance compared to existing models such as UNet, Deeplabv3+, ResUNet, SwinUNet, TransUNet, TransFuse, and UNetFormer on the Bijie and Luding datasets, achieving IoU values of 77.12% and 75.23%, respectively, with average improvements of 4.91% and 2.96%. This study introduces a novel perspective for landslide detection based on remote sensing images, focusing on how to effectively integrate the strengths of CNNs and Transformers for their application in landslide detection. Furthermore, it offers technical support for the application of hybrid models in landslide detection. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

17 pages, 4466 KiB

Open AccessArticle

Flooded Infrastructure Change Detection in Deeply Supervised Networks Based on Multi-Attention-Constrained Multi-Scale Feature Fusion

by Gang Qin, Shixin Wang, Futao Wang, Suju Li, Zhenqing Wang, Jinfeng Zhu, Ming Liu, Changjun Gu and Qing Zhao

Remote Sens. 2024, 16(22), 4328; https://doi.org/10.3390/rs16224328 - 20 Nov 2024

Viewed by 915

Abstract

Flood disasters are frequent, sudden, and have significant chain effects, seriously damaging infrastructure. Remote sensing images provide a means for timely flood emergency monitoring. When floods occur, emergency management agencies need to respond quickly and assess the damage. However, manual evaluation takes a [...] Read more.

Flood disasters are frequent, sudden, and have significant chain effects, seriously damaging infrastructure. Remote sensing images provide a means for timely flood emergency monitoring. When floods occur, emergency management agencies need to respond quickly and assess the damage. However, manual evaluation takes a significant amount of time; in current, commercial applications, the post-disaster flood vector range is used to directly overlay land cover data. On the one hand, land cover data are not updated in time, resulting in the misjudgment of disaster losses; on the other hand, since buildings block floods, the above methods cannot detect flooded buildings. Automated change-detection methods can effectively alleviate the above problems. However, the ability of change-detection structures and deep learning models for flooding to characterize flooded buildings and roads is unclear. This study specifically evaluated the performance of different change-detection structures and different deep learning models for the change detection of flooded buildings and roads in very-high-resolution remote sensing images. At the same time, a plug-and-play, multi-attention-constrained, deeply supervised high-dimensional and low-dimensional multi-scale feature fusion (MSFF) module is proposed. The MSFF module was extended to different deep learning models. Experimental results showed that the embedded MSFF performs better than the baseline model, demonstrating that MSFF can be used as a general multi-scale feature fusion component. After FloodedCDNet introduced MSFF, the detection accuracy of flooded buildings and roads changed after the data augmentation reached a maximum of 69.1% MIoU. This demonstrates its effectiveness and robustness in identifying change regions and categories from very-high-resolution remote sensing images. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

25 pages, 34342 KiB

Open AccessArticle

Quantifying the Geomorphological Susceptibility of the Piping Erosion in Loess Using LiDAR-Derived DEM and Machine Learning Methods

by Sisi Li, Sheng Hu, Lin Wang, Fanyu Zhang, Ninglian Wang, Songbai Wu, Xingang Wang and Zongda Jiang

Remote Sens. 2024, 16(22), 4203; https://doi.org/10.3390/rs16224203 - 11 Nov 2024

Viewed by 1088

Abstract

Soil piping erosion is an underground soil erosion process that is significantly underestimated or overlooked. It can lead to intense soil erosion and trigger surface processes such as landslides, collapses, and channel erosion. Conducting susceptibility mapping is a vital way to identify the [...] Read more.

Soil piping erosion is an underground soil erosion process that is significantly underestimated or overlooked. It can lead to intense soil erosion and trigger surface processes such as landslides, collapses, and channel erosion. Conducting susceptibility mapping is a vital way to identify the potential for soil piping erosion, which is of enormous significance for soil and water conservation as well as geological disaster prevention. This study utilized airborne radar drones to survey and map 1194 sinkholes in Sunjiacha basin, Huining County, on the Loess Plateau in Northwest China. We identified seventeen key hydrogeomorphological factors that influence sinkhole susceptibility and used six machine learning models—support vector machine (SVM), logistic regression (LR), Convolutional Neural Network (CNN), K-Nearest Neighbors (KNN), random forest (RF), and gradient boosting decision tree (GBDT)—for the susceptibility assessment and mapping of loess sinkholes. We then evaluated and validated the prediction results of various models using the area under curve (AUC) of the Receiver Operating Characteristic Curve (ROC). The results showed that all six of these machine learning algorithms had an AUC of more than 0.85. The GBDT model had the best predictive accuracy (AUC = 0.94) and model migration performance (AUC = 0.93), and it could find sinkholes with high and very high susceptibility levels in loess areas. This suggests that the GBDT model is well suited for the fine-scale susceptibility mapping of sinkholes in loess regions. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

24 pages, 8367 KiB

Open AccessArticle

Detecting Hailstorms in China from FY-4A Satellite with an Ensemble Machine Learning Model

by Qiong Wu, Yi-Xuan Shou, Yong-Guang Zheng, Fei Wu and Chun-Yuan Wang

Remote Sens. 2024, 16(18), 3354; https://doi.org/10.3390/rs16183354 - 10 Sep 2024

Viewed by 1067

Abstract

Hail poses a significant meteorological hazard in China, leading to substantial economic and agricultural damage. To enhance the detection of hail and mitigate these impacts, this study presents an ensemble machine learning model (BPNN+Dtree) that combines a backpropagation neural network (BPNN) and a [...] Read more.

Hail poses a significant meteorological hazard in China, leading to substantial economic and agricultural damage. To enhance the detection of hail and mitigate these impacts, this study presents an ensemble machine learning model (BPNN+Dtree) that combines a backpropagation neural network (BPNN) and a decision tree (Dtree). Using FY-4A satellite and ERA5 reanalysis data, the model is trained on geostationary satellite infrared data and environmental parameters, offering comprehensive, all-day, and large-area hail monitoring over China. The ReliefF method is employed to select 13 key features from 29 physical quantities, emphasizing cloud-top and thermodynamic properties over dynamic ones as input features for the model to enhance its hail differentiation capability. The BPNN+Dtree ensemble model harnesses the strengths of both algorithms, improving the probability of detection (POD) to 0.69 while maintaining a reasonable false alarm ratio (FAR) on the test set. Moreover, the model’s spatial distribution of hail probability more closely matches the observational data, outperforming the individual BPNN and Dtree models. Furthermore, it demonstrates improved regional applicability over overshooting top (OT)-based methods in the China region. The identified high-frequency hail areas correspond to the north-south movement of the monsoon rain belt and are consistent with the northeast-southwest belt distribution observed using microwave-based methods. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

22 pages, 20392 KiB

Open AccessArticle

AI-Driven Computer Vision Detection of Cotton in Corn Fields Using UAS Remote Sensing Data and Spot-Spray Application

by Pappu Kumar Yadav, J. Alex Thomasson, Robert Hardin, Stephen W. Searcy, Ulisses Braga-Neto, Sorin C. Popescu, Roberto Rodriguez III, Daniel E. Martin and Juan Enciso

Remote Sens. 2024, 16(15), 2754; https://doi.org/10.3390/rs16152754 - 27 Jul 2024

Viewed by 1513

Abstract

To effectively combat the re-infestation of boll weevils (Anthonomus grandis L.) in cotton fields, it is necessary to address the detection of volunteer cotton (VC) plants (Gossypium hirsutum L.) in rotation crops such as corn (Zea mays L.) and sorghum ( [...] Read more.

To effectively combat the re-infestation of boll weevils (Anthonomus grandis L.) in cotton fields, it is necessary to address the detection of volunteer cotton (VC) plants (Gossypium hirsutum L.) in rotation crops such as corn (Zea mays L.) and sorghum (Sorghum bicolor L.). The current practice involves manual field scouting at the field edges, which often leads to the oversight of VC plants growing in the middle of fields alongside corn and sorghum. As these VC plants reach the pinhead squaring stage (5–6 leaves), they can become hosts for boll weevil pests. Consequently, it becomes crucial to detect, locate, and accurately spot-spray these plants with appropriate chemicals. This paper focuses on the application of YOLOv5m to detect and locate VC plants during the tasseling (VT) growth stage of cornfields. Our results demonstrate that VC plants can be detected with a mean average precision (mAP) of 79% at an Intersection over Union (IoU) of 50% and a classification accuracy of 78% on images sized 1207 × 923 pixels. The average detection inference speed is 47 frames per second (FPS) on the NVIDIA Tesla P100 GPU-16 GB and 0.4 FPS on the NVIDIA Jetson TX2 GPU, which underscores the relevance and impact of detection speed on the feasibility of real-time applications. Additionally, we show the application of a customized unmanned aircraft system (UAS) for spot-spray applications through simulation based on the developed computer vision (CV) algorithm. This UAS-based approach enables the near-real-time detection and mitigation of VC plants in corn fields, with near-real-time defined as approximately 0.02 s per frame on the NVIDIA Tesla P100 GPU and 2.5 s per frame on the NVIDIA Jetson TX2 GPU, thereby offering an efficient management solution for controlling boll weevil pests. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

16 pages, 6768 KiB

Open AccessArticle

Landslide Susceptibility Assessment in Active Tectonic Areas Using Machine Learning Algorithms

by Tianjun Qi, Xingmin Meng and Yan Zhao

Remote Sens. 2024, 16(15), 2724; https://doi.org/10.3390/rs16152724 - 25 Jul 2024

Viewed by 1315

Abstract

The eastern margin of the Tibetan Plateau is one of the regions with the most severe landslide disasters on a global scale. With the intensification of seismic activity around the Tibetan Plateau and the increase in extreme rainfall events, the prevention of landslide [...] Read more.

The eastern margin of the Tibetan Plateau is one of the regions with the most severe landslide disasters on a global scale. With the intensification of seismic activity around the Tibetan Plateau and the increase in extreme rainfall events, the prevention of landslide disasters in the region is facing serious challenges. This article selects the Bailong River Basin located in this region as the research area, and the historical landslide data obtained from high-precision remote sensing image interpretation combined with field validation are used as the sample library. Using machine learning algorithms and data-driven landslide susceptibility assessment as the methods, 17 commonly used models and 17 important factors affecting the development of landslides are selected to carry out the susceptibility assessment. The results show that the BaggingClassifier model shows advantageous applicability in the region, and the landslide susceptibility distribution map of the Bailong River Basin was generated using this model. The results show that the road and population density are both high in very high and high susceptible areas, indicating that there is still a significant potential landslide risk in the basin. The quantitative evaluation of the main influencing factors emphasizes that distance to a road is the most important factor. However, due to the widespread utilization of ancient landslides by local residents for settlement and agricultural cultivation over hundreds of years, the vast majority of landslides are likely to have occurred prior to human settlement. Therefore, the importance of this factor may be overestimated, and the evaluation of the factors still needs to be dynamically examined in conjunction with the development history of the region. The five factors of NDVI, altitude, faults, average annual rainfall, and rivers have a secondary impact on landslide susceptibility. The research results have important significance for the susceptibility assessment of landslides in the complex environment of human–land interaction and for the construction of landslide disaster monitoring and early warning systems in the Bailong River Basin. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

22 pages, 11376 KiB

Open AccessArticle

Robust Landslide Recognition Using UAV Datasets: A Case Study in Baihetan Reservoir

by Zhi-Hai Li, An-Chi Shi, Huai-Xian Xiao, Zi-Hao Niu, Nan Jiang, Hai-Bo Li and Yu-Xiang Hu

Remote Sens. 2024, 16(14), 2558; https://doi.org/10.3390/rs16142558 - 12 Jul 2024

Cited by 1 | Viewed by 1496

Abstract

The task of landslide recognition focuses on extracting the location and extent of landslides over large areas, providing ample data support for subsequent landslide research. This study explores the use of UAV and deep learning technologies to achieve robust landslide recognition in a [...] Read more.

The task of landslide recognition focuses on extracting the location and extent of landslides over large areas, providing ample data support for subsequent landslide research. This study explores the use of UAV and deep learning technologies to achieve robust landslide recognition in a more rational, simpler, and faster manner. Specifically, the widely successful DeepLabV3+ model was used as a blueprint and a dual-encoder design was introduced to reconstruct a novel semantic segmentation model consisting of Encoder1, Encoder2, Mixer and Decoder modules. This model, named DeepLab for Landslide (DeepLab4LS), considers topographic information as a supplement to DeepLabV3+, and is expected to improve the efficiency of landslide recognition by extracting shape information from relative elevation, slope, and hillshade. Additionally, a novel loss function term—Positive Enhanced loss (PE loss)—was incorporated into the training of DeepLab4LS, significantly enhancing its ability to understand positive samples. DeepLab4LS was then applied to a UAV dataset of Baihetan reservoir, where comparative tests demonstrated its high performance in landslide recognition tasks. We found that DeepLab4LS has a stronger inference capability for landslides with less distinct boundary information, and delineates landslide boundaries more precisely. More specifically, in terms of evaluation metrics, DeepLab4LS achieved a mean intersection over union (mIoU) of 76.0% on the validation set, which is a substantial 5.5 percentage point improvement over DeepLabV3+. Moreover, the study also validated the rationale behind the dual-encoder design and the introduction of PE loss through ablation experiments. Overall, this research presents a robust semantic segmentation model for landslide recognition that considers both optical and topographic semantics of landslides, emulating the recognition pathways of human experts, and is highly suitable for landslide recognition based on UAV datasets. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

22 pages, 7648 KiB

Open AccessArticle

Fire-RPG: An Urban Fire Detection Network Providing Warnings in Advance

by Xiangsheng Li and Yongquan Liang

Fire 2024, 7(7), 214; https://doi.org/10.3390/fire7070214 - 26 Jun 2024

Cited by 3 | Viewed by 1751

Abstract

Urban fires are characterized by concealed ignition points and rapid escalation, making the traditional methods of detecting early stage fire accidents inefficient. Thus, we focused on the features of early stage fire accidents, such as faint flames and thin smoke, and established a [...] Read more.

Urban fires are characterized by concealed ignition points and rapid escalation, making the traditional methods of detecting early stage fire accidents inefficient. Thus, we focused on the features of early stage fire accidents, such as faint flames and thin smoke, and established a dataset. We found that these features are mostly medium-sized and small-sized objects. We proposed a model based on YOLOv8s, Fire-RPG. Firstly, we introduced an extra very small object detection layer to enhance the detection performance for early fire features. Next, we optimized the model structure with the bottleneck in GhostV2Net, which reduced the computational time and the parameters. The Wise-IoUv3 loss function was utilized to decrease the harmful effects of low-quality data in the dataset. Finally, we integrated the low-cost yet high-performance RepVGG block and the CBAM attention mechanism to enhance learning capabilities. The RepVGG block enhances the extraction ability of the backbone and neck structures, while CBAM focuses the attention of the model on specific size objects. Our experiments showed that Fire-RPG achieved an mAP of 81.3%, an improvement of 2.2%. In addition, Fire-RPG maintained high detection performance across various fire scenarios. Therefore, our model can provide timely warnings and accurate detection services. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

15 pages, 3788 KiB

Open AccessArticle

Wildfire Susceptibility Prediction Based on a CA-Based CCNN with Active Learning Optimization

by Qiuping Yu, Yaqin Zhao, Zixuan Yin and Zhihao Xu

Fire 2024, 7(6), 201; https://doi.org/10.3390/fire7060201 - 16 Jun 2024

Cited by 4 | Viewed by 1165

Abstract

Wildfires cause great losses to the ecological environment, economy, and people’s safety and belongings. As a result, it is crucial to establish wildfire susceptibility models and delineate fire risk levels. It has been proven that the use of remote sensing data, such as [...] Read more.

Wildfires cause great losses to the ecological environment, economy, and people’s safety and belongings. As a result, it is crucial to establish wildfire susceptibility models and delineate fire risk levels. It has been proven that the use of remote sensing data, such as meteorological and topographical data, can effectively predict and evaluate wildfire susceptibility. Accordingly, this paper converts meteorological and topographical data into fire-influencing factor raster maps for wildfire susceptibility prediction. The continuous convolutional neural network (CCNN for short) based on coordinate attention (CA for short) can aggregate different location information into channels of the network so as to enhance the feature expression ability; moreover, for different patches with different resolutions, the improved CCNN model does not need to change the structural parameters of the network, which improves the flexibility of the network application in different forest areas. In order to reduce the annotation of training samples, we adopt an active learning method to learn positive features by selecting high-confidence samples, which contributes to enhancing the discriminative ability of the network. We use fire probabilities output from the model to evaluate fire risk levels and generate the fire susceptibility map. Taking Chongqing Municipality in China as an example, the experimental results show that the CA-based CCNN model has a better classification performance; the accuracy reaches 91.7%, and AUC reaches 0.9487, which is 5.1% and 2.09% higher than the optimal comparative method, respectively. Furthermore, if an accuracy of about 86% is desired, our method only requires 50% of labeled samples and thus saves about 20% and 40% of the labeling efforts compared to the other two methods, respectively. Ultimately, the proposed model achieves the balance of high prediction accuracy and low annotation cost and is more helpful in classifying fire high warning zones and fire-free zones. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

12 pages, 3871 KiB

Open AccessArticle

Multitemporal Dynamics of Fuels in Forest Systems Present in the Colombian Orinoco River Basin Forests

by Walter Garcia-Suabita, Mario José Pacheco and Dolors Armenteras

Fire 2024, 7(6), 171; https://doi.org/10.3390/fire7060171 - 21 May 2024

Cited by 3 | Viewed by 1288

Abstract

In Colombia’s Orinoco, wildfires have a profound impact on ecosystem dynamics, particularly affecting savannas and forest–savanna transitions. Human activities have disrupted the natural fire regime, leading to increased wildfire frequency due to changes in land use, deforestation, and climate change. Despite extensive research [...] Read more.

In Colombia’s Orinoco, wildfires have a profound impact on ecosystem dynamics, particularly affecting savannas and forest–savanna transitions. Human activities have disrupted the natural fire regime, leading to increased wildfire frequency due to changes in land use, deforestation, and climate change. Despite extensive research on fire monitoring and prediction, the quantification of fuel accumulation, a critical factor in fire incidence, remains inadequately explored. This study addresses this gap by quantifying dead organic material (detritus) accumulation and identifying influencing factors. Using Brown transects across forests with varying fire intensities, we assessed fuel loads and characterized variables related to detritus accumulation over time. Employing factor analysis, principal components analysis, and a generalized linear mixed model, we determined the effects of various factors. Our findings reveal significant variations in biomass accumulation patterns influenced by factors such as thickness, wet and dry mass, density, gravity, porosity, and moisture content. Additionally, a decrease in fuel load over time was attributed to increased precipitation from three La Niña events. These insights enable more accurate fire predictions and inform targeted forest management strategies for fire prevention and mitigation, thereby enhancing our understanding of fire ecology in the Orinoco basin and guiding effective conservation practices. Full article

(This article belongs to the Topic AI for Natural Disasters Detection, Prediction and Modeling)

► Show Figures

Figure 1

Topic Menu

Topic Editors

AI for Natural Disasters Detection, Prediction and Modeling

Topic Information

Keywords

Participating Journals

Published Papers (14 papers)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI