A Spatiotemporal Contextual Model for Forest Fire Detection Using Himawari-8 Satellite Data

Early Stage Forest Fire Detection from Himawari-8 AHI Images Using a Modified MOD14 Algorithm Combined with Machine Learning

The early detection and rapid extinguishing of forest fires are effective in reducing their spread. Based on the MODIS Thermal Anomaly (MOD14) algorithm, we propose an early stage fire detection method from low-spatial-resolution but high-temporal-resolution images, observed by the Advanced Himawari Imager (AHI) onboard the geostationary meteorological satellite Himawari-8. In order to not miss early stage forest fire pixels with low temperature, we omit the potential fire pixel detection from the MOD14 algorithm and parameterize four contextual conditions included in the MOD14 algorithm as features. The proposed method detects fire pixels from forest areas using a random forest classifier taking these contextual parameters, nine AHI band values, solar zenith angle, and five meteorological values as inputs. To evaluate the proposed method, we trained the random forest classifier using an early stage forest fire data set generated by a time-reversal approach with MOD14 products and time-series AHI images in Australia. The results demonstrate that the proposed method with all parameters can detect fire pixels with about 90% precision and recall, and that the contribution of contextual parameters is particularly significant in the random forest classifier. The proposed method is applicable to other geostationary and polar-orbiting satellite sensors, and it is expected to be used as an effective method for forest fire detection.

Multifractal Analysis of MODIS Aqua and Terra Satellite Time Series of Normalized Difference Vegetation Index and Enhanced Vegetation Index of Sites Affected by Wildfires

The MODIS Aqua and Terra Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) time series acquired during nearly two decades (2000 to 2020) covering the area burned by the Camp Fire (California) in 2018 is investigated in this study by using the multifractal detrended fluctuation analysis in relation to the recovery process of vegetation after fire. In 2008, the same area was partially burned by two wildfires, the BTU Lightning Complex Fire and the Humboldt Fire. Our results indicate that all vegetation index time series are featured by six- and twelve-month modulating periodicities, with a larger spectral content at longer periods for two-fire-affected sites. Furthermore, two fires cause an increase of the persistence of the NDVI and EVI time series and an increase of the complexity, suggesting that the recovery process of vegetation dynamics of fire-affected sites is characterized by positive feedback mechanisms, driving the growth-generating phenomena, which become even more effective in those sites affected by two fires.

Development of a Novel Burned-Area Subpixel Mapping (BASM) Workflow for Fire Scar Detection at Subpixel Level

The accurate detection of burned forest area is essential for post-fire management and assessment, and for quantifying carbon budgets. Therefore, it is imperative to map burned areas accurately. Currently, there are few burned-area products around the world. Researchers have mapped burned areas directly at the pixel level that is usually a mixture of burned area and other land cover types. In order to improve the burned area mapping at subpixel level, we proposed a Burned Area Subpixel Mapping (BASM) workflow to map burned areas at the subpixel level. We then applied the workflow to Sentinel 2 data sets to obtain burned area mapping at subpixel level. In this study, the information of true fire scar was provided by the Department of Emergency Management of Hunan Province, China. To validate the accuracy of the BASM workflow for detecting burned areas at the subpixel level, we applied the workflow to the Sentinel 2 image data and then compared the detected burned area at subpixel level with in situ measurements at fifteen fire-scar reference sites located in Hunan Province, China. Results show the proposed method generated successfully burned area at the subpixel level. The methods, especially the BASM-Feature Extraction Rule Based (BASM-FERB) method, could minimize misclassification and effects due to noise more effectively compared with the BASM-Random Forest (BASM-RF), BASM-Backpropagation Neural Net (BASM-BPNN), BASM-Support Vector Machine (BASM-SVM), and BASM-notra methods. We conducted a comparison study among BASM-FERB, BASM-RF, BASM-BPNN, BASM-SVM, and BASM-notra using five accuracy evaluation indices, i.e., overall accuracy (OA), user’s accuracy (UA), producer’s accuracy (PA), intersection over union (IoU), and Kappa coefficient (Kappa). The detection accuracy of burned area at the subpixel level by BASM-FERB’s OA, UA, IoU, and Kappa is 98.11%, 81.72%, 74.32%, and 83.98%, respectively, better than BASM-RF’s, BASM-BPNN’s, BASM-SVM’s, and BASM-notra’s, even though BASM-RF’s and BASM-notra’s average PA is higher than BASM-FERB’s, with 89.97%, 91.36%, and 89.52%, respectively. We conclude that the newly proposed BASM workflow can map burned areas at the subpixel level, providing greater accuracy in regards to the burned area for post-forest fire management and assessment.

Investigating the Impact of Using IR Bands on Early Fire Smoke Detection from Landsat Imagery with a Lightweight CNN Model

Smoke plumes are the first things seen from space when wildfires occur. Thus, fire smoke detection is important for early fire detection. Deep Learning (DL) models have been used to detect fire smoke in satellite imagery for fire detection. However, previous DL-based research only considered lower spatial resolution sensors (e.g., Moderate-Resolution Imaging Spectroradiometer (MODIS)) and only used the visible (i.e., red, green, blue (RGB)) bands. To contribute towards solutions for early fire smoke detection, we constructed a six-band imagery dataset from Landsat 5 Thematic Mapper (TM) and Landsat 8 Operational Land Imager (OLI) with a 30-metre spatial resolution. The dataset consists of 1836 images in three classes, namely “Smoke”, “Clear”, and “Other_aerosol”. To prepare for potential on-board-of-small-satellite detection, we designed a lightweight Convolutional Neural Network (CNN) model named “Variant Input Bands for Smoke Detection (VIB_SD)”, which achieved competitive accuracy with the state-of-the-art model SAFA, with less than 2% of its number of parameters. We further investigated the impact of using additional Infra-Red (IR) bands on the accuracy of fire smoke detection with VIB_SD by training it with five different band combinations. The results demonstrated that adding the Near-Infra-Red (NIR) band improved prediction accuracy compared with only using the visible bands. Adding both Short-Wave Infra-Red (SWIR) bands can further improve the model performance compared with adding only one SWIR band. The case study showed that the model trained with multispectral bands could effectively detect fire smoke mixed with cloud over small geographic extents.

Forest Fire Detection of FY-3D Using Genetic Algorithm and Brightness Temperature Change

As one of China’s new generation polar-orbiting meteorological satellites, FengYun-3D (FY-3D) provides critical data for forest fire detection. Most of the existing related methods identify fire points by comparing the spatial features and setting thresholds empirically. However, they ignore temporal features that are associated with forest fires. Besides, they are difficult to generalize to multiple areas with different environmental characteristics. A novel method based on FY-3D combining the genetic algorithm and brightness temperature change detection is proposed in this work to improve these problems. After analyzing the spatial features of the FY-3D data, it adaptively detects potential fire points based on these features using the genetic algorithm, then filters the points with contextual information. To address the false alarms resulting from the confusing spectral characteristics between fire pixels and conventional hotspots, temporal information is introduced and the “MIR change rate” based on the multitemporal brightness temperature change is further proposed. In order to evaluate the performance of the proposed algorithm, several fire events occurring in different areas are used for testing. The Moderate-Resolution Imaging Spectroradiometer (MODIS) Thermal Anomalies/Fire products (MYD14) is chosen as the validation data to assess the accuracy of the proposed algorithm. A comparison of results demonstrates that the algorithm can identify fire points effectively and obtain a higher accuracy than the previous FY-3D algorithm.

Forest Fire Monitoring and Positioning Improvement at Subpixel Level: Application to Himawari-8 Fire Products

Forest fires are among the biggest threats to forest ecosystems and forest resources, and can lead to ecological disasters and social crises. Therefore, it is imperative to detect and extinguish forest fires in time to reduce their negative impacts. Satellite remote sensing, especially meteorological satellites, has been a useful tool for forest-fire detection and monitoring because of its high temporal resolution over large areas. Researchers monitor forest fires directly at pixel level, which usually presents a mixture of forest and fire, but the low spatial resolution of such mixed pixels cannot accurately locate the exact position of the fire, and the optimal time window for fire suppression can thus be missed. In order to improve the positioning accuracy of the origin of forest fire (OriFF), we proposed a mixed-pixel unmixing integrated with pixel-swapping algorithm (MPU-PSA) model to monitor the OriFFs in time. We then applied the model to the Japanese Himawari-8 Geostationary Meteorological Satellite data to obtain forest-fire products at subpixel level. In this study, the ground truth data were provided by the Department of Emergency Management of Hunan Province, China. To validate the positioning accuracy of MPU-PSA for OriFFs, we applied the model to the Himawari-8 satellite data and then compared the derived fire results with fifteen reference forest-fire events that occurred in Hunan Province, China. The results show that the extracted forest-fire locations using the proposed method, referred to as forest fire locations at subpixel (FFLS) level, were far closer to the actual OriFFs than those from the modified Himawari-8 Wild Fire Product (M-HWFP). This improvement will help to reduce false fire claims in the Himawari-8 Wild Fire Product (HWFP). We conducted a comparative study of M-HWFP and FFLS products using three accuracy-evaluation indexes, i.e., Euclidean distance, RMSE, and MAE. The mean distances between M-HWFP fire locations and OriFFs and between FFLS fire locations and OriFFs were 3362.21 m and 1294.00 m, respectively. The mean RMSEs of the M-HWFP and FFLS products are 1225.52 m and 474.93 m, respectively. The mean MAEs of the M-HWFP and FFLS products are 992.12 m and 387.13 m, respectively. We concluded that the newly proposed MPU-PSA method can extract forest-fire locations at subpixel level, providing higher positioning accuracy of forest fires for their suppression.

Mapping Fire Susceptibility in the Brazilian Amazon Forests Using Multitemporal Remote Sensing and Time-Varying Unsupervised Anomaly Detection

The economic and environmental impacts of wildfires have leveraged the development of new technologies to prevent and reduce the occurrence of these devastating events. Indeed, identifying and mapping fire-susceptible areas arise as critical tasks, not only to pave the way for rapid responses to attenuate the fire spreading, but also to support emergency evacuation plans for the families affected by fire-related tragedies. Aiming at simultaneously mapping and measuring the risk of fires in the forest areas of Brazil’s Amazon, in this paper we combine multitemporal remote sensing, derivative spectral indices, and anomaly detection into a fully unsupervised methodology. We focus our analysis on recent forest fire events that occurred in the Brazilian Amazon by exploring multitemporal images acquired by both Landsat-8 Operational Land Imager and Modis sensors. We experimentally confirm that the current methodology is capable of predicting fire outbreaks immediately at posterior instants, which attests to the operational performance and applicability of our approach to preventing and mitigating the impact of fires in Brazilian forest regions.

Adversarial Fusion Network for Forest Fire Smoke Detection

Recent advances suggest that deep learning has been widely used to detect smoke for early forest fire warnings. Despite its remarkable success, this approach has a number of problems in real life application. Deep neural networks only learn deep and abstract representations, while ignoring shallow and detailed representations. In addition, previous models have been trained on source domains but have generalized weakly on unseen domains. To cope with these problems, in this paper, we propose an adversarial fusion network (AFN), including a feature fusion network and an adversarial feature-adaptation network for forest fire smoke detection. Specifically, the feature fusion network is able to learn more discriminative representations by fusing abstract and detailed features. Meanwhile, the adversarial feature adaptation network is employed to improve the generalization ability and transfer gains of the AFN. Comprehensive experiments on two self-built forest fire smoke datasets, and three publicly available smoke datasets, validate that our method significantly improves the performance and generalization of smoke detection, particularly the accuracy of the detection of small amounts of smoke.

Active Fire Detection from Landsat-8 Imagery Using Deep Multiple Kernel Learning

Active fires are devastating natural disasters that cause socio-economical damage across the globe. The detection and mapping of these disasters require efficient tools, scientific methods, and reliable observations. Satellite images have been widely used for active fire detection (AFD) during the past years due to their nearly global coverage. However, accurate AFD and mapping in satellite imagery is still a challenging task in the remote sensing community, which mainly uses traditional methods. Deep learning (DL) methods have recently yielded outstanding results in remote sensing applications. Nevertheless, less attention has been given to them for AFD in satellite imagery. This study presented a deep convolutional neural network (CNN) “MultiScale-Net” for AFD in Landsat-8 datasets at the pixel level. The proposed network had two main characteristics: (1) several convolution kernels with multiple sizes, and (2) dilated convolution layers (DCLs) with various dilation rates. Moreover, this paper suggested an innovative Active Fire Index (AFI) for AFD. AFI was added to the network inputs consisting of the SWIR2, SWIR1, and Blue bands to improve the performance of the MultiScale-Net. In an ablation analysis, three different scenarios were designed for multi-size kernels, dilation rates, and input variables individually, resulting in 27 distinct models. The quantitative results indicated that the model with AFI-SWIR2-SWIR1-Blue as the input variables, using multiple kernels of sizes 3 × 3, 5 × 5, and 7 × 7 simultaneously, and a dilation rate of 2, achieved the highest F1-score and IoU of 91.62% and 84.54%, respectively. Stacking AFI with the three Landsat-8 bands led to fewer false negative (FN) pixels. Furthermore, our qualitative assessment revealed that these models could detect single fire pixels detached from the large fire zones by taking advantage of multi-size kernels. Overall, the MultiScale-Net met expectations in detecting fires of varying sizes and shapes over challenging test samples.

Supervised Machine Learning Approaches on Multispectral Remote Sensing Data for a Combined Detection of Fire and Burned Area

Bushfires pose a severe risk, among others, to humans, wildlife, and infrastructures. Rapid detection of fires is crucial for fire-extinguishing activities and rescue missions. Besides, mapping burned areas also supports evacuation and accessibility to emergency facilities. In this study, we propose a generic approach for detecting fires and burned areas based on machine learning (ML) approaches and remote sensing data. While most studies investigated either the detection of fires or mapping burned areas, we addressed and evaluated, in particular, the combined detection on three selected case study regions. Multispectral Sentinel-2 images represent the input data for the supervised ML models. First, we generated the reference data for the three target classes, burned, unburned, and fire, since no reference data were available. Second, the three regional fire datasets were preprocessed and divided into training, validation, and test subsets according to a defined schema. Furthermore, an undersampling approach ensured the balancing of the datasets. Third, seven selected supervised classification approaches were used and evaluated, including tree-based models, a self-organizing map, an artificial neural network, and a one-dimensional convolutional neural network (1D-CNN). All selected ML approaches achieved satisfying classification results. Moreover, they performed a highly accurate fire detection, while separating burned and unburned areas was slightly more challenging. The 1D-CNN and extremely randomized tree were the best-performing models with an overall accuracy score of 98 % on the test subsets. Even on an unknown test dataset, the 1D-CNN achieved high classification accuracies. This generalization is even more valuable for any use-case scenario, including the organization of fire-fighting activities or civil protection. The proposed combined detection could be extended and enhanced with crowdsourced data in further studies.

A Review on Early Forest Fire Detection Systems Using Optical Remote Sensing

The environmental challenges the world faces nowadays have never been greater or more complex. Global areas covered by forests and urban woodlands are threatened by natural disasters that have increased dramatically during the last decades, in terms of both frequency and magnitude. Large-scale forest fires are one of the most harmful natural hazards affecting climate change and life around the world. Thus, to minimize their impacts on people and nature, the adoption of well-planned and closely coordinated effective prevention, early warning, and response approaches are necessary. This paper presents an overview of the optical remote sensing technologies used in early fire warning systems and provides an extensive survey on both flame and smoke detection algorithms employed by each technology. Three types of systems are identified, namely terrestrial, airborne, and spaceborne-based systems, while various models aiming to detect fire occurrences with high accuracy in challenging environments are studied. Finally, the strengths and weaknesses of fire detection systems based on optical remote sensing are discussed aiming to contribute to future research projects for the development of early warning fire systems.

Analysis of Multifractal and Organization/Order Structure in Suomi-NPP VIIRS Normalized Difference Vegetation Index Series of Wildfire Affected and Unaffected Sites by Using the Multifractal Detrended Fluctuation Analysis and the Fisher-Shannon Analysis

The analysis of vegetation dynamics affected by wildfires contributes to the understanding of ecological changes under disturbances. The use of the Normalized Difference Vegetation Index (NDVI) of satellite time series can effectively contribute to this investigation. In this paper, we employed the methods of multifractal detrended fluctuation analysis (MFDFA) and Fisher–Shannon (FS) analysis to investigate the NDVI series acquired from the Visible Infrared Imaging Radiometer Suite (VIIRS) of the Suomi National Polar-Orbiting Partnership (Suomi-NPP). Four study sites that were covered by two different types of vegetation were analyzed, among them two sites were affected by a wildfire (the Camp Fire, 2018). Our findings reveal that the wildfire increases the heterogeneity of the NDVI time series along with their organization structure. Furthermore, the fire-affected and fire-unaffected pixels are quite well separated through the range of the generalized Hurst exponents and the FS information plane. The analysis could provide deeper insights on the temporal dynamics of vegetation that are induced by wildfire.

SmokeNet: Satellite Smoke Scene Detection Using Convolutional Neural Network with Spatial and Channel-Wise Attention

A variety of environmental analysis applications have been advanced by the use of satellite remote sensing. Smoke detection based on satellite imagery is imperative for wildfire detection and monitoring. However, the commonly used smoke detection methods mainly focus on smoke discrimination from a few specific classes, which reduces their applicability in different regions of various classes. To this end, in this paper, we present a new large-scale satellite imagery smoke detection benchmark based on Moderate Resolution Imaging Spectroradiometer (MODIS) data, namely USTC_SmokeRS, consisting of 6225 satellite images from six classes (i.e., cloud, dust, haze, land, seaside, and smoke) and covering various areas/regions over the world. To build a baseline for smoke detection in satellite imagery, we evaluate several state-of-the-art deep learning-based image classification models. Moreover, we propose a new convolution neural network (CNN) model, SmokeNet, which incorporates spatial and channel-wise attention in CNN to enhance feature representation for scene classification. The experimental results of our method using different proportions (16%, 32%, 48%, and 64%) of training images reveal that our model outperforms other approaches with higher accuracy and Kappa coefficient. Specifically, the proposed SmokeNet model trained with 64% training images achieves the best accuracy of 92.75% and Kappa coefficient of 0.9130. The model trained with 16% training images can also improve the classification accuracy and Kappa coefficient by at least 4.99% and 0.06, respectively, over the state-of-the-art models.

Relative Azimuthal-Angle Matching (RAM): A Screening Method for GEO-LEO Reflectance Comparison in Middle Latitude Forests

This study introduced a data screening method for comparing the reflectances in middle latitude forest regions collected by a Geostationary Earth Observing (GEO) satellite and a Low Earth Orbit (LEO) satellite. This method attempts to reduce the differences between the relative azimuth angles of the GEO and LEO observations. The method, called relative azimuthal-angle matching (RAM), takes advantage of the high temporal resolution of the GEO satellites, which enables collection of a wide range of relative azimuth angles within a day. The performance of the RAM method was evaluated using data in the visible and near-infrared bands collected by the Himawari-8/Advanced Himawari Imager (AHI) and the Terra/Moderate Resolution Imaging Spectroradiometer (MODIS). The consistency of the reflectance pairs of MODIS and AHI selected by the RAM method was compared with the consistency of the reflectance pairs acquired simultaneously by the two sensors. The data were matched pixel-by-pixel after applying atmospheric corrections and cloud screening. The results show that RAM improved the reflectance ratio by approximately 10% for the red and NIR bands on average relative to the simultaneous observations. Significant improvements in the two bands were observed (20%), especially among data collected in the fall and winter. Performance of RAM depends largely on season. Especially in summer, the reflectance pair chosen by RAM showed less consistency than solar zenith-angle matching (SZM). The results also indicated the relatively large influence of the spectral response functions on the green and red bands of the two sensors.

Detection and Monitoring of Forest Fires Using Himawari-8 Geostationary Satellite Data in South Korea

Geostationary satellite remote sensing systems are a useful tool for forest fire detection and monitoring because of their high temporal resolution over large areas. In this study, we propose a combined 3-step forest fire detection algorithm (i.e., thresholding, machine learning-based modeling, and post processing) using Himawari-8 geostationary satellite data over South Korea. This threshold-based algorithm filtered the forest fire candidate pixels using adaptive threshold values considering the diurnal cycle and seasonality of forest fires while allowing a high rate of false alarms. The random forest (RF) machine learning model then effectively removed the false alarms from the results of the threshold-based algorithm (overall accuracy ~99.16%, probability of detection (POD) ~93.08%, probability of false detection (POFD) ~0.07%, and 96% reduction of the false alarmed pixels for validation), and the remaining false alarms were removed through post-processing using the forest map. The proposed algorithm was compared to the two existing methods. The proposed algorithm (POD ~ 93%) successfully detected most forest fires, while the others missed many small-scale forest fires (POD ~ 50–60%). More than half of the detected forest fires were detected within 10 min, which is a promising result when the operational real-time monitoring of forest fires using more advanced geostationary satellite sensor data (i.e., with higher spatial and temporal resolutions) is used for rapid response and management of forest fires.