Burn Scar Mapping Research Articles

Forest Fire Burn Scar Mapping Based on Modified Image Super-Resolution Reconstruction via Sparse Representation

Forest Fire Burn Scar Mapping Based on Modified Image Super-Resolution Reconstruction via Sparse Representation

Regional-Scale Assessment of Burn Scar Mapping in Southwestern Amazonia Using Burned Area Products and CBERS/WFI Data Cubes

Fires are one of the main sources of disturbance in fire-sensitive ecosystems such as the Amazon. Any attempt to characterize their impacts and establish actions aimed at combating these events presupposes the correct identification of the affected areas. However, accurate mapping of burned areas in humid tropical forest regions remains a challenging task. In this paper, we evaluate the performance of four operational BA products (MCD64A1, Fire_cci, GABAM and MapBiomas Fogo) on a regional scale in the southwestern Amazon and propose a new approach to BA mapping using fraction images extracted from data cubes of the Brazilian orbital sensors CBERS-4/WFI and CBERS-4A/WFI. The methodology for detecting burned areas consisted of applying the Linear Spectral Mixture Model to the images from the CBERS-4/WFI and CBERS-4A/WFI data cubes to generate shadow fraction images, which were then segmented and classified using the ISOSEG non-supervised algorithm. Regression and similarity analyses based on regular grid cells were carried out to compare the BA mappings. The results showed large discrepancies between the mappings in terms of total area burned, land use and land cover affected (forest and non-forest) and spatial location of the burned area. The global products MCD64A1, GABAM and Fire_cci tended to underestimate the area burned in the region, with Fire_cci underestimating BA by 88%, while the regional product MapBiomas Fogo was the closest to the reference, underestimating by only 7%. The burned area estimated by the method proposed in this work (337.5 km2) was 12% higher than the reference and showed a small difference in relation to the MapBiomas Fogo product (18% more BA). These differences can be explained by the different datasets and methods used to detect burned areas. The adoption of global products in regional studies can be critical in underestimating the total area burned in sensitive regions. Our study highlights the need to develop approaches aimed at improving the accuracy of current global products, and the development of regional burned area products may be more suitable for this purpose. Our proposed approach based on WFI data cubes has shown high potential for generating more accurate regional burned area maps, which can refine BA estimates in the Amazon.

Full-polarimetric burn scar mapping – the differences of active fire and post-fire situations

ABSTRACTFire is a threat to human lives, infrastructure, and forestry. Satellite-based Earth observations enable a fast, efficient, and reliable estimation of burnt area. In most cases, optical satellite data are used for burn scar detection. However, smoke and cloud coverage strongly limits the suitability of optical imagery for rapid mapping of burn scars. Here, an automated procedure based on quad-polarized L-band Synthetic Aperture Radar (SAR) data is presented to enable fast burn scar mapping independently of the weather and smoke conditions. Full-polarimetric SAR data enable the decomposition of the SAR backscatter into different scattering mechanisms, describing the scatterer more precisely. Important differences in the polarimetric backscattering behaviour during the active fire and post-fire situations are reported. While the SPAN difference is best suited for burn scar mapping during active fire situations, the Entropy-Anisotropy-Alpha and the generalized Freeman–Durden decompositions showed the best suitability for burnt area mapping several months after the end of the fire. The proposed methodology is an object-based image analysis approach based on change detection. The big fire event which affected Fort McMurray, Alberta, Canada, in May–June 2016 was investigated as a case study.

Burned Area Mapping of an Escaped Fire into Tropical Dry Forest in Western Madagascar Using Multi-Season Landsat OLI Data

A human-induced fire cleared a large area of tropical dry forest near the Ankoatsifaka Research Station at Kirindy Mitea National Park in western Madagascar over several weeks in 2013. Fire is a major factor in the disturbance and loss of global tropical dry forests, yet remotely sensed mapping studies of fire-impacted tropical dry forests lag behind fire research of other forest types. Methods used to map burns in temperature forests may not perform as well in tropical dry forests where it can be difficult to distinguish between multiple-age burn scars and between bare soil and burns. In this study, the extent of forest lost to stand-replacing fire in Kirindy Mitea National Park was quantified using both spectral and textural information derived from multi-date satellite imagery. The total area of the burn was 18,034 ha. It is estimated that 6% (4761 ha) of the Park’s primary tropical dry forest burned over the period 23 June to 27 September. Half of the forest burned (2333 ha) was lost in the large conflagration adjacent to the Research Station. The best model for burn scar mapping in this highly-seasonal tropical forest and pastoral landscape included the differenced Normalized Burn Ratio (dNBR) and both uni- and multi-temporal measures of greenness. Lessons for burn mapping of tropical dry forest are much the same as for tropical dry forest mapping—highly seasonal vegetation combined with a mixture of background spectral information make for a complicated analysis and may require multi-temporal imagery and site specific techniques.

A Fully Automatic Burnt Area Mapping Processor Based on AVHRR Imagery—A TIMELINE Thematic Processor

The German Aerospace Center’s (DLR) TIMELINE project (“Time Series Processing of Medium Resolution Earth Observation Data Assessing Long-Term Dynamics in our Natural Environment”) aims to develop an operational processing and data management environment to process 30 years of National Oceanic and Atmospheric Administration (NOAA)—Advanced Very High-Resolution Radiometer (AVHRR) raw data into Level (L) 1b, L2, and L3 products. This article presents the current status of the fully automated L3 burnt area mapping processor, which is based on multi-temporal datasets. The advantages of the proposed approach are (I) the combined use of different indices to improve the classification result, (II) the provision of a fully automated processor, (III) the generation and usage of an up-to-date cloud-free pre-fire dataset, (IV) classification with adaptive thresholding, and (V) the assignment of five different probability levels to the burnt areas detected. The results of the AVHRR data-based burn scar mapping processor were validated with the Moderate Resolution Imaging Spectroradiometer (MODIS) burnt area product MCD64 at four different European study sites. In addition, the accuracy of the AVHRR-based classification and that of the MCD64 itself were assessed by means of Landsat imagery.

A Simple Normalized Difference Approach to Burnt Area Mapping Using Multi-Polarisation C-Band SAR

In fire-prone ecosystems, periodic fires are vital for ecosystem functioning. Fire managers seek to promote the optimal fire regime by managing fire season and frequency requiring detailed information on the extent and date of previous burns. This paper investigates a Normalised Difference α-Angle (NDαI) approach to burn-scar mapping using C-band data. Polarimetric decompositions are used to derive α-angles from pre-burn and post-burn scenes and NDαI is calculated to identify decreases in vegetation between the scenes. The technique was tested in an area affected by a wildfire in January 2016 in the Western Cape, South Africa. The quad-pol H-A-α decomposition was applied to RADARSAT-2 data and the dual-pol H-α decomposition was applied to Sentinel-1A data. The NDαI results were compared to a burn scar extracted from Sentinel-2A data. High overall accuracies of 97.4% (Kappa = 0.72) and 94.8% (Kappa = 0.57) were obtained for RADARSAT-2 and Sentinel-1A, respectively. However, large omission errors were found and correlated strongly with areas of high local incidence angle for both datasets. The combined use of data from different orbits will likely reduce these errors. Furthermore, commission errors were observed, most notably on Sentinel-1A results. These errors may be due to the inability of the dual-pol H-α decomposition to effectively distinguish between scattering mechanisms. Despite these errors, the results revealed that burnt areas could be extracted and were in good agreement with the results from Sentinel-2A. Therefore, the approach can be considered in areas where persistent cloud cover or smoke prevents the extraction of burnt area information using conventional multispectral approaches.

Land cover mapping with emphasis to burnt area delineation using co-orbital ALI and Landsat TM imagery

In this study, the potential of EO-1 Advanced Land Imager (ALI) radiometer for land cover and especially burnt area mapping from a single image analysis is investigated. Co-orbital imagery from the Landsat Thematic Mapper (TM) was also utilised for comparison purposes. Both images were acquired shortly after the suppression of a fire occurred during the summer of 2009 North-East of Athens, the capital of Greece. The Maximum Likelihood (ML), Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs) classifiers were parameterised and subsequently applied to the acquired satellite datasets. Evaluation of the land use/cover mapping accuracy was based on the error matrix statistics. Also, the McNemar test was used to evaluate the statistical significance of the differences between the approaches tested. Derived burnt area estimates were validated against the operationally deployed Services and Applications For Emergency Response (SAFER) Burnt Scar Mapping service.All classifiers applied to either ALI or TM imagery proved flexible enough to map land cover and also to extract the burnt area from other land surface types. The highest total classification accuracy and burnt area detection capability was returned from the application of SVMs to ALI data. This was due to the SVMs ability to identify an optimal separating hyperplane for best classes’ separation that was able to better utilise ALI's advanced technological characteristics in comparison to those of TM sensor. This study is to our knowledge the first of its kind, effectively demonstrating the benefits of the combined application of SVMs to ALI data further implying that ALI technology may prove highly valuable in mapping burnt areas and land use/cover if it is incorporated into the development of Landsat 8 mission, planned to be launched in the coming years.

Towards a pan-European burnt scar mapping methodology based on single date medium resolution optical remote sensing data

A two stage approach for burnt scar detection from single date multispectral medium spatial resolution optical remote sensing data (AWIFS) has been developed. The approach includes first an identification of burnt scar seeds based on a learning algorithm followed by a region growing process. An Artificial Neural Network (ANN) and a Classification tree (CT) were tested as learning algorithms. Both learning algorithms were coupled with a bootstrap aggregation. Training data for the classifiers were obtained from MODIS-based polygons generated by the Rapid Damage Assessment (RDA) module of the European Forest Fire Information System (EFFIS), to which different levels of filtering were applied. The outcomes were validated against datasets generated from visual interpretation of ETM+ scenes. The method was tested in two locations of Portugal and Greece. Both ANN and CT alternatives produced similar results, with kappa coefficients close to 0.80 in the Greek location and higher than 0.70 in the Portuguese location. In test sites, more than 80% and 90% of burnt areas larger than 10ha and 50ha respectively were detected. The results show that filtering the training dataset reduces the overestimation of burnt areas and produce higher accuracies.

Burnt area delineation from a uni-temporal perspective based on Landsat TM imagery classification using Support Vector Machines

Information on burnt area is of critical importance in many applications as for example in assessing the disturbance of natural ecosystems due to a fire or in proving important information to policy makers on the land cover changes for establishing restoration policies of fire-affected regions. Such information is commonly obtained through remote sensing image thematic classification and a wide range of classifiers have been suggested for this purpose. The objective of the present study has been to investigate the use of Support Vector Machines (SVMs) classifier combined with multispectral Landsat TM image for obtaining burnt area mapping. As a case study a typical Mediterranean landscape in Greece was used, in which occurred one of the most devastating fires during the summer of 2007. Accuracy assessment was based on the classification overall statistical accuracy results and also on comparisons of the derived burnt area estimates versus validated estimates from the Risk-EOS Burnt Scar Mapping service. Results from the implementation of the SVM using diverse kernel functions showed an average overall classification accuracy of 95.87% and a mean kappa coefficient of 0.948, with the burnt area class always clearly separable from all the other classes used in the classification scheme. Total burnt area estimate computed from the SVM was also in close agreement with that from Risk-EOS (mean difference of less than 1%). Analysis also indicated that, at least for the studied here fire, the inclusion of the two middle infrared spectral bands TM5 and TM7 of TM sensor as well as the selection of the kernel function in SVM implementation have a negligible effect in both the overall classification performance and in the delineation of total burnt area. Overall, results exemplified the appropriateness of the spatial and spectral resolution of the Landsat TM imagery combined with the SVM in obtaining rapid and cost-effective post-fire analysis. This is of considerable scientific and practical value, given the present open access to the archived and new observations from this satellite radiometer globally.

A comparative analysis of a fixed thresholding vs. a classification tree approach for operational burn scar detection and mapping

The scope of this paper is to demonstrate, evaluate and compare two burn scar mapping (BSM) approaches developed and applied operationally in the framework of the RISK-EOS service element project within the Global Monitoring for Environment and Security (GMES) program funded by ESA (http://www.risk-eos.com). The first method is the BSM_NOA, a fixed thresholding method using a set of specifically designed and combined image enhancements, whilst the second one is the BSM_ITF, a decision tree classification approach based on a wide range of biophysical parameters. The two methods were deployed and compared in the framework of operational mapping conditions set by RISK-EOS standards, based either on sets of uni- or multi-temporal satellite images acquired by Landsat 5 TM and SPOT 4 HRV. The evaluation of the performance of the two methods showed that either in uni- or multi-temporal acquisition mode, the two methods reach high detection capability rates ranging from 80% to 91%. At the same time, the minimum burnt area detected was of 0.9–1.0ha, despite the coarser spatial resolution of Landsat 5 TM sensor. Among the advantages of the satellite-based approaches compared to conventional burn scar mapping, are cost-efficiency, repeatability, flexibility, and high spatial and thematic accuracy from local to country level. Following the catastrophic fire season of 2007, burn scar maps were generated using BSM_NOA for the entirety of Greece and BSM_ITF for south France in the framework of the RISK-EOS/GMES Services Element project.

An automatic method for burn scar mapping using support vector machines

Wildfires release large amounts of carbon, smoke and aerosols that strongly impact the global climatic system. Burn scar is an important parameter when modelling the impact of wildfires on the ecosystem and the climatic system. We have developed an automatic burn scar mapping method using daily Moderate Resolution Imaging Spectroradiometer (MODIS) data, in which the Global Environment Monitoring Index (GEMI), a vegetation index VI3T and a new index, GEMI‐Burn scar (GEMI‐B), were used together to enhance the differences between burned and unburned pixels related to vegetation photosynthesis, surface temperature and vegetation water content, respectively, and an automatic region growing method based on Support Vector Machines (SVMs) was used to classify burn scars without any predefined threshold. A case study was carried out to validate the new method at the border area between Mongolia and China, where a wildfire took place in May 2003. The results show that the burn scar area extracted by the new method is consistent with that from Landsat Thematic Mapper (TM) data with high accuracy. The sound performance of the new technique is due to the following reasons: (1) multiple features of burn scar spectra were combined and used, (2) a reasonable assumption was made stating that the neighbourhoods of active fires (hotspots) are most likely to be burn scars, (3) an SVM classifier was adopted that works well with a small number of training samples, and (4) an iterative classification procedure was developed that is capable of running continuous training for the SVM classifier to deal with the transitionary features of burn scar pixels. The results suggest that the new index GEMI‐B and automatic mapping method based on SVMs have the potential to be applied to near real‐time burn scar mapping in grassland areas.

Quantifying the impact of cloud obscuration on remote sensing of active fires in the Brazilian Amazon

Vegetation fires remain as one of the most important processes governing land use and land cover change in tropical areas. The large area extent of fire prone areas associated with human activities makes satellite remote sensing of active fires a valuable tool to help monitor biomass burning in those regions. However, identification of active fire fronts under optically thick clouds is not possible through passive remote sensing, often resulting in omission errors. Previous analyses of fire activity either ignored the cloud obscuration problem or applied corrections based on the assumption that fire occurrence is not impacted by the presence of clouds. In this study we addressed the cloud obscuration problem in the Brazilian Amazon region using a pixel based probabilistic approach, using information on previous fire occurrence, precipitation and land use. We implemented the methodology using data from the geostationary GOES imager, covering the entire diurnal cycle of fire activity and cloud occurrence. Our assessment of the method indicated that the cloud adjustment reproduced the number of potential fires missed within 1.5% and 5% of the true fire counts on annual and monthly bases respectively. Spatially explicit comparison with high resolution burn scar maps in Acre state showed a reduction of omission error (from 58.3% to 43.7%) and only slight increase of commission error (from 6.4% to 8.8%) compared to uncorrected fire counts. A basin-wide analysis of corrected GOES fire counts during 2005 showed a mean cloud adjustment factor of approximately 11%, ranging from negligible adjustment in the central and western part of the Brazilian Amazon to as high as 50% in parts of Roraima, Para and Mato Grosso.

AVHRR-based mapping of fires in Russia: New products for fire management and carbon cycle studies

A new database of fire activity in Russia derived from 1-km resolution remote sensing imagery is presented and discussed. The procedure used to generate this burned-area product is described, including active-fire detection and burn-scar mapping approaches. Fire detection makes use of a probabilistic procedure using image data from the United States National Oceanic and Atmospheric Administration's (NOAA) advanced very high resolution radiometer (AVHRR) system. Using the combination of AVHRR data collected at the Krasnoyarsk, Russia, high-resolution picture transmission (HRPT) receiving station, and data from the NOAA Satellite Active Archive (SAA), fire maps are being created for all of Russia for 1995 to 1997 and all of Eastern Russia (east of the Ural Mountains) for 1995 to 2002. This mapping effort has resulted in the most complete set of historic fire maps available for Russia. An initial validation indicates that the burned-area estimates are conservative because the approaches do not detect smaller fires, and, in many cases, fire areas are slightly underestimated. Analyses using the fire database showed that an average of 7.7×10 6 ha yr −1 of fire occurred in Eastern Russia between 1996 and 2002 and that fire was widely dispersed in different regions. The satellite-based burned-area estimates area were two to five times greater than those contained in official government burned-area statistics. The data show that there is significant interannual variability in area burned, ranging between a low of 1.5×10 6 ha in 1997 to a high of 12.1×10 6 ha in 2002. Seasonal patterns of fire are similar to patterns seen in the North American boreal region, with large-fire seasons experiencing more late-season burning (in August and September) than during low-fire years. There was a distinct zonal distribution of fires in Russia; 65% of the area burned occurred in the taiga zone, which includes southern, middle, and northern taiga subzones, 20% in the steppe and forest steppe zones, 12% in the mixed forest zone, and 3% in the tundra and forest-tundra zones. Lands classified as forest experienced 55% of all burned area, while crops and pastures, swamps and bogs, and grass and shrubs land cover categories experienced 13% to 15% each. Finally, the utility of the products is discussed in the context of fire management and carbon cycling.

A dynamic algorithm for wildfire mapping with NOAA/AVHRR data

A wildfire-mapping algorithm is proposed based on fire dynamics, called the dynamic algorithm. It is applied to daily NOAA/AVHRR/HRPT data for wildland areas (scrub, chaparral, grassland, marsh, riparian forest, woodland, rangeland and forests) in California for September and October 1999. Daily AVHRR images acquired from two successive days are compared for active fire detection and burn scar mapping. The algorithm consists of four stages: data preparation; hotspot detection; burn scar mapping; and final confirmation of potential burn scar pixels. Preliminary comparisons between the result mapped by the dynamic algorithm and the fire polygons collected by the California Department of Forestry and Fire Protection through ground survey indicate that the algorithm can track burn scars at different developmental stages at a daily level. The comparisons between wildfire mapping results produced by a modified version of an existing algorithm and the dynamic algorithm also indicate this point. This is the major contribution of this algorithm to wildfire detection methods. The dynamic algorithm requires highly precise registration between consecutive images.

Areal estimates of fragmented land cover: Effects of pixel size and model-based corrections

The effects of coarse spatial resolution on the apparent extent of fragmented cover types on digital maps were assessed and a model-based method for correcting for underestimation due to omission of small fragments was tested. Burn scar maps based on Landsat Multispectral Scanner (MSS) imagery and a map of ponds in Arctic tundra based on ERS-1 synthetic aperture radar (SAR) imagery were spatially degraded to simulate EOS Moderate-Resolution Imaging Spectroradiometer (MODIS) and 100 m SAR maps. Omission of small fragments on the degraded imagery decreased the apparent extent by as much as 50%. The apparent extent on degraded maps, where various degrees of systematic tendency towards commission were simulated by identification of pixels with less than 50% burn scar or open water, was either close to that measured on the original map (for small classification bias) or overestimated (large classification bias). An area correction procedure based on fitting sizes to a lognormal distribution with censored data was tested. The area correction compensated for 25 to 70% of the underestimation relative to the original maps. Area correction led to inflated areal estimates with biased maps or with a Pareto rather than the lognormal distribution model, indicating that the areal correction procedure must be used judiciously if improved estimates are to be expected.