A Burned Area Mapping Algorithm for Sentinel-2 Data Based on Approximate Reasoning and Region Growing

Madagascar's burned area from Sentinel-2 imagery (2016-2022): Four times higher than from lower resolution sensors

Madagascar is one of the most burned regions in the world, to the point that it has been called the 'Isle of fire' or the 'Burning Island'. An accurate characterization of the burned area (BA) is crucial for understanding the true situation and impacts of fires on this island, where there is an active scientific debate on how fire affects multiple environmental and socioeconomic aspects, and how fire regimes should be in a complex context with differing interests. Despite this, recent advances have revealed that BA in Madagascar is poorly characterised by the currently available global BA products. In this work, we present, validate, and explore a BA database at 20 m spatial resolution for Madagascar covering the period 2016-2022. The database was built based on 75,010 Sentinel-2 images using a two-phase BA detection algorithm. The validation with independent long-term reference units showed Dice coefficients ≥79 %, omission errors ≤24 %, commission errors ≤18 %, and a relative bias ≥ - 8 %. An intercomparison with other available global BA products (GABAM, FireCCI51, C3SBA11, or MCD64) demonstrated that our product (i) exhibits temporal consistency, (ii) represents a significant accuracy improvement, as it reduces BA underestimations by about eightfold, (iii) yields BA estimates four times higher, and (iv) shows enhanced capability in detecting fires of all sizes. The observed BA spatial patterns were heterogeneous across the island, with 32 % of the grasslands burning annually, in contrast to other land cover types such as the dense tropical forest where <2 % burned every year. We conclude that the BA characterization in Madagascar must be addressed using imagery at spatial resolution higher than MODIS or Sentinel-3 (≥250 m), and temporal resolution higher than Landsat (16 days) to deal with cloudiness, the rapid attenuation of burn scars signals, and small fire patches.

Burned Area Detection Using Multi-Sensor SAR, Optical, and Thermal Data in Mediterranean Pine Forest

Burned area (BA) mapping of a forest after a fire is required for its management and the determination of the impacts on ecosystems. Different remote sensing sensors and their combinations have been used due to their individual limitations for accurate BA mapping. This study analyzes the contribution of different features derived from optical, thermal, and Synthetic Aperture Radar (SAR) images to extract BA information from the Turkish red pine (Pinus brutia Ten.) forest in a Mediterranean ecosystem. In addition to reflectance values of the optical images, Normalized Burn Ratio (NBR) and Land Surface Temperature (LST) data are produced from both Sentinel-2 and Landsat-8 data. The backscatter of C-band Sentinel-1 and L-band ALOS-2 SAR images and the coherence feature derived from the Interferometric SAR technique were also used. The pixel-based random forest image classification method is applied to classify the BA detection in 24 scenarios created using these features. The results show that the L-band data provided a better contribution than C-band data and the combination of features created from Landsat LST, NBR, and coherence of L-band ALOS-2 achieved the highest accuracy, with an overall accuracy of 96% and a Kappa coefficient of 92.62%.

A Fully Automatic, Interpretable and Adaptive Machine Learning Approach to Map Burned Area from Remote Sensing

The paper proposes a fully automatic algorithm approach to map burned areas from remote sensing characterized by human interpretable mapping criteria and explainable results. This approach is partially knowledge-driven and partially data-driven. It exploits active fire points to train the fusion function of factors deemed influential in determining the evidence of burned conditions from reflectance values of multispectral Sentinel-2 (S2) data. The fusion function is used to compute a map of seeds (burned pixels) that are adaptively expanded by applying a Region Growing (RG) algorithm to generate the final burned area map. The fusion function is an Ordered Weighted Averaging (OWA) operator, learnt through the application of a machine learning (ML) algorithm from a set of highly reliable fire points. Its semantics are characterized by two measures, the degrees of pessimism/optimism and democracy/monarchy. The former allows the prediction of the results of the fusion as affected by more false positives (commission errors) than false negatives (omission errors) in the case of pessimism, or vice versa; the latter foresees if there are only a few highly influential factors or many low influential ones that determine the result. The prediction on the degree of pessimism/optimism allows the expansion of the seeds to be appropriately tuned by selecting the most suited growing layer for the RG algorithm thus adapting the algorithm to the context. The paper illustrates the application of the automatic method in four study areas in southern Europe to map burned areas for the 2017 fire season. Thematic accuracy at each site was assessed by comparison to reference perimeters to prove the adaptability of the approach to the context; estimated average accuracy metrics are omission error = 0.057, commission error = 0.068, Dice coefficient = 0.94 and relative bias = 0.0046.