No abstract

Nowadays Earth observation satellites, in particular Landsat, provide a valuable help to forest managers in post-fire operations; being the base of post-fire damage maps that enable to analyze fire impacts and to develop vegetation recovery plans. Sentinel-2A MultiSpectral Instrument (MSI) records data in similar spectral wavelengths that Landsat 8 Operational Land Imager (OLI), and has higher spatial and temporal resolutions. This work compares two types of satellite-based maps for evaluating fire damage in a large wildfire (around 8000 ha) located in Sierra de Gata (central-western Spain) on 6-11 August 2015. 1) burn severity maps based exclusively on Landsat data; specifically, on differenced Normalized Burn Ratio (dNBR) and on its relative versions (Relative dNBR, RdNBR, and Relativized Burn Ratio, RBR) and 2) burn severity maps based on the same indexes but combining pre-fire data from Landsat 8 OLI with post-fire data from Sentinel-2A MSI data. Combination of both Landsat and Sentinel-2 data might reduce the time elapsed since forest fire to the availability of an initial fire damage map. Interpretation of ortho-photograph Pléiades 1 B data (1:10,000) provided us the ground reference data to measure the accuracy of both burn severity maps. Results showed that Landsat based burn severity maps presented an adequate assessment of the damage grade (κ statistic = 0.80) and its spatial distribution in wildfire emergency response. Further using both Landsat and Sentinel-2 MSI data the accuracy of burn severity maps, though slightly lower (κ statistic = 0.70) showed an adequate level for be used by forest managers.

No abstract

No abstract

This study shows an accurate and fast methodology in order to evaluate fire severity classes of large forest fires. A single Landsat Enhanced Thematic Mapper multispectral image was utilized in this study with the aim of mapping fire severity classes (high, moderate and low) using a combined-approach based in an spectral mixing model and object-based image analysis. A large wildfire in the Northwest of Spain is used to test the model. Fraction images obtained by Landsat unmixing were used as input data in the object-based image analysis. A multilevel segmentation and a classification were carried out by using membership functions. This method was compared with other simplest ones in order to evaluate the suitability to distinguish between the three fire severity classes above mentioned. McNemar’s test was used to evaluate the statistical significance of the difference between approaches tested in this study. The combined approach achieved the highest accuracy reaching 97.32% and kappa index of agreement of 95.96% and improving accuracy of individual classes.