First evaluation of fire severity retrieval from PRISMA hyperspectral data

We evaluated the utility of the composite burn index (CBI) for estimating fire severity in Alaskan black spruce forests by comparing data from 81 plots located in 2004 and 2005 fire events. We collected data to estimate the CBI and quantify crown damage, percent of trees standing after the fire, depth of the organic layer remaining after the fire, depth of burning in the surface organic layer (absolute and relative), and the substrate layer exposed by the fire. To estimate pre-fire organic layer depth, we collected data in 15 unburned stands to develop relationships between total organic layer depth and measures of the adventitious root depth above mineral soil and below the surface of the organic layer. We validated this algorithm using data collected in 17 burned stands where pre-fire organic layer depth had been measured. The average total CBI value in the black spruce stands was 2.46, with most of the variation a result of differences in the CBI observed for the substrate layer. While a quadratic equation using the substrate component of CBI was a relatively strong predictor of mineral soil exposure as a result of fire (R2 = 0.61, P < 0.0001, F = 60.3), low correlations were found between the other measures of fire severity and the CBI (R2 = 0.00–0.37). These results indicate that the CBI approach has limited potential for quantifying fire severity in these ecosystems, in particular organic layer consumption, which is an important factor to understand how ecosystems will respond to changing climate and fire regimes in northern regions.

Background Fire behaviour assessments of past wildfire events have major implications for anticipating post-fire ecosystem responses and fuel treatments to mitigate extreme fire behaviour of subsequent wildfires. Aims This study evaluates for the first time the potential of remote sensing techniques to provide explicit estimates of fire type (surface fire, intermittent crown fire, and continuous crown fire) in Mediterranean ecosystems. Methods Random Forest classification was used to assess the capability of spectral indices and multiple endmember spectral mixture analysis (MESMA) image fractions (char, photosynthetic vegetation, non-photosynthetic vegetation) retrieved from Sentinel-2 data to predict fire type across four large wildfires Key results MESMA fraction images procured more accurate fire type estimates in broadleaf and conifer forests than spectral indices, without remarkable confusion among fire types. High crown fire likelihood in conifer and broadleaf forests was linked to a post-fire MESMA char fractional cover of about 0.8, providing a direct physical interpretation. Conclusions Intrinsic biophysical characteristics such as the fractional cover of char retrieved from sub-pixel techniques with physical basis are accurate to assess fire type given the direct physical interpretation. Implications MESMA may be leveraged by land managers to determine fire type across large areas, but further validation with field data is advised.

The boreal region stores a large proportion of the world's terrestrial carbon (C) and is subject to high-intensity, stand-replacing wildfires that release C and nitrogen (N) stored in biomass and soils through combustion. While severity and extent of fires drives overall emissions, methods for accurately estimating fire severity are poorly tested in this unique region where organic soil combustion is responsible for a large proportion of total emissions. We tested a method using adventitious roots on black spruce trees (Picea mariana) in combination with canopy allometry to reconstruct prefire organic soil layers and canopy biomass in boreal black spruce forests of Alaska (USA), thus providing a basis for more accurately quantifying fire severity levels. We calibrated this adventitious-root-height method in unburned spruce stands and then tested it by comparing our biomass and soils estimates reconstructed in burned stands with actual prefire stand measurements. We applied this approach to 38 black spruce stands burned in 2004 in Alaska, where we measured organic soil and stand characteristics and estimated the amount of soil and canopy biomass, as well as C and N pools, consumed by fire. These high-intensity quantitative estimates of severity were significantly correlated to a semiquantitative visual rapid assessment tool, the composite burn index (CBI). This index has proved useful for assessing fire severity in forests in the western United States but has not yet been widely tested in the boreal forest. From our study, we conclude that using postfire measurements of adventitious roots on black spruce trees in combination with soils and tree data can be used to reconstruct prefire organic soil depths and biomass pools, providing accurate estimates of fire severity and emissions. Furthermore, using our quantitative reconstruction we show that CBI is a reasonably good predictor of biomass and soil C loss at these sites, and it shows promise for rapidly estimating fire severity across a wide range of boreal black spruce forest types, especially where the use of high-intensity measurements may be limited by cost and time.

Forest fires are key ecological disturbances that influence vegetation dynamics and soil microbial processes central to carbon and nutrient cycling. While fire frequency and severity are increasing globally, the microbial mechanisms underlying ecosystem recovery remain inadequately understood. We used high-throughput amplicon sequencing to evaluate short-term effects of low- and high-severity fires on soil microbial diversity and co-occurrence networks following fire disturbance in a temperate forest. Fire severity had no significant impact on microbial α-diversity, but significantly altered β-diversity. Mantel tests indicated that soil pH and belowground biomass were the primary environmental drivers of bacterial and fungal community turnover under different fire severities. Further, network analyses revealed distinct microbial responses to fire severity: low-severity fire primarily restructured bacterial associations, whereas high-severity fire disrupted both bacterial and fungal networks. These findings suggest that microbial community structure and interactions are differentially sensitive to fire severity, with implications for soil functional resilience and ecosystem restoration strategies in fire-affected forests.

Evaluation of spectral unmixing techniques using MODIS in a structurally complex savanna environment for retrieval of green vegetation, nonphotosynthetic vegetation, and soil fractional cover

The quantitative assessment of forest fire severity is significant for understanding the changes in ecological processes caused by fire disturbances. As a novel spectral index derived from the multi-objective optimization algorithm, the Analytic Burned Area Index (ABAI) was originally designed for mapping burned areas. However, the performance of the ABAI in detecting forest fire severity has not been addressed. To fill this gap, this study utilizes a ground-based dataset of fire severity (the composite burn index, CBI) to validate the effectiveness of the ABAI in detecting fire severity. First, the effectiveness of the ABAI regarding forest fire severity was validated using uni-temporal images from Sentinel-2 and Landsat 8 OLI. Second, fire severity accuracy derived from the ABAI with bi-temporal images from both sensors was evaluated. Finally, the performance of the ABAI was tested with different sensors and compared with representative spectral indices. The results show that (1) the ABAI demonstrates significant advantages in terms of accuracy and stability in assessing fire severity, particularly in areas with large numbers of terrain shadows and severe burn regions; (2) the ABAI also shows great advantages in assessing regional forest fire severity when using only uni-temporal remotely sensed data, and it performed almost as well as the dNBR in bi-temporal images. (3) The ABAI outperforms commonly used indices with both Sentinel-2 and Landsat 8 data, indicating that the ABAI is normally more generalizable and powerful and provides an optional spectral index for fire severity evaluation.

In this paper we evaluated the potential use of Multiple Endmember Spectral Mixture Analysis (MESMA) applied to EO-1 Hyperion hyperspectral data to separate land covers (soil = dunes and dry fields; green vegetation = pinus, eucalyptus and grasslands; water = without sediments, with sediments, and with chlorophyll; and shade), in the southern state of Rio Grande do Sul, Brazil. The approach involved (a) preprocessing and atmospheric correction of Hyperion image; (b) sequential use of Minimum Noise Fraction (MNF), Pixel Purity Index (PPI) and n-Dimensional Visualizer techniques in the visible to shortwave range for the initial selection of a group of endmember, and another group of pixels for model validation; (c) use of the software Visualization and Image Processing for Environmental Research (VIPER) Tools to perform the final selection of endmembers based on the spectral library, and to obtain MESMA models; and (d) evaluation of resulting fraction images and RMSE values to determine the optimal number of endmembers of the MESMA model. Results showed that a four-endmember MESMA model described the diversity of the scene components, including that of materials within the same class (e.g., pinus and eucalyptus) and produced the largest fractions and the lowest RMSE values on a per-pixel basis. Results also showed the performance of MESMA applied to Hyperion data to discriminate properly land covers in the coastal plains, even considering the low signal-to-noise ratio of the instrument.

Brachypodium sylvaticum, a shade-tolerant, forest dwelling, and aggressive invasive grass native to Eurasia, is a noxious weed in California, Oregon, and Washington. This species could cause ecosystem collapse by altering forest fire regimes. To examine interactions with fire, we divided two Willamette National Forest sites into eight units and randomly selected half for treatment with prescribed fire in spring 2011. We assessed the effect ofB. sylvaticumon fire (severity and intensity) as well as the effect of fire onB. sylvaticum(cover, seedling emergence, and dispersal). We found thatB. sylvaticumcover decreased fire severity but had no effect on intensity. Furthermore, fire severity influencedB. sylvaticumcover; in areas receiving low-severity fire, the grass increased from 21 ± 15.05 to 34 ± 15.81%, but in areas of high-severity fire, cover remained consistently around 0% (0 ± 0% cover in yr 1 to 0.2 5± 0.25% in yr 3). In the field, prescribed fire decreased seedling emergence by 32% compared to controls, but not in an associated greenhouse experiment. However, in the greenhouse, severely burned plots had zero emergence, compared to 0.29 ± 0.14 seedlings low-severity m−2plot. Fire severity also influenced dispersal in the field; we monitored plots with < 0.5% coverB. sylvaticuminitially; when these plots experienced low severity fire, they had greaterB. sylvaticumcover (increasing 1,200%), suggesting increased dispersal with less severe fires. High-severity dispersal plots did not experience increased cover. High severity fires have the potential to control the grass, but low-severity fires will likely increase its cover.

Context Obligate-seeding woody plants with long reproductive maturity periods and no soil seed banks are threatened with decline as climate change drives more frequent and severe fires, such as the extensive 2019–2020 wildfires in south-eastern Australia. Aims This study aimed to investigate the effects of fire intervals and severity on the persistence of one such species, Banksia cunninghamii (Hairpin Banksia), in temperate forests. Methods We measured post-fire seedling recruitment of B. cunninghamii at 25 sites in Victoria, burned at various severities in the 2019–2020 wildfires and with differing prior fire intervals. A Bayesian framework was used to model the relationship between seedling numbers, fire severity and fire interval. A spatial analysis compared a species distribution model for B. cunninghamii with fire severity and fire intervals. Key results There was a low chance of B. cunninghamii recruitment (<25%) at sites that either had burned eucalypt canopies or a preceding fire interval of less than 12 years. Sixty-seven percent of its distribution in the south-east of the state of Victoria was mapped as burned at high severity (burned eucalypt canopies) between 1998 and 2020, or burned at shorter than 12 year intervals between 1960 and 2020, although some B. cunninghamii populations will have persisted due to the patchiness of past burns. Conclusions Banksia cunninghamii is vulnerable to local extinctions in the wildfire-affected areas if fires occur again before plants reach maturity, or if high fire severity destroys seeds. Implications More frequent and severe wildfires mean that burn planning needs to consider the reproductive cycles of serotinous obligate-seeding plants.

Vegetation recovery drivers at short-term after fire are plant community-dependent in mediterranean burned landscapes

Synergies between VSWIR and TIR data for the urban environment: An evaluation of the potential for the Hyperspectral Infrared Imager (HyspIRI) Decadal Survey mission

Degradation of cattle pastures is a management concern that influences future land use in Amazonia. However, “degradation” is poorly defined and has different meanings for ranchers, ecologists, and policy makers. Here we analyze pasture degradation using objective scalars of photosynthetic vegetation (PV), nonphotosynthetic vegetation (NPV), and exposed soil (S) derived from Landsat imagery. A general, probabilistic spectral mixture model decomposed satellite spectral reflectance measurements into subpixel estimates of PV, NPV, and S covers at ranches in western and eastern Amazonia. Most pasture management units at all ranches fell along a single line of decreasing PV with increasing NPV and S, which could be considered a degradation continuum. The ranch with the highest stocking densities and most intensive management had greater NPV and S than a less intensively managed ranch. The number of liming, herbiciding, and disking treatments applied to each pasture management unit was positively correlated with NPV and negatively correlated with PV. Although these objective scalars revealed signs of degradation, intensive management kept exposed soil to <40% cover and maintained economically viable cattle production over several decades. In ranches with few management inputs, the high PV cover in young pastures declined with increasing pasture age, while NPV and S increased, even where grazing intensity was low. Both highly productive pastures and vigorous regrowth of native vegetation cause high PV values. Analysis of spectral properties holds promise for identifying areas where grazing intensity has exceeded management inputs, thus increasing coverage of senescent foliage and exposed soil.

As wildfire activity increases in many regions of the world, it is imperative that we understand how key components of fire‐prone ecosystems respond to spatial variation in fire characteristics. Pollinators provide a foundation for ecological communities by assisting in the reproduction of native plants, yet our understanding of how pollinators such as wild bees respond to variation in fire severity is limited, particularly for forest ecosystems. Here, we took advantage of a natural experiment created by a large‐scale, mixed‐severity wildfire to provide the first assessment of how wild bee communities are shaped by fire severity in mixed‐conifer forest. We sampled bees in the Douglas Fire Complex, a 19,000‐ha fire in southern Oregon,USA, to evaluate how bee communities responded to local‐scale fire severity. We found that fire severity served a strong driver of bee diversity: 20 times more individuals and 11 times more species were captured in areas that experienced high fire severity relative to areas with the lowest fire severity. In addition, we found pronounced seasonality in the local bee community, with more individuals and more species captured during late summer, especially in severely burned regions of the landscape. Two critical habitat components for maintaining bee populations—flowering plants and boring insect exit holes used by cavity‐nesting bees—also increased with fire severity. Although we detected shifts in the relative abundance of several bee and plant genera along the fire severity gradient, the two most abundant bee genera (BombusandHalictus) responded positively to high fire severity despite differences in their typical foraging ranges. Our study demonstrates that within a large wildfire mosaic, severely burned forest contained the most diverse wild bee communities. This finding has particularly important implications for biodiversity in fire‐prone areas given the expected expansion of wildfires in the coming decades.

Many studies have examined how fuels, topography, climate, and fire weather influence fire severity. Less is known about how different forest management practices influence fire severity in multi-owner landscapes, despite costly and controversial suppression of wildfires that do not acknowledge ownership boundaries. In 2013, the Douglas Complex burned over 19,000ha of Oregon & California Railroad (O&C) lands in Southwestern Oregon, USA. O&C lands are composed of a checkerboard of private industrial and federal forestland (Bureau of Land Management, BLM) with contrasting management objectives, providing a unique experimental landscape to understand how different management practices influence wildfire severity. Leveraging Landsat based estimates of fire severity (Relative differenced Normalized Burn Ratio, RdNBR) and geospatial data on fire progression, weather, topography, pre-fire forest conditions, and land ownership, we asked (1) what is the relative importance of different variables driving fire severity, and (2) is intensive plantation forestry associated with higher fire severity? Using Random Forest ensemble machine learning, we found daily fire weather was the most important predictor of fire severity, followed by stand age and ownership, followed by topographic features. Estimates of pre-fire forest biomass were not an important predictor of fire severity. Adjusting for all other predictor variables in a general least squares model incorporating spatial autocorrelation, mean predicted RdNBR was higher on private industrial forests (RdNBR 521.85±18.67 [mean ± SE]) vs. BLM forests (398.87±18.23) with a much greater proportion of older forests. Our findings suggest intensive plantation forestry characterized by young forests and spatially homogenized fuels, rather than pre-fire biomass, were significant drivers of wildfire severity. This has implications for perceptions of wildfire risk, shared fire management responsibilities, and developing fire resilience for multiple objectives in multi-owner landscapes.

With an increase in the frequency and severity of wildfires across the globe and resultant changes to long-established fire regimes, the mapping of fire severity is a vital part of monitoring ecosystem resilience and recovery. The emergence of unoccupied aircraft systems (UAS) and compact sensors (RGB and LiDAR) provide new opportunities to map fire severity. This paper conducts a comparison of metrics derived from UAS Light Detecting and Ranging (LiDAR) point clouds and UAS image based products to classify fire severity. A workflow which derives novel metrics describing vegetation structure and fire severity from UAS remote sensing data is developed that fully utilises the vegetation information available in both data sources. UAS imagery and LiDAR data were captured pre- and post-fire over a 300 m by 300 m study area in Tasmania, Australia. The study area featured a vegetation gradient from sedgeland vegetation (e.g., button grass 0.2m) to forest (e.g., Eucalyptus obliqua and Eucalyptus globulus 50m). To classify the vegetation and fire severity, a comprehensive set of variables describing structural, textural and spectral characteristics were gathered using UAS images and UAS LiDAR datasets. A recursive feature elimination process was used to highlight the subsets of variables to be included in random forest classifiers. The classifier was then used to map vegetation and severity across the study area. The results indicate that UAS LiDAR provided similar overall accuracy to UAS image and combined (UAS LiDAR and UAS image predictor values) data streams to classify vegetation (UAS image: 80.6%; UAS LiDAR: 78.9%; and Combined: 83.1%) and severity in areas of forest (UAS image: 76.6%, UAS LiDAR: 74.5%; and Combined: 78.5%) and areas of sedgeland (UAS image: 72.4%; UAS LiDAR: 75.2%; and Combined: 76.6%). These results indicate that UAS SfM and LiDAR point clouds can be used to assess fire severity at very high spatial resolution.