Evaluation of pre/post-fire differenced spectral indices for assessing burn severity in a Mediterranean environment with Landsat Thematic Mapper

In this study several pre/post-fire differenced spectral indices for assessing burn severity in a Mediterranean environment are evaluated. GeoCBI (Geo Composite Burn Index) field data of burn severity were correlated with remotely sensed measures, based on the NBR (Normalized Burn Ratio), the NDMI (Normalized Difference Moisture Index) and the NDVI (Normalized Difference Vegetation Index). In addition, the strength of the correlation was evaluated for specific fuel types and the influence of the regression model type is pointed out. The NBR was the best remotely sensed index for assessing burn severity, followed by the NDMI and the NDVI. For this case study of the 2007 Peloponnese fires, results show that the GeoCBI–dNBR (differenced NBR) approach yields a moderate–high R ²?=?0.65. Absolute indices outperformed their relative equivalents, which accounted for pre-fire vegetation state. The GeoCBI–dNBR relationship was stronger for forested ecotypes than for shrub lands. The relationship between the field data and the dNBR and dNDMI (differenced NDMI) was nonlinear, while the GeoCBI–dNDVI (differenced NDVI) relationship appeared linear.     

GeoCBI: A modified version of the Composite Burn Index for the initial assessment of the short-term burn severity from remotely sensed data

Burn severity estimation is a key factor in the post-fire management. Previous studies using remotely sensed data to retrieve burn severity, as measured by the Composite Burn Index (CBI), have found inconsistencies, since spectral indices work well in some ecosystems but not in others. These inconsistencies may be caused by the lack of spectral uniqueness in the CBI definition, or by the performance of the spectral indices used. This paper analyses the former aspect, using a simulation analysis to study the relationships between the CBI and reflectance. Subsequently, a modified version of this index, called GeoCBI, is proposed to improve the retrieval of burn severity from remotely sensed data. GeoCBI takes into account the fraction of cover (FCOV) of the different vegetation strata used to compute the CBI. Moreover, it also includes the changes in the leaf area index (LAI) for the intermediate and tall tree strata (D+E). Field and simulation results show that GeoCBI is more consistently related to spectral reflectance than CBI for different ranges of burn severities, while keeping its ecological meaning.     

The temporal dimension of differenced Normalized Burn Ratio (dNBR) fire/burn severity studies: The case of the large 2007 Peloponnese wildfires in Greece

The temporal dimension of differenced Normalized Burn Ratio (dNBR) fire/burn severity studies was studied for the case of the large 2007 Peloponnese wildfires in Greece. Fire severity is defined as the degree of environmental change as measured immediately post-fire, whereas burn severity combines the direct fire impact and ecosystems responses. Geo Composite Burn Index (GeoCBI), two pre-/post-fire differenced Thematic Mapper (TM) dNBR assessments and a Moderate Resolution Imaging Spectroradiometer (MODIS) dNBR time series were used to analyze the temporal dimension. MODIS dNBR time series were calculated based on the difference between the NBR of the burned and control pixels, which were retrieved using time series similarity of a pre-fire year. The analysis incorporated the optimality statistic, which evaluates index performance based on displacements in the mid-infrared-near infrared bi-spectral space. Results showed a higher correlation between field and TM data early post-fire (R² = 0.72) than one-year post-fire (R² = 0.56). Additionally, mean dNBR (0.56 vs. 0.29), the dNBR standard deviation (0.29 vs. 0.19) and mean optimality (0.65 vs. 0.47) were clearly higher for the initial assessment than for the extended assessment. This is due to regenerative processes that obscured first-order fire effects impacting the suitability of the dNBR to assess burn severity in this case study. This demonstrates the importance of the lag timing, i.e. time since fire, of an assessment, especially in a quickly recovering Mediterranean ecosystem. The MODIS time series was used to study intra-annual changes in index performance. The seasonal timing of an assessment highly impacts what is actually measured. This seasonality affected both the greenness of herbaceous resprouters and the productivity of the control pixels, which is land cover specific. Appropriate seasonal timing of an assessment is therefore of paramount importance to anticipate false trends (e.g. caused by senescence). Although these findings are case study specific, it can be expected that similar temporal constraints affect assessments in other ecoregions. Therefore, within the limitations of available Landsat imagery, caution is recommended for the temporal dimension when assessing post-fire effects. This is crucial, especially for studies that aim to evaluate trends in fire/burn severity across space and time. Also, clarification in associated terminology is suggested.     

Properties of X-, C- and L-band repeat-pass interferometric SAR coherence in Mediterranean pine forests affected by fires

Synthetic Aperture Radar (SAR) data has been investigated to determine the relationship between burn severity and interferometric coherence at three sites affected by forest fires in a hilly Mediterranean environment. Repeat-pass SAR images were available from the TerraSAR-X, ERS-1/2, Envisat ASAR and ALOS PALSAR sensors. Coherence was related to measurements of burn severity (Composite Burn Index) and remote sensing estimates expressed by the differenced normalized burn ratio (dNBR) index. In addition, the effects of topography and weather on coherence estimates were assessed. The analysis for a given range of local incidence angle showed that the co-polarized coherence increases with the increase of burn severity at X- and C-band whereas cross-polarized coherence was practically insensitive to burn severity. Higher sensitivity to burn severity was found at L-band for both co- and cross-polarized channels. The association strength between coherence and burn severity was strongest for images acquired under stable, dry environmental conditions. When the local incidence angle is accounted for the determination coefficients increased from 0.6 to 0.9 for X- and C-band. At L-band the local incidence angle had less influence on the association strength to burn severity.     

Understanding burn severity sensing in Arctic tundra: exploring vegetation indices,suboptimal assessment timing and the impact of increasing pixel size

Little is known about how satellite imagery can be used to describe burn severity in tundra landscapes. The Anaktuvuk River Fire (ARF) in 2007 burned over 1000 km² of tundra on the North Slope of Alaska, creating a mosaic of small (1 m²) to large (>100 m²) patches that differed in burn severity. The ARF scar provided us with an ideal landscape to determine if a single-date spectral vegetation index can be used once vegetation recovery began and to independently determine how pixel size influences burn severity assessment. We determine and explore the sensitivity of several commonly used vegetation indices to variation in burn severity across the ARF scar and the influence of pixel size on the assessment and classification of tundra burn severity. We conducted field surveys of spectral reflectance at the peak of the first growing season post-fire (extended assessment period) at 18 field sites that ranged from high to low burn severity. In comparing single-date indices, we found that the two-band enhanced vegetation index (EVI2) was highly correlated with normalized burn ratio (NBR) and better distinguished among three burn severity classes than both the NBR and the normalized difference vegetation index (NDVI). We also show clear evidence that shortwave infrared (SWIR) reflectivity does not vary as a function of burn severity. By comparing a Quickbird scene (2.4 m pixels) to simulated 30 and 250 m pixel scenes, we are able to confirm that while the moderate spatial resolution of the Landsat Thematic Mapper (TM) sensor (30 m) is sufficient for mapping tundra burn severity, the coarser resolution of the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor (250 m) is not well matched to the fine scale of spatial heterogeneity in the ARF burn scar.     

Mapping burned areas and burn severity patterns in SW Australian eucalypt forest using remotely-sensed changes in leaf area index

Remote sensing is the most practical method available to managers of fire-prone forests for quantifying and mapping fire impacts. Differenced Normalised Burn Ratio (ΔNBR) is among the most widely used spectral indices for the mapping of burn severity but is difficult to interpret in terms of fire-related changes in key biophysical attributes and processes. We propose to quantify burn severity as a change in the leaf area index (ΔLAI) of a stand. LAI is a key biophysical attribute of forests, and is central to understanding their water and carbon cycles. Previous studies have suggested that changes in canopy LAI may be a major contributor to ΔNBR and to the composite burn index (CBI) that is frequently used in combination with the NBR to assess burn severity on the ground. We applied remotely-sensed ΔLAI to map burn severity in jarrah (Eucalyptus marginata) forest in south-western Australia burnt during the January 2005 Perth Hills wildfires. Ground-based digital photography was used to measure LAI in typical stands representing the full range of canopy densities present in the study area as well as variation in the time since the last fire. Regression models for the prediction of LAI were developed using NBR, the Normalised Difference Vegetation Index (NDVI) or the Simple Ratio (SR) as the independent variable. All three LAI models had equally high coefficients of determination (R²: 0.87) and small root mean squared errors (RMSE: 0.27–0.28). ΔLAI was calculated as the difference between pre- and post-fire LAI, predicted using imagery from January 2004 and February 2005, respectively. The area affected by the January 2005 fire and the burn severity patterns within that area were mapped using ΔLAI and ΔNBR. Landscape patterns of burn severity obtained from differencing pre- and post-fire LAI were similar to those mapped by ΔNBR. We conclude that fire-affected areas and burn severity patterns in the northern jarrah forest can be objectively mapped using remotely-sensed changes in LAI, while offering the important advantage over NBR of being readily interpretable in the wider context of ecological forest management.     

Post-fire vegetation response as a proxy to quantify the magnitude of burn severity in tropical peatland

In recent years, fires in tropical forests in Southeast Asia have become more frequent and widespread, resulting in an increased need to evaluate fire impacts at a landscape scale. We examine whether post-fire vegetation regrowth can be used as a proxy to evaluate burn severity in a peatland landscape in Central Kalimantan, Indonesian Borneo, that has been subject to frequent fires. Several single- and bi-temporal indices as well as spectral fraction endmembers derived from either a post-fire image or a combination of pre- and post-fire images obtained by the Landsat sensor were examined. Spectral data were correlated with vegetation variables obtained from in situ measurements collected 4 years after the last fire. Of the tested spectral data, the bi-temporal and single normalized burn ratio (dNBR and NBR) showed the strongest correlations with the sets of vegetation variables (i.e. total woody aboveground biomass, tree density, and number of trees <10 cm diameter at breast height (DBH)). The results of an analysis of variance (ANOVA) and Tukey's multiple comparison of means test confirmed that NBR, dNBR, and the normalized difference water index could delineate four regrowth classes, thus confirming their utility in separating areas subjected to a single fire from those affected by multiple fires (MFs) as well as for discrimination between fires of differing severity. The results (a) provide evidence of the long-lasting impact that MFs have on forest recovery in this ecosystem and (b) confirm that vegetation response can be used as a proxy to quantify burn severity in locations affected by MFs.     

Detecting post-fire burn severity and vegetation recovery using multitemporal remote sensing spectral indices and field-collected composite burn index data in a ponderosa pine forest

It is challenging to detect burn severity and vegetation recovery because of the relatively long time period required to capture the ecosystem characteristics. Multitemporal remote sensing data can provide multitemporal observations before, during and after a wildfire, and can improve the change detection accuracy. The goal of this study is to examine the correlations between multitemporal spectral indices and field-observed burn severity, and to provide a practical method to estimate burn severity and vegetation recovery. The study site is the Jasper Fire area in the Black Hills National Forest, South Dakota, that burned during August and September 2000. Six multitemporal Landsat images acquired from 2000 (pre-fire), 2001 (post-fire), 2002, 2003, 2005 and 2007 were used to assess burn severity. The normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), normalized burn ratio (NBR), integrated forest index (IFI) and the differences of these indices between the pre-fire and post-fire years were computed and analysed with 66 field-based composite burn index (CBI) plots collected in 2002. Results showed that differences of NDVI and differences of EVI between the pre-fire year and the first two years post-fire were highly correlated with the CBI scores. The correlations were low beyond the second year post-fire. Differences of NBR had good correlation with CBI scores in all study years. Differences of IFI had low correlation with CBI in the first year post-fire and had good correlation in later years. A CBI map of the burnt area was produced using regression tree models and the multitemporal images. The dynamics of four spectral indices from 2000 to 2007 indicated that both NBR and IFI are valuable for monitoring long-term vegetation recovery. The high burn severity areas had a much slower recovery than the moderate and low burn areas.     

Densemask RCNN: A Hybrid Model for Skin Burn Image Classification and Severity Grading

In the medical image processing, automatic segmentation of burn images is one of the critical tasks in the classification of skin burn into normal and burn area. Traditional models identify the burns from the image and distinguish the region as burn and non-burn regions. However, the earlier models cannot accurately classify the wound region and also requires more time in the prediction of burns. Also, the burn depth analysis is an important factor for the calculation of the percentage of burn depth i.e. degree of severity is analyzed by Total body surface area (TBSA). For those issues, we design a hybrid approach named DenseMask Regional convolutional neural network (RCNN) approach for segmenting the skin burn region based on the various degrees of burn severity. In this, hybrid integration of Mask-region based convolution neural network CNN (Mask R-CNN) and dense pose estimation are integrated into DenseMask RCNN that calculate the full-body human pose and performs semantic segmentation. At first, we use the Residual Network with a dilated convolution using a weighted mapping model to generate the dense feature map. Then the feature map is fed into the Region proposal network (RPN) which utilizes a Feature pyramid network (FPN) to detect the objects at different scales of location and pyramid level from the input images. For the accurate alignment of pixel-to-pixel labels, we introduce a Region of interest (RoI)-pose align module that properly aligns the objects based on the human pose with the characteristics of scale, right-left, translation, and left–right flip to a standard scale. After the alignment task, a cascaded fully convolutional architecture is employed on the top of the RoI module that performs mask segmentation and dense pose regression task simultaneously. Finally, the transfer learning model classifies the detected burn regions into three classes of wound depths. Experimental analysis is performed on the burn dataset and the result obtained shows better accuracy than the state-of-art approaches.

     

Comparison of AVIRIS and Landsat ETM+ detection capabilities for burn severity

Our study compares data on burn severity collected from multi-temporal Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) with similar data from the Enhanced Thematic Mapper Plus (ETM+) using the differenced Normalized Burn Ratio (dNBR). Two AVIRIS and ETM+ data acquisitions recorded surface conditions immediately before the Hoover Fire began to spread rapidly and again the following year. Data were validated with 63 field plots using the Composite Burn Index (CBI). The relationship between spectral channels and burn severity was examined by comparing pre- and post-fire datasets. Based on the high burn severity comparison, AVIRIS channels 47 and 60 at wavelengths of 788 and 913 nm showed the greatest negative response to fire. Post-fire reflectance values decreased the most on average at those wavelengths, while channel 210 at 2370 nm showed the greatest positive response on average. Fire increased reflectance the most at that wavelength over the entire measured spectral range. Furthermore, channel 210 at 2370 nm exhibited the greatest variation in spectral response, suggesting potentially high information content for fire severity. Based on general remote sensing principles and the logic of variable spectral responses to fire, dNBR from both sensors should produce useful results in quantifying burn severity. The results verify the band-response relationships to burn severity as seen with ETM+ data and confirm the relationships by way of a distinctly different sensor system.     

Estimating burn severity from Landsat dNBR and RdNBR indices across western Canada

National parks in western Canada experience wildland fire events at differing frequencies, intensities, and burn severities. These episodic disturbances have varying implications for various biotic and abiotic processes and patterns. To predict burn severity, the differenced Normalized Burn Ratio (dNBR) algorithm, derived from Landsat imagery, has been used extensively throughout the wildland fire community. In Canada, few accuracy assessments have been undertaken to compare the accuracy of the dNBR algorithm to its relative form (RdNBR). To investigate the accuracies of these two algorithms in Canada's National Parks, we hypothesized that RdNBR would outperform dNBR in two specific applications based on former research by Miller and Thode (2007). The first was the capacity of the RdNBR to produce more accurate results than dNBR over a wide range of fires and secondly in pre-fire landscapes with low canopy closure and high heterogeneity. To investigate these questions, dNBR and RdNBR indices were extracted from Landsat imagery and compared to the measurements of the Composite Burn Index (Key & Benson, 2006). Following this, best fit models were developed and statistically tested at the individual, regional, overall, and vegetative levels. We then developed confusion matrices to assess the relative strength and weakness of each model. As an additional means of comparing model accuracy, we tested Hall et al.'s (2008) non-linear model in estimating burn severity for the study's western boreal region and individual fires. The results indicate that across all fires, the RdNBR-derived model did not estimate burn severity more accurately than dNBR (65.2% versus 70.2% classification accuracy, respectively) nor in the heterogeneous and low canopy cover landscapes. In addition, we conclude that RdNBR is no more effective than dNBR at the regional, individual, and fine-scale vegetation levels. The Hall et al. (2008) model was found to estimate burn severity in the western boreal region with a higher overall kappa than both the dNBR and RdNBR study models. The results herein support the continued research and pursuit of developing regional remote sensing derived models in western Canada.     

Landsat digital analysis of the initial recovery of burned tundra at Kokolik River, Alaska

During late July and early August 1977, a wildfire burned 48km² of tundra in northwestern Alaska near the Kokolik River. The environmental effects of the fire were studied in the field and from aircraft and Landsat data. Three categories of burn severity were delineated using an August 1977 Landsat scene acquired shortly after the fire stopped. Measurable reflectance increased in all three categories by the following year as determined from a Landsat image acquired in August 1978. Regrowth of vegetation in the year following the fire was measured using Landsat digital data and compared with field measurements from selected portions of the burned area. Live vascular plant cover doubled in one of the severely burned portions of the area and increased 33% in a lightly burned portion as determined from field measurements. Landsat-derived measurements showed an increase of 62.5% in reflectance for the severely burned areas, and 53% for the lightly burned areas, which is attributed to regrowth of vegetation. Within the most severely burned portion, 9.6 out of a total of 13.3 km² showed minimal recovery based on the Landsat-derived spectral data. Within the lightly burned portion, 5.9 out of a total of 13.5 km² showed the same range of spectral values as did the control areas. Prefire terrain and vegetation conditions were found to influence burn severity. The drier high-relief areas generally burned more completely than lower-lying wet areas. Satellite data acquired after the fire confirmed this for much of the burned area.     

Burn severity estimation from remotely sensed data: Performance of simulation versus empirical models

Burn severity is a key factor in post-fire assessment and its estimation is traditionally restricted to field work and empirical fitting from remotely sensed data. However, the first method is limited in terms of spatial coverage and cost effectiveness and the second is site- and data-specific. Since alternative approaches based on radiative transfer models (RTM) have been usefully applied in retrieving several biophysical plant parameters (leaf area index, water and dry matter content, chlorophyll), this paper has applied the inversion of a simulation model to estimate burn severity in terms of the Composite Burn Index (CBI). The performance of the model inversion method was compared to standard empirical techniques. The study area chosen was a large forest fire in central Spain which occurred in July 2005. The model inversion showed the most accurate estimation for high severity levels (for CBI > 2.7, RMSE = 0.30) and for unburned areas (CBI < 0.5, RMSE = 0). In both methodologies, the error associated to CBI from 0.5 to 2.7 was not acceptable (RMSE > 0.7), because it is higher than 25% of the total range of the index. Finally, burn severity maps from both methods were compared.     

High spatial resolution burn severity mapping of the New Jersey Pine Barrens with WorldView-3 near-infrared and shortwave infrared imagery

The WorldView-3 (WV-3) sensor, launched in 2014, is the first high-spatial resolution scanner to acquire imagery in the shortwave infrared (SWIR). A spectral ratio of the SWIR combined with the near-infrared (NIR) can potentially provide an effective differentiation of wildfire burn severity. Previous high spatial resolution sensors were limited to data from the visible and NIR for mapping burn severity, for example using the normalized difference vegetation index (NDVI). Drawing on a study site in the Pine Barrens of New Jersey, USA, we investigate optimal processing methods for analysing WV-3 data, with a focus on the pre-fire minus post-fire differenced normalized burn ratio (dNBR). Although the imagery, originally acquired with a 3.7 m instantaneous field of view, was aggregated to 7.5 m pixels by DigitalGlobe due to current licensing constraints, a slight additional smoothing of the data was nevertheless found to help reduce noise in the multi-temporal dNBR imagery. The highest coefficient of determination (R²) of the regressions of dNBR with the field-based composite burn index was obtained with a dNBR ratio produced with the NIR1 and SWIR6 bands. Only a very small increase in R² was found when dNBR was calculated using the average of NIR1 and NIR2 for the NIR bands, and SWIR5 to SWIR8 for the SWIR bands. dNBR calculated using SWIR1 as the NIR band produced notably lower R² values than when either NIR1 or NIR2 were used. Differenced NDVI data was found to produce models with a much lower R² than dNBR, emphasizing the importance of the shortwave infrared region for monitoring fire severity. High spatial resolution dNBR data from WV-3 can potentially provide valuable information on finer details regarding burn severity patterns than can be obtained from Landsat 30 m data.     

Assessment of fire severity and species diversity in the southern Appalachians using Landsat TM and ETM+ imagery

Relatively little is known about the disturbance ecology of large wildfires in the southern Appalachians. The occurrence of a 4000-ha wildfire in the Linville Gorge Wilderness area in western North Carolina has provided a rare opportunity to study a large fire with a range of severities. The objectives of this study were to 1) assess the potential for using multi-temporal Landsat imagery to map fire severity in the southern Appalachians, 2) examine the influences of topography and forest community type on the spatial pattern of fire severity; and 3) examine the relationship between predicted fire severity and changes in species richness. A non-linear regression equation predicted a field-based composite burn index (CBI) as a function of change in the Normalized Burn Ratio (dNBR) with an R² of 0.71. Fire severity was highest on drier landforms located on upper hillslopes, ridges, and on southwest aspects, and was higher in pine communities than in other forest types. Predicted CBI was positively correlated with changes in species richness and with the post-fire cover of pine seedlings (Pinus virginiana, P. rigida, and P. pungens), suggesting that burn severity maps can be used to predict community-level fire effects across large landscapes. Despite the relatively large size of this fire for the southern Appalachians, severity was strongly linked to topographic variability and pre-fire vegetation, and spatial variation in fire severity was correlated with changes in species richness. Thus, the Linville Gorge fire appears to have generally reinforced the ecological constraints imposed by underlying environmental gradients.     

Postfire soil burn severity mapping with hyperspectral image unmixing

Burn severity is mapped after wildfires to evaluate immediate and long-term fire effects on the landscape. Remotely sensed hyperspectral imagery has the potential to provide important information about fine-scale ground cover components that are indicative of burn severity after large wildland fires. Airborne hyperspectral imagery and ground data were collected after the 2002 Hayman Fire in Colorado to assess the application of high resolution imagery for burn severity mapping and to compare it to standard burn severity mapping methods. Mixture Tuned Matched Filtering (MTMF), a partial spectral unmixing algorithm, was used to identify the spectral abundance of ash, soil, and scorched and green vegetation in the burned area. The overall performance of the MTMF for predicting the ground cover components was satisfactory (r² = 0.21 to 0.48) based on a comparison to fractional ash, soil, and vegetation cover measured on ground validation plots. The relationship between Landsat-derived differenced Normalized Burn Ratio (dNBR) values and the ground data was also evaluated (r² = 0.20 to 0.58) and found to be comparable to the MTMF. However, the quantitative information provided by the fine-scale hyperspectral imagery makes it possible to more accurately assess the effects of the fire on the soil surface by identifying discrete ground cover characteristics. These surface effects, especially soil and ash cover and the lack of any remaining vegetative cover, directly relate to potential postfire watershed response processes.     

MODIS-informed greenness responses to daytime land surface temperature fluctuations and wildfire disturbances in the Alaskan Yukon River Basin

Pronounced climate warming and increased wildfire disturbances are known to modify forest composition and control the evolution of the boreal ecosystem over the Yukon River Basin (YRB) in interior Alaska. In this study, we evaluate the post-fire green-up rate using the normalized difference vegetation index (NDVI) derived from 250 m 7 day eMODIS (an alternative and application-ready type of Moderate Resolution Imaging Spectroradiometer (MODIS) data) acquired between 2000 and 2009. Our analyses indicate measureable effects on NDVI values from vegetation type, burn severity, post-fire time, and climatic variables. The NDVI observations from both fire scars and unburned areas across the Alaskan YRB showed a tendency of an earlier start to the growing season (GS); the annual variations in NDVI were significantly correlated to daytime land surface temperature (LST) fluctuations; and the rate of post-fire green-up depended mainly on burn severity and the time of post-fire succession. The higher average NDVI values for the study period in the fire scars than in the unburned areas between 1950 and 2000 suggest that wildfires enhance post-fire greenness due to an increase in post-fire evergreen and deciduous species components.     

Calibration and validation of the relative differenced Normalized Burn Ratio (RdNBR) to three measures of fire severity in the Sierra Nevada and Klamath Mountains, California, USA

Multispectral satellite data have become a common tool used in the mapping of wildland fire effects. Fire severity, defined as the degree to which a site has been altered, is often the variable mapped. The Normalized Burn Ratio (NBR) used in an absolute difference change detection protocol (dNBR), has become the remote sensing method of choice for US Federal land management agencies to map fire severity due to wildland fire. However, absolute differenced vegetation indices are correlated to the pre-fire chlorophyll content of the vegetation occurring within the fire perimeter. Normalizing dNBR to produce a relativized dNBR (RdNBR) removes the biasing effect of the pre-fire condition. Employing RdNBR hypothetically allows creating categorical classifications using the same thresholds for fires occurring in similar vegetation types without acquiring additional calibration field data on each fire. In this paper we tested this hypothesis by developing thresholds on random training datasets, and then comparing accuracies for (1) fires that occurred within the same geographic region as the training dataset and in similar vegetation, and (2) fires from a different geographic region that is climatically and floristically similar to the training dataset region but supports more complex vegetation structure. We additionally compared map accuracies for three measures of fire severity: the composite burn index (CBI), percent change in tree canopy cover, and percent change in tree basal area. User's and producer's accuracies were highest for the most severe categories, ranging from 70.7% to 89.1%. Accuracies of the moderate fire severity category for measures describing effects only to trees (percent change in canopy cover and basal area) indicated that the classifications were generally not much better than random. Accuracies of the moderate category for the CBI classifications were somewhat better, averaging in the 50%-60% range. These results underscore the difficulty in isolating fire effects to individual vegetation strata when fire effects are mixed. We conclude that the models presented here and in Miller and Thode ([Miller, J.D. & Thode, A.E., (2007). Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (dNBR). Remote Sensing of Environment, 109, 66-80.]) can produce fire severity classifications (using either CBI, or percent change in canopy cover or basal area) that are of similar accuracy in fires not used in the original calibration process, at least in conifer dominated vegetation types in Mediterranean-climate California.     

Characterization of post-fire surface cover, soils, and burn severity at the Cerro Grande Fire, New Mexico, using hyperspectral and multispectral remote sensing

Forest fires leave behind a changed ecosystem with a patchwork of surface cover that includes ash, charred organic matter, soils and soil minerals, and dead, damaged, and living vegetation. The distributions of these materials affect post-fire processes of erosion, nutrient cycling, and vegetation regrowth. We analyzed high spatial resolution (2.4 m pixel size) Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) data collected over the Cerro Grande fire, to map post-fire surface cover into 10 classes, including ash, soil minerals, scorched conifer trees, and green vegetation. The Cerro Grande fire occurred near Los Alamos, New Mexico, in May 2000. The AVIRIS data were collected September 3, 2000. The surface cover map revealed complex patterns of ash, iron oxide minerals, and clay minerals in areas of complete combustion. Scorched conifer trees, which retained dry needles heated by the fire but not fully combusted by the flames, were found to cover much of the post-fire landscape. These scorched trees were found in narrow zones at the edges of completely burned areas. A surface cover map was also made using Landsat Enhanced Thematic Mapper plus (ETM+) data, collected September 5, 2000, and a maximum likelihood, supervised classification. When compared to AVIRIS, the Landsat classification grossly overestimated cover by dry conifer and ash classes and severely underestimated soil and green vegetation cover. In a comparison of AVIRIS surface cover to the Burned Area Emergency Rehabilitation (BAER) map of burn severity, the BAER high burn severity areas did not capture the variable patterns of post-fire surface cover by ash, soil, and scorched conifer trees seen in the AVIRIS map. The BAER map, derived from air photos, also did not capture the distribution of scorched trees that were observed in the AVIRIS map. Similarly, the moderate severity class of Landsat-derived burn severity maps generated from the differenced Normalized Burn Ratio (dNBR) calculation had low agreement with the AVIRIS classes of scorched conifer trees. Burn severity and surface cover images were found to contain complementary information, with the dNBR map presenting an image of degree of change caused by fire and the AVIRIS-derived map showing specific surface cover resulting from fire.     

Ten years of forest cover change in the Sierra Nevada detected using Landsat satellite image analysis

The Sierra Nevada of California is a region where large wildfires have been suppressed for over a century. A detailed geographic record of recent vegetation regrowth and disturbance patterns in forests of the Sierra Nevada remains a gap that can be filled with remote-sensing data. Landsat Thematic Mapper imagery was analysed to detect 10 years of recent changes (between 2000 and 2009) in forest vegetation cover for areas burned by wildfires between years of 1995 and 1999 in the region. Results confirmed the prevalence of regrowing forest vegetation during the period 2000 and 2009 over 17% of the combined burned areas. Classification of these regrowing forest vegetation areas by the Landsat normalized burn ratio (NBR) showed that there was a marked increase in average disturbance index (ΔDI) values in the transitions from low to moderate to high burn severity classes. Within the five combined wildfire perimeters, 45% of the high burn severity area delineated by the RdNBR analysis was covered by regrowing forest detected between 2000 and 2009. In contrast, a notable fraction (12%) of the entire 5 km (unburned) buffer area outside the 1995–1999 fires perimeters showed decline in forest cover, and not nearly as many regrowing forest areas, covering only 3% of all the 1995–1999 buffer areas combined. Based on comparison of these results to ground-based survey data and high-resolution aerial images, the Landsat difference index (ΔDI) methodology can fulfil much of the need for consistent, low-cost monitoring of changes due to climate and biological factors in western forest regrowth following stand-replacing disturbances.