Search

The Composite Burn Index (CBI) is commonly linked to remotely sensed data to understand spatial and temporal patterns of burn severity. However, a comprehensive understanding of the tradeoffs between different methods used to model CBI with remotely sensed data is lacking. To help understand the current state of the science, provide a blueprint towards conducting broad-scale meta-analyses, and identify key decision points and potential rationale, we conducted a review of studies that linked remotely sensed data to continuous estimates of burn severity measured with the CBI and related methods. We provide a roadmap of the different methodologies applied and examine potential rationales used to justify them. Our findings largely reflect methods applied in North America – particularly in the western USA – due to the high number of studies in that region. We find the use of different methods across studies introduces variations that make it difficult to compare outcomes. Additionally, the existing suite of comparative studies focuses on one or few of many possible sources of uncertainty. Thus, compounding error and propagation throughout the many decisions made during analysis is not well understood. Finally, we suggest a broad set of methodological information and key rationales for decision-making that could facilitate future reviews.

BACKGROUND: In explaining and predicting spatial vegetation patterns, ecologists have increasingly favoured the use of climatic water balance variables, including actual evapotranspiration (AET) and climatic water deficit (CWD), for representing the hydrologic and energetic environment experienced by plants. Much of the interest in these variables lies in their hypothesized potential to characterize biologically relevant environmental variation more directly than simple climate variables, such as precipitation and temperature. Practically, obtaining AET and CWD values across space requires hydrologic process models that involve assumptions, including assumptions about vegetation transpiration rates. However, transpiration parameter values are rarely known with precision and can vary several‐fold within and among vegetation types. APPROACH: We evaluate the extent to which assumptions about vegetation physiology in water balance models affect (a) relative spatial variation in modelled water balance values and (b) ecological inferences that are derived from analyses using water balance variables. We demonstrate an approach for identifying inferences that are robust to these assumptions. RESULTS: Assumptions about vegetation physiology can substantially affect relative spatial variation in modelled water balance values. More importantly, such assumptions can also substantially affect the inferences (e.g., expected vegetation distributions) drawn from ecological analyses that use water balance variables as predictors. Water balance variables are less sensitive to assumptions in environmental settings with abundant water supply, where AET variation is driven primarily by available energy (e.g., temperature and insolation), but they can be highly sensitive to assumptions in drier environments. CONCLUSION: Because of their sensitivity to assumptions, water balance variables are not unambiguously superior to simpler climatic and topographic variables, such as precipitation and temperature. However, they retain some advantages, primarily related to their mechanistic incorporation of interactions between water and energy, which may support their use in applications where sensitivity to hydrologic modelling assumptions is low or of minor concern.

Changing disturbance regimes and climate can overcome forest ecosystem resilience. Following high-severity fire, forest recovery may be compromised by lack of tree seed sources, warmer and drier postfire climate, or short-interval reburning. A potential outcome of the loss of resilience is the conversion of the prefire forest to a different forest type or nonforest vegetation. Conversion implies major, extensive, and enduring changes in dominant species, life forms, or functions, with impacts on ecosystem services. In the present article, we synthesize a growing body of evidence of fire-driven conversion and our understanding of its causes across western North America. We assess our capacity to predict conversion and highlight important uncertainties. Increasing forest vulnerability to changing fire activity and climate compels shifts in management approaches, and we propose key themes for applied research coproduced by scientists and managers to support decision-making in an era when the prefire forest may not return.

Forest-restoration efforts are increasing in the western United States in response to realized and expected changes in climate and disturbance regimes. Managers are challenged to find practical and defensible targets to shift forest composition, structure, and spatial pattern to a more resistant and resilient state. The Northeast Washington Forest Vision 2020 project on the Colville National Forest presented an opportunity to map and use previously uncaptured mesic stand-level historical reference conditions to a large restoration project. We reconstructed historical forest conditions in 12 plots across a range of plant-association groups and mapped five restoration treatment units after implementation. We evaluated treatment effectiveness both in terms of meeting the prescriptions’ stated objectives and by similarity to observed reference conditions using metrics of density, species composition, clump-size patterns, and open-space patterns. We found that dry plant associations were historically dominated by distributed clumps of large shade-intolerant trees, whereas cold mesic plant associations were structured as a gap-matrix spatial pattern. Treatments were effective at meeting prescribed density and species-composition targets, but generally resulted in stands that were overly uniform or clumped compared to historical reference conditions.

Post-fire tree mortality is a spatially structured process driven by interacting factors across multiple scales. However, empirical models of fire-caused tree mortality are generally not spatially explicit, do not differentiate among scales, and do not differentiate immediate from delayed mortality. We aimed to quantify cross-scale linkages between forest structure—including spatial patterns of trees—and the progression of mortality 1–4 years post-fire in terms of rates, causes, and underlying demography. We used data from a long-term study site in the Sierra Nevada, California to build a post-fire tree mortality model predicted by lidar-measured estimates of structure. We calculated structural metrics at scales from individual trees to 90 × 90 m neighborhoods and combined them with metrics for topography, site water balance, and burn weather to predict immediate and delayed post-fire tree mortality. Mortality rates decreased while average diameter of newly killed trees increased each year post-fire. Burn weather predictors as well as interactive terms across scales improved model fit and parsimony. Including landscape-scale information improved finer-scale predictions but not vice versa. The amount of fuel, fuel configuration, and burning conditions predicted total mortality at broader scales while tree group-scale fuel connectivity, tree species fire tolerance, and local stresses predicted the fine-scale distribution, timing, and agents of mortality. Landscape-scale conditions provide the template upon which finer-scale variation in post-fire tree mortality is arranged. Post-fire forest structure is associated with the etiologies of different mortality agents, and so landscape-level heterogeneity is a key part of ecosystem stability and resilience.

As the frequency of large, severe fires increases, detecting the drivers of spatial fire severity patterns is key to predicting controls provided by weather, fuels, topography, and management. Identify the biophysical and management drivers of severity patterns and their spatial variability across the 2013 Rim Fire, Sierra Nevada, California, USA. Random forest models were developed separately for reburned and fire-excluded (> 80 year) areas within Yosemite National Park (NP) and Stanislaus National Forest (NF). Models included biophysical, past disturbance, and spatial autocorrelation (SA) predictors. Variable importance was assessed globally and locally. Variance partitioning was used to assess pure and shared variance among predictors. High spatial variability in the relative dominance of predictors existed across burn days and between land ownerships. Fire weather was a dominant top-down control during plume-dominated fire spread days. However, bottom-up controls from fuels and topography created local, fine-scale heterogeneity throughout. Reburn severity correlated with previous severity suggesting strong landscape memory, particularly in Yosemite NP. SA analysis showed broad-scale spatial dependencies and high shared variance among predictors. Wildfires are inherently a multi-scaled process. Spatial structure in environmental variables create broad-scale patterns and dependencies among drivers leading to regions of similar fire behavior, while local bottom-up drivers generate fine-scaled heterogeneity. Identifying the conditions under which top-down factors overwhelm bottom-up controls can help managers monitor and manage wildfires to achieve both suppression and restoration goals. Restoration targeting both surface and ladder fuels can mediate future fire severity even under extreme weather conditions.

One challenge that land managers face in the southern Sierra Nevada is how to balance conservation of fisher (Pekania pennanti) habitat with the reintroduction of fire. The fisher population in the southern Sierra Nevada is of high conservation priority, due to its small population size, genetic isolation, and the risk of habitat loss due to wildfire and fuel reduction activities. It is unknown whether contemporary forests altered by fire suppression can support fisher habitat following the reintroduction of fire. We examined whether patch-scale forest habitat conditions used by fishers in a landscape with limited recent fire (Dinkey study area) also exist in a landscape with restored fire (Yosemite study area). We developed random forest and logistic regression models using lidar-derived forest structure metrics to distinguish reproductive den presence (n = 261) from randomly-generated “available” points (n = 261) within an estimate of the female population home range. The full logistic regression model correctly classified (under cross-validation) 69.5% of observations and the random forest model correctly classified 74.3%. The parsimonious logistic regression model we selected had comparable accuracy to the full model (correctly classified 68.8% of observations) and included the following variables: cover >2 m, 95th percentile height, and 25th percentile height. Partial dependence plots suggest thresholds at which predicted probability of reproductive den presence exceeds 50%: cover >2 m greater than 60%, 95th percentile height of at least 32 m, and 25th percentile height between 4 m and 14 m. We found that comparable thresholds of forest cover and tree height exist in burned areas in Yosemite; 43.0% of burned pixels had similar lidar-derived forest structural characteristics to those predictive of reproductive dens in Dinkey. Areas with similar forest structures occurred within a range of fire severities and years since the most recent fire, and particularly in low-severity fire conditions (mean differenced normalized burn ratio [dNBR] value: 128.4). These results are promising for land managers who face the challenge of simultaneously reducing the risk of high-severity fire and conserving fisher habitat, however more research is needed to conclude whether suitable fisher habitat can exist in burned areas at all scales of selection and for all activities of the fisher population.

Restoration of fire-dependent forests is often guided by reference conditions from forests with an active fire regime, thought to be resilient to current and future disturbances and stresses. Reference conditions are usually based on historical data or reconstruction, which greatly limits the scale and completeness of data that can be collected. In the Sierra Nevada of California, large areas with reintroduced active fire regimes coupled with extensive lidar data coverage provide the unique opportunity to develop a contemporary regional reference condition dataset across a wide gradient of biophysical conditions. We developed this dataset with a focus on three questions: (1) What is the geographic and environmental distribution of restored active-fire forest areas in the Sierra Nevada mixed-conifer zone? (2) What are the ranges of variation in forest structure and spatial patterns across reference areas? And (3) How do stand density, tree clumping, and canopy opening patterns vary by topography and climate in reference areas? We analyzed fire history and environmental conditions over 10.8 million ha, including 3.9 million ha in the Sierra Nevada mixed-conifer zone, and found 30,377 ha of restored active-fire areas. Although reference areas were distributed throughout the Sierra Nevada they were more abundant on National Park lands (81% of reference areas) than National Forest lands and were associated with higher lightning strike density. Lidar-measured ranges of variation in reference condition structure were broad, with tree densities of 6–320 trees ha−1 (median 107 trees ha−1), basal area of 0.01–113 m2 ha−1 (median 21 m2 ha−1), average size of closely associated clumps of trees from >1 to 207 trees (median 3.1 trees), and average percent of stand area >6 m from the nearest canopy ranging from 0% to 100% (median 5.1%). These ranges correspond well with past studies reporting density and spatial patterns of contemporary and historical active-fire reference stands in the Sierra Nevada, except this study observed greater total variation due to the much greater spatial extent of sampling. Within the montane forest zone, reference areas at middle elevations had lower density (86 vs. 121 trees ha−1), basal area, (13.7 vs. 31 m2 ha−1), and mean clump size (2.7 vs. 4.0 trees) compared to lower- and higher-elevation reference areas, while ridgetops had lower density (101 vs. 115 trees ha−1), basal area (19.6 vs. 24.1 m2 ha−1), and mean clump size (3.0 vs. 3.3 trees) and more open space (7.4% vs. 5.1%) than other landforms. Many of the relationships between physiography and reference structure were context-dependent, suggesting that management practices should create heterogeneous forest structure congruent with local climatic and topographic factors influencing stand conditions.

Mountain ecosystems contain strong elevational gradients in climate and vegetation that shape species distributions and the structure of animal communities. Nevertheless, studies of habitat selection for individual species rarely account for such gradients that often result in species being managed uniformly across their range, which may not improve conservation as intended. Therefore, we characterized variation in nocturnal habitat selection by 18 GPS‐tagged California spotted owls (Strix occidentalis occidentalis) along a 1400‐m elevational gradient in the Sierra Nevada, California. We characterized three‐dimensional forest structure with light detection and ranging data that we used in mixed‐effects resource‐ and step‐selection analyses of owl habitat selection. At lower elevations, owls selected stands with shorter trees, sites closer to hard edges between tall forests and open areas, sites with less diversity in forest seral types and sites with more ridge and southwest aspects. In contrast, owls at higher elevations selected the opposite. Within public forests that had taller trees and within their home range core (45% kernel density estimate of GPS points) areas, owls selected forests with less and more canopy cover at low and high elevations, respectively. Outside of their core areas, owls selected areas with fewer and more tall trees at low and high elevations, respectively. These findings may be explained by elevational gradients in prey distribution and variation in owl diet because owls consume more woodrats (Neotoma spp; earlier seral species) at lower elevations and more flying squirrels (Glaucomys sabrinus; older forest species) at higher elevations. Thus, at low elevations and in areas unlikely to support nesting, spotted owls could benefit from management that promotes woodrat habitat by encouraging oak regeneration and creating small brushy openings within forests with shorter (younger) trees. Conversely, at higher elevations, (1) enhancing flying squirrel habitat by promoting large trees and denser canopy on mesic sites and (2) managing for greater cover type diversity on southwest‐facing slopes and ridgetops is more likely to improve foraging habitat quality for spotted owls. The patterns of owl selection over elevational gradients has not been explicitly considered in most habitat management plans but clearly would improve management throughout mountain ecosystems.

We review science-based adaptation strategies for western North American (wNA) forests that include restoring active fire regimes and fostering resilient structure and composition of forested landscapes. As part of the review, we address common questions associated with climate adaptation and realignment treatments that run counter to a broad consensus in the literature. These include the following: (1) Are the effects of fire exclusion overstated? If so, are treatments unwarranted and even counterproductive? (2) Is forest thinning alone sufficient to mitigate wildfire hazard? (3) Can forest thinning and prescribed burning solve the problem? (4) Should active forest management, including forest thinning, be concentrated in the wildland urban interface (WUI)? (5) Can wildfires on their own do the work of fuel treatments? (6) Is the primary objective of fuel reduction treatments to assist in future firefighting response and containment? (7) Do fuel treatments work under extreme fire weather? (8) Is the scale of the problem too great? Can we ever catch up? (9) Will planting more trees mitigate climate change in wNA forests? And (10) is post-fire management needed or even ecologically justified? Based on our review of the scientific evidence, a range of proactive management actions are justified and necessary to keep pace with changing climatic and wildfire regimes and declining forest heterogeneity after severe wildfires. Science-based adaptation options include the use of managed wildfire, prescribed burning, and coupled mechanical thinning and prescribed burning as is consistent with land management allocations and forest conditions. Although some current models of fire management in wNA are averse to short-term risks and uncertainties, the long-term environmental, social, and cultural consequences of wildfire management primarily grounded in fire suppression are well documented, highlighting an urgency to invest in intentional forest management and restoration of active fire regimes.

Understanding the patterns and relationships between vegetation productivity and climatic conditions is essential for predicting the future impacts of climate change. Climate change is altering precipitation patterns and increasing temperatures in the Southwest United States. The large-scale and long-term effects of these changes on vegetation productivity are not well understood. This study investigates the patterns and relationships between seasonal vegetation productivity, represented by Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI), and the Standardized Precipitation Evapotranspiration Index (SPEI) across the Mojave, Sonoran, and Chihuahuan Deserts and the Apache Highlands of the Southwest United States over 16 years from 2000 to 2015. To examine the spatiotemporal gradient and response of vegetation productivity to dry and wet conditions, we evaluated the linear trend of different SPEI timescales and correlations between NDVI and SPEI. We found that all four ecoregions are experiencing more frequent and severe drought conditions in recent years as measured by negative SPEI trends and severe negative SPEI values. We found that changes in NDVI were more strongly correlated with winter rather than summer water availability. Investigating correlations by vegetation type across all four ecoregions, we found that grassland and shrubland productivity were more dependent on summer water availability whereas sparse vegetation and forest productivity were more dependent on winter water availability. Our results can inform resource management and enhance our understanding of vegetation vulnerability to climate change.

Background tree mortality is a complex process that requires large sample sizes and long timescales to disentangle the suite of ecological factors that collectively contribute to tree stress, decline, and eventual mortality. Tree mortality associated with acute disturbance events, in contrast, is conspicuous and frequently studied, but there remains a lack of research regarding the role of background mortality processes in mediating the severity and delayed effects of disturbance. We conducted an empirical study by measuring the rates, causes, and spatial pattern of mortality annually among 32,989 individual trees within a large forest demography plot in the Sierra Nevada. We characterized the relationships between background mortality, compound disturbances (fire and drought), and forest spatial structure, and we integrated our findings with a synthesis of the existing literature from around the world to develop a conceptual framework describing the spatio‐temporal signatures of background and disturbance‐related tree mortality. The interactive effects of fire, drought, and background mortality processes altered the rate, spatial structuring, and ecological consequences of mortality. Before fire, spatially non‐random mortality was only evident among small (1

We evaluated the effects of postfire management on forest structure in mixed-conifer forests of northeastern Washington, USA. Postfire treatments were harvest-only, harvest combined with planting, planting-only, and postfire prescribed fire. We used aerial light detection and ranging (LiDAR) to measure vertical and horizontal components of postfire forest structure over a period of 2 to 32 years after fires. We compared treated areas to control areas with similar bioclimatic environments and past fire severity. We used niche overlap statistics to quantify distributions of individual forest structure components and PERMANOVA to assess forest structural response to the presence or absence of treatments, past fire severity, time since treatment, and bioclimatic setting. Harvest alone after fire decreased dominant tree height and reduced vertical canopy complexity and the cover of tall trees. Harvest combined with planting increased dominant tree height, vertical complexity, and cover in lower height strata. Planting and prescribed fires showed little difference in forest structure relative to untreated controls. Overall, the burn severity of the initial fire was the strongest influence on postfire structure, and many aspects of vertical and horizontal forest structure showed little difference with increasing time since fire.

In the western US, wildfires are modifying the structure, composition, and patterns of forested landscapes at rates that far exceed mechanical thinning and prescribed fire treatments. There are conflicting narratives as to whether these wildfires are restoring landscape resilience to future climate and wildfires. To evaluate the landscape-level work of wildfires, we assessed four subwatersheds in eastern Washington, USA that experienced large wildfires in 2014, 2015, or 2017 after more than a century of fire exclusion and extensive timber harvest. We compared pre- and post-fire landscape conditions to an ecoregion-specific historical (HRV) and future range of variation (FRV) based on empirically established reference conditions derived from a large dataset of historical aerial photo imagery. These four wildfires proved to be a blunt restoration tool, moving some attributes towards more climate-adapted conditions and setting others back. Fires reduced canopy cover and decreased overall tree size and canopy complexity, which moved them into, or slightly outside, the FRV ranges. Moderate- and low-severity fire generally shifted closed-canopy forest structure to open-canopy classes. Patches of high-severity fire shifted patterns of forest, woodland, grassland, and shrubland towards or beyond the HRV ranges and within the FRV ranges by increasing the total area and size of non-forest patches. However, large patches of high-severity fire in dry and moist mixed-conifer forests homogenized landscape patterns beyond FRV ranges towards simplified conditions dominated by non-forest vegetation types. Fires realigned and reconnected landscape patterns with the topo-edaphic template in some cases, but pre-existing fragmentation and spatial mismatches were compounded in many others. Patches of large-tree, closed-canopy forest were reduced by high-severity fire, and the potential to restore more climate-adapted large-tree, open-canopy forest was lost. Re-establishing landscape patterns with desired patch sizes of forest, in particular patches with large trees, will take many decades to centuries and may not occur in drier locations or where seed trees are no longer present. While large wildfires burning during extreme fire weather conditions can move some attributes towards HRV and FRV ranges, intentionally planned mechanical and prescribed-fire treatments that are integrated with strategic wildfire response will better prepare and adapt landscapes for future wildfires and climate.

We investigated the relative importance of daily fire weather, landscape position, climate, recent forest and fuels management, and fire history to explaining patterns of remotely-sensed burn severity – as measured by the Relativized Burn Ratio – in 150 fires occurring from 2001 to 2019, which burned conifer forests of northeastern Washington State, USA. Daily fire weather, annual precipitation anomalies, and species’ fire resistance traits were important predictors of wildfire burn severity. In areas burned within the past two to three decades, prior fire decreased the severity of subsequent burns, particularly for the first 16 postfire years. In areas managed before a wildfire, thinning and prescribed burning treatments lowered burn severity relative to untreated controls. Prescribed burning was the most effective treatment at lowering subsequent burn severity, and prescribed burned areas were usually unburned or burned at low severity in subsequent wildfires. Patches that were harvested and planted

Post-fire landscapes are the frontline of forest ecosystem change. As such, they represent opportunities to foster conditions that are better adapted to future climate and wildfires with post-fire management. In western US landscapes, post-fire management has been mostly defined by short-term emergency mitigation measures, salvage harvest to recover economic value, and replanting to achieve full stocking. These approaches are largely incongruent with ecologically based forest management due to their limited scope and objectives. Here, we develop a framework for ecologically based post-fire management. Post-fire management principles are to (i) protect large-diameter trees and fire refugia; (ii) anticipate future fuel accumulation from post-fire tree mortality; (iii) reinitiate and maintain stabilizing fire-vegetation feedbacks; (iv) differentiate between climate- and dispersal-mediated transitions to non-forest; and (v) align species composition and structure with future fire regimes and climate. Stand-scale management strategies to implement these principles include (i) maintain or enhance forest resilience; (ii) restore forest conditions and resist transition to non-forest; and (iii) accept or facilitate transition to non-forest. Determining where and over what extent to deploy these stand-scale strategies in large, burned landscapes is informed by a post-fire landscape evaluation, and expressed with a landscape prescription. A post-fire landscape evaluation is a data-driven characterization of current vegetation conditions, including the immediate changes caused by wildfire, and includes a departure analysis—an evaluation of current conditions against reference conditions. The landscape prescription provides guidance about the distribution of different successional patches and their sizes across the topographic template and identifies priority areas for different post-fire treatments. We develop a geospatial framework to integrate ecological principles with a post-fire landscape evaluation that can be readily applied to management planning after wildfire. We illustrate application of these principles through the development of landscape prescriptions for two watersheds, each burned in a recent large fire, in northeast Washington, USA. Use of ecologically based post-fire management principles and landscape evaluations can help shift often contentious debates over salvage harvesting towards a more productive dialogue around how to best adapt landscapes to future conditions.

A century of land-use changes and intensive plantation-style forestry in the Pacific Northwest have resulted in a substantial deficit of structurally complex late-successional forests, which serve key ecological and social functions. Restoration treatments can theoretically accelerate the development of structural complexity in simplified forest stands, but our understanding of the short-term effects of such treatments on forest structure and gap patterns is still emerging. In this study, we used a bi-temporal airborne lidar dataset to assess short-term (10 years post-treatment) effects of restoration treatments on the structural complexity of forest stands in the Ellsworth Creek Preserve in southwestern Washington. We also compared forest structure and gap patterns between treated stands and existing late-successional areas within the preserve. Restoration treatments in older stands (60- to 80-years old) resulted in a greater increase in the mean and variability of both vertical and horizontal structural complexity and a considerable increase in the density of canopy gaps, indicating accelerated development towards increased structural complexity. Treatments in the younger stands (20- to 30-years old) primarily accelerated the development of vertical canopy complexity. Older treated stands within the preserve will require natural growth in canopy height, increased regeneration density, and filling in and diversification of canopy gaps to reach the structural complexity of late-successional forests in the preserve, but accelerating these developments may require additional restoration-focused treatments. Forest managers in the region should consider restoration treatments as a viable option for improving short-term structural complexity in simplified forest stands. We also encourage long-term management and monitoring plans that continue to address the deficit of late-successional forests in the region.

Restoration of western dry forests in the USA often focuses on reducing fuel loads. In the range of the spotted owl, these treatments may reduce canopy cover and tree density, which could reduce preferred habitat conditions for the owl and other sensitive species. In particular, high canopy cover (≥70%) has been widely reported to be an important feature of spotted owl habitat, but averages of stand-level forest cover do not provide important information on foliage height and gap structure. To provide better quantification of canopy structure, we used airborne LiDAR imagery to identify canopy cover in different height strata and the size and frequency of gaps that were associated with owl nest sites, protected activity centers (PACs), and territories within four study areas and 316 owl territories. Although total canopy cover was high in nest stands and PAC areas, the cover in tall (>48 m) trees was the canopy structure most highly selected for, while cover in lower strata (2–16 m) was avoided compared to availability in the surrounding landscape. Tall tree cover gradually decreased and lower strata cover increased as distance increased from the nest. Large (>1000 m2) gaps were not found near nests, but otherwise there was no difference in gap frequencies and sizes between PACs and territories and the surrounding landscape. Using cluster analysis we classified canopy conditions into 5 structural classes and 4 levels of canopy cover to assess the relationship between total canopy cover and tree size within nest sites, PACs, and territories. High canopy cover (≥70%) mostly occurs when large tree cover is high, indicating the two variables are often confounded. Our results suggest that the cover of tall trees may be a better predictor of owl habitat than total canopy cover because the latter can include cover in the 2–16 m strata – conditions that owls actually avoid. Management strategies designed to preserve and facilitate the growth of tall trees while reducing the cover and density of understory trees may improve forest resilience to drought and wildfire while also maintaining or promoting the characteristics of owl habitat.

Satellite-derived fire severity metrics are a foundational tool used to estimate fire effects at the landscape scale. Changes in surface characteristics permit reasonably accurate delineation between burned and unburned areas, but variability in severity within burned areas is much more challenging to detect. Previous studies have relied primarily on categorical data to calibrate severity indices in terms of classification accuracy, but this approach does not readily translate into an expected amount of error in terms of actual tree mortality. We addressed this issue by examining a dataset of 40,370 geolocated trees that burned in the 2013 California Rim Fire using 36 Landsat-derived burn severity indices. The differenced Normalized Burn Ratio (dNBR) performed reliably well, but the differenced SWIR:NIR ratio most accurately predicted percent basal area mortality and the differenced normalized vegetation index (dNDVI) most accurately predicted percent mortality of stems ≥10 cm diameter at breast height. Relativized versions of dNBR did not consistently improve accuracy; the relativized burn ratio (RBR) was generally equivalent to dNBR while RdNBR had consistently lower accuracy. There was a high degree of variability in observed tree mortality, especially at intermediate spectral index values. This translated into a considerable amount of uncertainty at the landscape scale, with an expected range in estimated percent basal area mortality greater than 37% for half of the area burned (>50,000 ha). In other words, a 37% range in predicted mortality rate was insufficient to capture the observed mortality rate for half of the area burned. Uncertainty was even greater for percent stem mortality, with half of the area burned exceeding a 46% range in predicted mortality rate. The high degree of uncertainty in tree mortality that we observed challenges the confidence with which Landsat-derived spectral indices have been used to measure fire effects, and this has broad implications for research and management related to post-fire landscape complexity, distribution of seed sources, or persistence of fire refugia. We suggest ways to account for uncertainty that will facilitate a more nuanced and ecologically-accurate interpretation of fire effects. This study makes three key contributions to the field of remote sensing of fire effects: 1) we conducted the most comprehensive comparison to date of all previously published severity indices using the largest contiguous set of georeferenced tree mortality field data and revealed that the accuracy of both absolute and relative spectral indices depends on the tree mortality metric of interest;2) we conducted this study in a single, large fire that enabled us to isolate variability due to intrinsic, within-landscape factors without the additional variance due to extrinsic factors associated with different biogeographies or climatic conditions; and 3) we identified the range in tree mortality that may be indistinguishable based on spectral indices derived from Landsat satellites, and we demonstrated how this variability translates into a considerable amount of uncertainty in fire effects at the landscape scale.

Research Highlights: We used airborne lidar to assess the multi-scalar patterns of forest structure across a large (471,000 hectare), multi-owner landscape of the Oregon Coast Range, USA. The results of this study can be used in the development and evaluation of conservation strategies focused on forest management. Background and Objectives: Human management practices reflect policy and economic decisions and shape forest structure through direct management and modification of disturbance regimes. Previous studies have found that land ownership affects forest cover, patch dynamics, structure, and ecosystem function and services. However, prior assessments of forest structure across landscapes and ownerships have been limited by a lack of high-fidelity forest structure measurements across a large spatial extent. We addressed three research questions: (1) What distinct classes of forest structure exist across our study area? (2) How does the distribution and pattern of forest structure vary among types of owners at scales of patches, ownership types, and subregion, and is this independent of property size? and (3) What implications do the fine and sub-regional scale patterns have for landscape configuration goals under recent updates to the Northwest Forest Plan? Materials and Methods: We examined forest structure patterns by identifying six statistically distinct classes of forest structure and then examining their distribution across and within ownership types. We used these structure classes to examine their area within each ownership class, mean patch size, and intermixing at multiple scales. Results and Conclusion: We found that the six different forest structure classes in the study area can be interpreted as two assemblages: production-style forests, principally on private lands, and structurally complex forests, principally on public lands. We found that land ownership objectives resulted in distinct landscape patterns of forest structure as measured by mean structure class patch size and intermixing of different structure class patches. Finally, we found that forest structure differed between public and private lands but differed comparatively little among ownership types within those two broad categories.

This paper presents a prototype Carbon Monitoring System (CMS) developed to produce regionally unbiased annual estimates of aboveground biomass (AGB). Our CMS employed a bottom-up, two-step modeling strategy beginning with a spatially and temporally biased sample: project datasets collected and contributed by US Forest Service (USFS) and other forestry stakeholders in 29 different project areas in the northwestern USA. Plot-level AGB estimates collected in the project areas served as the response variable for predicting AGB primarily from lidar metrics of canopy height and density (R2 = 0.8, RMSE = 115 Mg ha−1, Bias = 2 Mg ha−1). This landscape model was used to map AGB estimates at 30 m resolution where lidar data were available. A stratified random sample of AGB pixels from these landscape-level AGB maps then served as training data for predicting AGB regionally from Landsat image time series variables processed through LandTrendr. In addition, climate metrics calculated from downscaled 30 year climate normals were considered as predictors in both models, as were topographic metrics calculated from elevation data; these environmental predictors allowed AGB estimation over the full range of observations with the regional model (R2 = 0.8, RMSE = 152 Mg ha−1, Bias = 9 Mg ha−1), including higher AGB values (>400 Mg ha−1) where spectral predictors alone saturate. For both the landscape and regional models, the machine-learning algorithm Random Forests (RF) was consistently applied to select predictor variables and estimate AGB. We then calibrated the regional AGB maps using field plot data systematically collected without bias by the national Forest Inventory and Analysis (FIA) Program. We found both our project landscape and regional, annual AGB estimates to be unbiased with respect to FIA estimates (Biases of 1% and 0.7%, respectively) and conclude that they are well suited to inform forest management and planning decisions by our contributing stakeholders. Social media abstract Lidar-based biomass estimates can be upscaled with Landsat data to regionally unbiased annual maps.

Background Satellite-based aboveground forest biomass maps commonly form the basis of forest biomass and carbon stock mapping and monitoring, but biomass maps likely vary in performance by region and as a function of spatial scale of aggregation. Assessing such variability is not possible with spatially-sparse vegetation plot networks. In the current study, our objective was to determine whether high-resolution lidar-based and moderate-resolution Landsat-base aboveground live forest biomass maps converged on similar predictions at stand- to landscape-levels (10 s to 100 s ha) and whether such differences depended on biophysical setting. Specifically, we examined deviations between lidar- and Landsat-based biomass mapping methods across scales and ecoregions using a measure of error (normalized root mean square deviation), a measure of the unsystematic deviations, or noise (Pearson correlation coefficient), and two measures related to systematic deviations, or biases (intercept and slope of a regression between the two sets of predictions). Results Compared to forest inventory data (0.81-ha aggregate-level), lidar and Landsat-based mean biomass predictions exhibited similar performance, though lidar predictions exhibited less normalized root mean square deviation than Landsat when compared with the reference plot data. Across aggregate-levels, the intercepts and slopes of regression equations describing the relationships between lidar- and Landsat-based biomass predictions stabilized (i.e., little additional change with increasing area of aggregates) at aggregate-levels between 10 and 100 ha, suggesting a consistent relationship between the two maps at landscape-scales. Differences between lidar- and Landsat-based biomass maps varied as a function of forest canopy heterogeneity and composition, with systematic deviations (regression intercepts) increasing with mean canopy cover and hardwood proportion within forests and correlations decreasing with hardwood proportion. Conclusions Deviations between lidar- and Landsat-based maps indicated that satellite-based approaches may represent general gradients in forest biomass. Ecoregion impacted deviations between lidar and Landsat biomass maps, highlighting the importance of biophysical setting in determining biomass map performance across aggregate scales. Therefore, regardless of the source of remote sensing (e.g., Landsat vs. lidar), factors affecting the measurement and prediction of forest biomass, such as species composition, need to be taken into account whether one is estimating biomass at the plot, stand, or landscape scale.

Recent and projected increases in the frequency and severity of large wildfires in the western U.S. makes understanding the factors that strongly affect landscape fire patterns a management priority for optimizing treatment location. We compared the influence of variations in the local environment on burn severity patterns on the large 2013 Rim fire that burned under extreme drought with those of previous smaller fires for a study area in the Sierra Nevada, California, USA. Although much of the Rim fire burned during plume-dominated conditions resulting in large high-severity patches, our study area burned under milder fire weather resulting in a mix of fire severities. In our study area the Rim fire produced a higher proportion of moderate- and high-severity effects than occurred in previous fires. Random forest modeling explained up to 63% of the Rim fire burn variance using seven predictors: time since previous fire, actual evapotranspiration (AET), climatic water deficit (Deficit), previous maximum burn severity, burning index, slope position, and solar radiation. Models using only a subset of biophysical predictors (AET, Deficit, slope position and steepness, and solar radiation) explained 55% of the Rim fire and 58% of the maximum fire burn severity of previous fires. The relationship of burn severity to patterns of AET, however, reversed for the Rim fire (positive) compared to earlier fires (negative). Measurements of pre-Rim fire forest structure from LiDAR did not improve our ability to explain burn severity patterns. We found that accounting for spatial autocorrelation in burn severity and biophysical environment was important to model quality and stability. Our results suggest water balance and topography can help predict likely burn severity patterns under moderate climate and fire weather conditions, providing managers with general guidance for prioritizing fuel treatments and identifying where fire is less likely to burn with higher severities even for locations with higher forest density and canopy cover.

Mountainous topography creates fine-scale environmental mosaics that vary in precipitation, temperature, insolation, and slope position. This mosaic in turn influences fuel accumulation and moisture and forest structure. We studied these the effects of varying environmental conditions across a 27,104 ha landscape within Yosemite National Park, California, USA, on the number of fires and burn severity (measured from Landsat data for 1984–2010) and on canopy cover at two heights (>2 m and 2–8 m) and dominant tree height (measured with airborne LiDAR data). We used site water balance (actual evapotranspiration and climatic water deficit) and topography (slope position, slope, and insolation) as environmental predictors. Random forest modeling showed that environmental conditions predicted substantial portions of the variations in fire and forest structure: e.g., 85–93% of the variation in whether a location did not burn, burned once, or burned twice; 64% of the variation in the burn severity; and 72% of the variation in canopy cover >2 m for unburned forests, 64% for once-burned forests, and 59% for twice-burned forests. Environmental conditions also predicted a substantial portion of forest structure following one and two fires, even though the post-fire forest structures were substantially different than pre-fire structures. This suggests a feedback mechanism in which local fire regimes and pre-fire forest structures are related to local environments, and their interaction produces post-fire structures also related to local environments. Among environmental predictors, water balance had the greatest explanatory power, followed by slope position, and then by slope and insolation. Managers could use our methods to help select reference areas that match environmental conditions, identify areas at risk for fires that endanger critical habitat or other resources, and identify climate analog areas to help anticipate and plan for climate change.

The fate of live forest biomass is largely controlled by growth and disturbance processes, both natural and anthropogenic. Thus, biomass monitoring strategies must characterize both the biomass of the forests at a given point in time and the dynamic processes that change it. Here, we describe and test an empirical monitoring system designed to meet those needs. Our system uses a mix of field data, statistical modeling, remotely-sensed time-series imagery, and small-footprint lidar data to build and evaluate maps of forest biomass. It ascribes biomass change to specific change agents, and attempts to capture the impact of uncertainty in methodology. We find that A common image framework for biomass estimation and for change detection allows for consistent comparison of both state and change processes controlling biomass dynamics. Regional estimates of total biomass agree well with those from plot data alone. The system tracks biomass densities up to 450–500 Mg ha−1 with little bias, but begins underestimating true biomass as densities increase further. Scale considerations are important. Estimates at the 30 m grain size are noisy, but agreement at broad scales is good. Further investigation to determine the appropriate scales is underway. Uncertainty from methodological choices is evident, but much smaller than uncertainty based on choice of allometric equation used to estimate biomass from tree data. In this forest-dominated study area, growth and loss processes largely balance in most years, with loss processes dominated by human removal through harvest. In years with substantial fire activity, however, overall biomass loss greatly outpaces growth. Taken together, our methods represent a unique combination of elements foundational to an operational landscape-scale forest biomass monitoring program.

Following changes in vegetation structure and pattern, along with a changing climate, large wildfire incidence has increased in forests throughout the western United States. Given this increase, there is great interest in whether fuels treatments and previous wildfire can alter fire severity patterns in large wildfires. We assessed the relative influence of previous fuels treatments (including wildfire), fire weather, vegetation, and water balance on fire-severity in the Rim Fire of 2013. We did this at three different spatial scales to investigate whether the influences on fire severity changed across scales. Both fuels treatments and previous low to moderate-severity wildfire reduced the prevalence of high-severity fire. In general, areas without recent fuels treatments and areas that previously burned at high severity tended to have a greater proportion of high-severity fire in the Rim Fire. Areas treated with prescribed fire, especially when combined with thinning, had the lowest proportions of high severity. The proportion of the landscape burned at high severity was most strongly influenced by fire weather and proportional area previously treated for fuels or burned by low to moderate severity wildfire. The proportion treated needed to effectively reduce the amount of high severity fire varied by spatial scale of analysis, with smaller spatial scales requiring a greater proportion treated to see an effect on fire severity. When moderate and high-severity fire encountered a previously treated area, fire severity was significantly reduced in the treated area relative to the adjacent untreated area. Our results show that fuels treatments and low to moderate-severity wildfire can reduce fire severity in a subsequent wildfire, even when burning under fire growth conditions. These results serve as further evidence that both fuels treatments and lower severity wildfire can increase forest resilience.

Mosaics of tree clumps and openings are characteristic of forests dominated by frequent, low- and moderate-severity fires. When restoring these fire-suppressed forests, managers often try to reproduce these structures to increase ecosystem resilience. We examined unburned and burned forest structures for 1937 0.81 ha sample areas in Yosemite National Park, USA. We estimated severity for fires from 1984 to 2010 using the Landsat-derived Relativized differenced Normalized Burn Ratio (RdNBR) and measured openings and canopy clumps in five height strata using airborne LiDAR data. Because our study area lacked concurrent field data, we identified methods to allow structural analysis using LiDAR data alone. We found three spatial structures, canopy-gap, clump-open, and open, that differed in spatial arrangement and proportion of canopy and openings. As fire severity increased, the total area in canopy decreased while the number of clumps increased, creating a patchwork of openings and multistory tree clumps. The presence of openings > 0.3 ha, an approximate minimum gap size needed to favor shade-intolerant pine regeneration, increased rapidly with loss of canopy area. The range and variation of structures for a given fire severity were specific to each forest type. Low- to moderate-severity fires best replicated the historic clump-opening patterns that were common in forests with frequent fire regimes. Our results suggest that managers consider the following goals for their forest restoration: 1) reduce total canopy cover by breaking up large contiguous areas into variable-sized tree clumps and scattered large individual trees; 2) create a range of opening sizes and shapes, including ~ 50% of the open area in gaps > 0.3 ha; 3) create multistory clumps in addition to single story clumps; 4) retain historic densities of large trees; and 5) vary treatments to include canopy-gap, clump-open, and open mosaics across project areas to mimic the range of patterns found for each forest type in our study.

A century of fire suppression has left fire-dependent forests of the western United States increasingly vulnerable to wildfire, drought, and insects. Forest managers are trying to improve resilience using treatments such as mechanical thinning and prescribed fire; however, operational and resource constraints limit treatments to a fraction of the needed area each year. An alternative is to let wildfires burn under less-than-extreme fire weather where human lives and infrastructure are not at risk. We examined post-fire forest structure using airborne lidar data to determine whether a single wildland fire following an extended fire-free period could produce forest structures resembling fire-resilient historical conditions. We studied forest structures resulting from these “first-entry” fires in a forest with a history of timber management (2008 American River Complex Fires, Tahoe National Forest) and in a wilderness area (2009 Big Meadow Fire, Yosemite National Park). We compared the results of these first-entry fires with nearby reference areas that had experienced 2+ fires that burned predominately at low- and moderate-severity. We identified visible overstory trees from the lidar data and examined their patterns in terms of individuals, tree clumps, and openings. We found that moderate-severity fire effects in these first-entry fires produced similar patterns to the reference areas with area in openings at approximately 40% and trees predominately in small (2 to 4 trees) and medium (5–9 trees) clumps High-severity fire produced mortality likely to lead to large canopy openings that were historically uncharacteristic in these forests. As burn severity increased, the amount of the residual canopy area represented by taller trees (>16 m and especially >32 m) decreased, which could result from fires preferentially killing taller trees or from locations with taller trees more commonly experiencing lower burn severities. Our study suggests that first-entry fires allowed to burn under less-than-extreme conditions can reproduce spatial patterns resembling historical conditions resilient to fires and drought but possibly at the disproportionate expense of larger trees.

While fire shapes the structure of forests and acts as a keystone process, the details of how fire modifies forest structure have been difficult to evaluate because of the complexity of interactions between fires and forests. We studied this relationship across 69.2 km 2 of Yosemite National Park, USA, that was subject to 32 fires ⩾40 ha between 1984 and 2010. Forests types included ponderosa pine ( Pinus ponderosa ), white fir-sugar pine ( Abies concolor / Pinus lambertiana ), and red fir ( Abies magnifica ). We estimated and stratified burned area by fire severity using the Landsat-derived Relativized differenced Normalized Burn Ratio (RdNBR). Airborne LiDAR data, acquired in July 2010, measured the vertical and horizontal structure of canopy material and landscape patterning of canopy patches and gaps. Increasing fire severity changed structure at the scale of fire severity patches, the arrangement of canopy patches and gaps within fire severity patches, and vertically within tree clumps. Each forest type showed an individual trajectory of structural change with increasing fire severity. As a result, the relationship between estimates of fire severity such as RdNBR and actual changes appears to vary among forest types. We found three arrangements of canopy patches and gaps associated with different fire severities: canopy-gap arrangements in which gaps were enclosed in otherwise continuous canopy (typically unburned and low fire severities); patch-gap arrangements in which tree clumps and gaps alternated and neither dominated (typically moderate fire severity); and open-patch arrangements in which trees were scattered across open areas (typically high fire severity). Compared to stands outside fire perimeters, increasing fire severity generally resulted first in loss of canopy cover in lower height strata and increased number and size of gaps, then in loss of canopy cover in higher height strata, and eventually the transition to open areas with few or no trees. However, the estimated fire severities at which these transitions occurred differed for each forest type. Our work suggests that low severity fire in red fir forests and moderate severity fire in ponderosa pine and white fir-sugar pine forests would restore vertical and horizontal canopy structures believed to have been common prior to the start of widespread fire suppression in the early 1900s. The fusion of LiDAR and Landsat data identified post-fire structural conditions that would not be identified by Landsat alone, suggesting a broad applicability of combining Landsat and LiDAR data for landscape-scale structural analysis for fire management.

Post-fire tree mortality models are vital tools used by forest land managers to predict fire effects, estimate delayed mortality and develop management prescriptions. We evaluated the performance of mortality models within the First Order Fire Effects Model (FOFEM) software, and compared their performance to locally-parameterised models based on five different forms. We evaluated all models at the individual tree and stand levels with a dataset comprising 34174 trees from a mixed-conifer forest in the Sierra Nevada, California that burned in the 2013 Rim Fire. We compared stand-level accuracy across a range of spatial scales, and we used point pattern analysis to test the accuracy with which mortality models predict post-fire tree spatial pattern. FOFEM under-predicted mortality for the three conifers, possibly because of the timing of the Rim Fire during a severe multi-year drought. Locally-parameterised models based on crown scorch were most accurate in predicting individual tree mortality, but tree diameter-based models were more accurate at the stand level for Abies concolor and large-diameter Pinus lambertiana, the most abundant trees in this forest. Stand-level accuracy was reduced by spatially correlated error at small spatial scales, but stabilised at scales ≥1ha. The predictive error of FOFEM generated inaccurate predictions of post-fire spatial pattern at small scales, and this error could be reduced by improving FOFEM model accuracy for small trees.

LiDAR measurements of canopy structure can be used to classify forest stands into structural stages to study spatial patterns of canopy structure, identify habitat, or plan management actions. A key assumption in this process is that differences in canopy structure based on forest age and elevation are consistent with predictions from models of stand development. Three LiDAR metrics (95th percentile height, rumple, and canopy density) were computed for 59 secondary and 35 primary forest plots in the Pacific Northwest, USA. Hierarchical clustering identified two precanopy closure classes, two low-complexity postcanopy closure classes, and four high-complexity postcanopy closure classes. Forest development models suggest that secondary plots should be characterized by low-complexity classes and primary plots characterized by high-complexity classes. While the most and least complex classes largely confirmed this relationship, intermediate-complexity classes were unexpectedly composed of both secondary and primary forest types. Complexity classes were not associated with elevation, except that primary Tsuga mertensiana (Bong.) Carrière (mountain hemlock) plots were complex. These results suggest that canopy structure does not develop in a linear fashion and emphasize the importance of measuring structural conditions rather than relying on development models to estimate structural complexity across forested landscapes.

Forest structure, as measured by the physical arrangement of trees and their crowns, is a fundamental attribute of forest ecosystems that changes as forests progress through suc;cessional stages. We examined whether LiDAR data could be used to directly assess the successional stage of forests by determining the degree to which the LiDAR data would show the same relative ranking of structural development among sites as would traditional field measurements. We sampled 94 primary and secondary sites (19–93, 223–350, and 600 years old) from three conifer forest zones in western Washington state, USA, in the field and with small-footprint, discrete return LiDAR. Seven sets of LiDAR metrics were tested to measure canopy structure. Ordinations using the of LiDAR 95th percentile height, rumple, and canopy density metrics had the strongest correlations with ordinations using two sets of field metrics (Procrustes R = 0.72 and 0.78) and a combined set of LiDAR and field metrics (Procrustes R = 0.95). These results suggest that LiDAR can accurately characterize forest successional stage where field measurements are not available. This has important implications for enabling basic and applied studies of forest structure at stand to landscape scales.

Megafires have lasting social, ecological, and economic impacts and are increasing in the western contiguous United States. Because of their infrequent nature, there is a limited sample of megafires to investigate their unique behavior, drivers, and relationship to forest management practices. One approach is to characterize critical information pre-, during, and post-fire using remote sensing. In August 2013, the Rim Fire burned 104,131 ha and in September 2014, the King Fire burned 39,545 ha. Both fires occurred in California's Sierra Nevada. The areas burned by these fires were fortuitously surveyed by airborne campaigns, which provided the most recent remote sensing technologies not currently available from satellite. Technologies include an imaging spectrometer spanning the visible to shortwave infrared (0.38-2.5 μm), a multispectral, high-spatial resolution thermal infrared (3.5-13 μm) spectroradiometer, and Light Detection and Ranging that provide spatial resolutions of pixels from 1 × 1 m to 35 × 35 m. Because of the unique information inherently derived from these technologies before the fires, the areas were subsequently surveyed after the fires. We processed and provide free dissemination of these airborne datasets as products of surface reflectance, spectral metrics and forest structural metrics ( http://dx.doi.org/10.3334/ORNLDAAC/1288). These data products provide a unique opportunity to study relationships among and between remote sensing observations and fuel and fire characteristics (e.g., fuel type, condition, structure, and fire severity). The novelty of these data is not only in the unprecedented types of information available from them before, during, and after two megafires, but also in the synergistic use of multiple state of the art technologies for characterizing the environment. The synergy of these data can provide novel information that can improve maps of fuel type, structure, abundance, and condition that may improve predictions of megafire behavior and effects, thus aiding management before, during, and after such events. Key questions that these data could address include: What drives, extinguishes, and results from megafires? How does megafire behavior relate to fire and fuel management? How does the size and severity of a megafire affect the ecological recovery of the system?