LiDAR Variables Research Articles

Bridging gaps: On the performance of airborne LiDAR to model wood mouse-habitat structure relationships in pine forests.

LiDAR technology has firmly contributed to strengthen the knowledge of habitat structure-wildlife relationships, though there is an evident bias towards flying vertebrates. To bridge this gap, we investigated and compared the performance of LiDAR and field data to model habitat preferences of wood mouse (Apodemus sylvaticus) in a Mediterranean high mountain pine forest (Pinus sylvestris). We recorded nine field and 13 LiDAR variables that were summarized by means of Principal Component Analyses (PCA). We then analyzed wood mouse’s habitat preferences using three different models based on: (i) field PCs predictors, (ii) LiDAR PCs predictors; and (iii) both set of predictors in a combined model, including a variance partitioning analysis. Elevation was also included as a predictor in the three models. Our results indicate that LiDAR derived variables were better predictors than field-based variables. The model combining both data sets slightly improved the predictive power of the model. Field derived variables indicated that wood mouse was positively influenced by the gradient of increasing shrub cover and negatively affected by elevation. Regarding LiDAR data, two LiDAR PCs, i.e. gradients in canopy openness and complexity in forest vertical structure positively influenced wood mouse, although elevation interacted negatively with the complexity in vertical structure, indicating wood mouse’s preferences for plots with lower elevations but with complex forest vertical structure. The combined model was similar to the LiDAR-based model and included the gradient of shrub cover measured in the field. Variance partitioning showed that LiDAR-based variables, together with elevation, were the most important predictors and that part of the variation explained by shrub cover was shared. LiDAR derived variables were good surrogates of environmental characteristics explaining habitat preferences by the wood mouse. Our LiDAR metrics represented structural features of the forest patch, such as the presence and cover of shrubs, as well as other characteristics likely including time since perturbation, food availability and predation risk. Our results suggest that LiDAR is a promising technology for further exploring habitat preferences by small mammal communities.

Using remote-sensing data to assess habitat selection of a declining passerine at two spatial scales

Detailed information on habitat needs is integral to identify conservation measures for declining species. However, field data on habitat structure is typically limited in extent. Remote sensing has the potential to overcome these limitations of field-based studies. We aimed to assess abiotic and biotic characteristics of territories used by the declining wood warbler (Phylloscopus sibilatrix), a forest-interior migratory passerine, at two spatial scales by evaluating a priori expectations of habitat selection patterns. First, territories established by males before pairing, referred to as pre-breeding territories, were compared to pseudo-absence control areas located in the wider forested landscape (first spatial scale, Nterritories = 66, Ncontrols = 66). Second, breeding territories of paired wood warblers were compared to true-absence control areas located immediately close-by in the forest (second spatial scale, Nterritories = 78, Ncontrols = 78). Habitat variables predominantly described forest structure and were mainly based on first and last pulse lidar (light detection and ranging) data. Occurrence of pre-breeding territories was related to vegetation height, vertical diversity and stratification, canopy cover, inclination and solar radiation. Occurrence of breeding territories was associated to vegetation height, vertical diversity and inclination. Territory selection at the two spatial scales addressed was governed by similar factors. With respect to conservation, habitat suitability for wood warblers could be retained by maintaining a shifting mosaic of stand ages and structures at large spatial scales. Moreover, leaf-off lidar variables have the potential to contribute to understanding the ecological niche of species in predominantly deciduous forests.

Estimation of Above Ground Forest Biomass and Carbon Stock by Integrating Lidar, Satellite Image and Field Measurement in Nepal

For the first time in South Asia, the model-based Lidar Assisted Multisource Program (LAMP) was tested in 23500 km2 TAL area of Nepal by integrating 5% LiDAR sampling, wall-to-wall Rapid Eye satellite image and a representative field inventory to estimate Above Ground Biomass (AGB) and carbon stock. The average 1.26/m2LiDAR point density recorded by the scanner was used to measure canopy height and build a model using LiDAR variables and model coefficients. The developed LAMP model successfully estimated the AGB of the study area. The research tells that the study area comprises almost 50% forest cover with an average 211.63 t/ha AGB.Standing carbon stock was converted from AGB by multiplying the 0.47 which is default carbon fraction. Average standing carbon stock is 99.47 t/ha in the study area. The LAMP method found that the standing total AGB was 214.85-208.41 t/ha at a 95% confidence level and the FRA field-plot AGB estimate is 210.09/ha. This correspondence at this level of confidence means that the LAMP estimates are as accurate as those of the field-based inventory.J. Nat. Hist. Mus. Vol. 28, 2014: 160-170

Comparison of small-footprint discrete return and full waveform airborne lidar data for estimating multiple forest variables

The quantification of forest ecosystems is important for a variety of purposes, including the assessment of wildlife habitat, nutrient cycles, timber yield and fire propagation. This research assesses the estimation of forest structure, composition and deadwood variables from small-footprint airborne lidar data, both discrete return (DR) and full waveform (FW), acquired under leaf-on and leaf-off conditions. The field site, in the New Forest, UK, includes managed plantation and ancient, semi-natural, coniferous and deciduous woodland. Point clouds were rendered from the FW data through Gaussian decomposition. An area-based regression approach (using Akaike Information Criterion analysis) was employed, separately for the DR and FW data, to model 23 field-measured forest variables. A combination of plot-level height, intensity/amplitude and echo-width variables (the latter for FW lidar only) generated from both leaf-on and leaf-off point cloud data were utilised, together with individual tree crown (ITC) metrics from rasterised leaf-on height data. Statistically significant predictive models (p b 0.05) were generated for all 23 forest metrics using both the DR and FW lidar datasets, with R2 values for the best fit models in the range R2= 0.43–0.94 for the DR data and R2= 0.28–0.97 for the FW data (with normalised RMSE values being 18%–66% and 16%–48% respectively). For all but two forest metrics the difference between the NRMSE of the best performing DR and FW models was ≤7%, and there was an even split (11:12) as to which lidar dataset (DR or FW) generated the best model per forest metric. Overall, the DR data performed better at modelling structure variables, whilst the FW data performed better at modelling composition and deadwood variables. Neither showed a clear advantage at modelling variables from a particular vegetation layer (canopy, shrub or ground). Height, intensity/ amplitude, and ITC-derived crown variables were shown to be important inputs across the best performing models (DR or FW), but the additional echo-width variables available from FW point data were relatively unimportant. Of perhaps greater significance to the choice between lidar data type (i.e. DR or FW) in determining the predictive power of the best performing models was the selection of leaf-on and/or leaf-off data. Of the 23 best models, 10 contained both leaf-on and leaf-off lidar variables, whilst 11 contained only leaf-on and two only leaf-off data. We therefore conclude that although FW lidar has greater vertical profile information than DR lidar, the greater complimentary information about the entire forest canopy profile that is available from both leaf-on and leaf-off data is of more benefit to forest inventory, in general, than the selection between DR or FW lidar.

Estimation of standing wood volume and species composition in managed nemoral multi-layer mixed forests by using nearest neighbour classifier, multispectral satellite images and airborne lidar data / Puistute liigilise koosseisu ja tüvemahu hindamine k-lähima naabri meetodil mitmerindelistes majandatavates segametsades

Abstract. Nearest neighbour techniques are useful for constructing maps of forest inventory variables based on sample plot and auxiliary data from remote sensing. The most problematic issue of the nearest neighbour technique is possible systematic bias in the estimated values. In this study a 15 by 15 km test site in nemoral multilayer mixed forests was established in Laeva, Estonia. A set of 444 circular sample plots was used as reference set. Airborne lidar data and Landsat-8 Operational Land Imager image were used to construct five different feature variable sets consisting of original variables and alternatively principal components. The response variables were wood volume of first tree layer, wood volume of the second tree layer and main species code. A special test was carried out where a substantial amount of Silver birch dominated plots were removed from the reference set. The wood volume prediction validation was carried out on 89 forest growth sample plots and on 2290 forest stands. Species composition prediction was validated on 986 forest stands. As in many previous studies the results confirmed superiority of airborne lidar variables over spectral variables for wood volume estimation. The first three principal components of airborne lidar variables and first five principal components of all possible original feature variables contained over 99% of the information and performed well in imputations. The imputed wood volume at small values was overestimated and underestimated at large values regardless of used reference set. The feature variable sets containing spectral data performed better for species composition imputation. There was a forest age dependent discrepancy in predicted species proportions: birch and spruce proportions were underestimated in young stands and overestimated in older stands while proportion of aspen had exactly the opposite errors. The lack of fit depended slightly on the feature variable sets. The birch dominated plot partial removal from the reference set changed the predicted proportion of species but did not remove the forest age dependent lack of fit. The result can be important for the studies in which bootstrap samples are used to estimate error statistics for nearest neighbour technique based forest inventory variable maps.

Landsat Time Series and Lidar as Predictors of Live and Dead Basal Area Across Five Bark Beetle-Affected Forests

Bark beetle-caused tree mortality affects important forest ecosystem processes. Remote sensing methodologies that quantify live and dead basal area (BA) in bark beetle-affected forests can provide valuable information to forest managers and researchers. We compared the utility of light detection and ranging (lidar) and the Landsat-based detection of trends in disturbance and recovery (LandTrendr) algorithm to predict total, live, dead, and percent dead BA in five bark beetle-affected forests in Alaska, Arizona, Colorado, Idaho, and Oregon, USA. The BA response variables were predicted from lidar and LandTrendr predictor variables using the random forest (RF) algorithm. RF models explained 28%-61% of the variation in BA responses. Lidar variables were better predictors of total and live BA, whereas LandTrendr variables were better predictors of dead and percent dead BA. RF models predicting percent dead BA were applied to lidar and LandTrendr grids to produce maps, which were then compared to a gridded dataset of tree mortality area derived from aerial detection survey (ADS) data. Spearman correlations of beetle-caused tree mortality metrics between lidar, LandTrendr, and ADS were low to moderate; low correlations may be due to plot sampling characteristics, RF model error, ADS data subjectivity, and confusion caused by the detection of other types of forest disturbance by LandTrendr. Provided these sources of error are not too large, our results show that lidar and LandTrendr can be used to predict and map live and dead BA in bark beetle-affected forests with moderate levels of accuracy.

Mapping and monitoring Mount Graham red squirrel habitat with Lidar and Landsat imagery

The Mount Graham red squirrel (Tamiasciurus hudsonicus grahamensis) is an endemic subspecies located in the Pinaleño Mountains of southeast Arizona. Living in a conifer forest on a sky-island surrounded by desert, the Mount Graham red squirrel is one of the rarest mammals in North America. Over the last two decades, drought, insect infestations, and fire destroyed much of its habitat. A federal recovery team is working on a plan to recover the squirrel and detailed information is necessary on its habitat requirements and population dynamics. Toward that goal I developed and compared three probabilistic models of Mount Graham red squirrel habitat with a geographic information system and logistic regression. Each model contained the same topographic variables (slope, aspect, elevation), but the Landsat model contained a greenness variable (Normalized Difference Vegetation Index) extracted from Landsat, the Lidar model contained three forest-inventory variables extracted from lidar, while the Hybrid model contained Landsat and lidar variables. The Hybrid model produced the best habitat classification accuracy, followed by the Landsat and Lidar models, respectively. Landsat-derived forest greenness was the best predictor of habitat, followed by topographic (elevation, slope, aspect) and lidar (tree height, canopy bulk density, and live basal area) variables, respectively. The Landsat model's probabilities were significantly correlated with all 12 lidar variables, indicating its utility for habitat mapping. While the Hybrid model produced the best classification results, only the Landsat model was suitable for creating a habitat time series or habitat–population function between 1986 and 2013. The techniques I highlight should prove valuable in the development of Landsat- or lidar-based habitat models range wide.

Modelling Forest α-Diversity and Floristic Composition — On the Added Value of LiDAR plus Hyperspectral Remote Sensing

The decline of biodiversity is one of the major current global issues. Still, there is a widespread lack of information about the spatial distribution of individual species and biodiversity as a whole. Remote sensing techniques are increasingly used for biodiversity monitoring and especially the combination of LiDAR and hyperspectral data is expected to deliver valuable information. In this study spatial patterns of vascular plant community composition and α-diversity of a temperate montane forest in Germany were analysed for different forest strata. The predictive power of LiDAR (LiD) and hyperspectral (MNF) datasets alone and combined (MNF+LiD) was compared using random forest regression in a ten-fold cross-validation scheme that included feature selection and model tuning. The final models were used for spatial predictions. Species richness could be predicted with varying accuracy (R2 = 0.26 to 0.55) depending on the forest layer. In contrast, community composition of the different layers, obtained by multivariate ordination, could in part be modelled with high accuracies for the first ordination axis (R2 = 0.39 to 0.78), but poor accuracies for the second axis (R2 ≤ 0.3). LiDAR variables were the best predictors for total species richness across all forest layers (R2 LiD = 0.3, R2 MNF = 0.08, R2 MNF+LiD = 0.2), while for community composition across all forest layers both hyperspectral and LiDAR predictors achieved similar performances (R2 LiD = 0.75, R2 MNF = 0.76, R2 MNF+LiD = 0.78). The improvement in R2 was small (≤0.07)—if any—when using both LiDAR and hyperspectral data as compared to using only the best single predictor set. This study shows the high potential of LiDAR and hyperspectral data for plant biodiversity modelling, but also calls for a critical evaluation of the added value of combining both with respect to acquisition costs.

Allometric equation choice impacts lidar-based forest biomass estimates: A case study from the Sierra National Forest, CA

Regression modeling of biomass estimates from an airborne, multiple return lidar system using regional biomass allometric equations differs significantly from those using national scale Jenkins allometric equations with respect to the amount of variation explained, variables selected and variables of importance. Our discrete return lidar data were collected in September 2007 at 121 plots in a conifer dominated forest site in the Sierra Nevada mountains that include a full range of forest density. We regressed field plot-level estimates of biomass derived from field data and two different allometric equations with a range of lidar metrics. We compared regression performance across eight models: (1) point clouds alone, (2) point clouds with an empirical relationship between DBH and height (i.e., volume), (3) individual tree-level metrics, and (4) all data combined, and across two allometric equations – (A) Forest Inventory Analysis (FIA), and (B) Jenkins. In lower biomass plots, the reference above ground biomass (AGB) estimates from regional allometric equations and Jenkins equations were closely related; in plots with large biomass they were different. This finding suggests that published equations from large biomass plots are either not readily available or less represented in national scale allometric equation compiling. Models using reference AGBs calculated from regional allometric equations performed much better than those using reference AGBs calculated from Jenkins allometric equations. In these cases adjusted R2 improvement ranged from 0.07 to 0.11. The regression model that used regional allometric equations with lidar metrics and individual tree data provided the best overall R2 (0.79) with lowest RMSE suggesting that in most conditions regional biomass equations should be preferred over national equations. The inclusion of volumetric metrics shows that lidar variables are more sensitive to the reference AGBs calculated from regional allometric equations, and care should be taken when substituting regional equations using national scale compiled allometric equations in regional biomass studies. In addition, consistent with previous studies, the mean height of individual trees identified was chosen by both models with both reference AGBs calculated from regional allometric equations and those calculated from Jenkins equations, supporting the need to identify individual trees for biomass prediction. Based on these results, we conclude that the selection of allometric equations can influence the capacity of lidar data to estimate biomass significantly, and a careful selection of allometric equations is required for regional lidar biomass studies.

Modeling forest songbird species richness using LiDAR-derived measures of forest structure

Conservation of biodiversity requires information at many spatial scales in order to detect and preserve habitat for many species, often simultaneously. Vegetation structure information is particularly important for avian habitat models and has largely been unavailable for large areas at the desired resolution. Airborne LiDAR, with its combination of relatively broad coverage and fine resolution provides existing new opportunities to map vegetation structure and hence avian habitat. Our goal was to model the richness of forest songbirds using forest structure information obtained from LiDAR data. In deciduous forests of southern Wisconsin, USA, we used discrete-return airborne LiDAR to derive forest structure metrics related to the height and density of vegetation returns, as well as composite variables that captured major forest structural elements. We conducted point counts to determine total forest songbird richness and the richness of foraging, nesting, and forest edge-related habitat guilds. A suite of 35 LiDAR variables were used to model bird species richness using best-subsets regression and we used hierarchical partitioning analysis to quantify the explanatory power of each variable in the multivariate models. Songbird species richness was correlated most strongly with LiDAR variables related to canopy and midstory height and midstory density (R 2 = 0.204, p < 0.001). Richness of species that nest in the midstory was best explained by canopy height variables (R 2 = 0.197, p < 0.001). Species that forage on the ground responded to mean canopy height and the height of the lower canopy (R 2 = 0.149, p < 0.005) while aerial foragers had higher richness where the canopy was tall and dense and the midstory more sparse (R 2 = 0.216, p < 0.001). Richness of edge-preferring species was greater where there were fewer vegetation returns but higher density in the understory (R 2 = 0.153, p < 0.005). Forest interior specialists responded positively to a tall canopy, developed midstory, and a higher proportion of vegetation returns (R 2 = 0.195, p < 0.001). LiDAR forest structure metrics explained between 15 and 20% of the variability in richness within deciduous forest songbird communities. This variability was associated with vertical structure alone and shows how LiDAR can provide a source of complementary predictive data that can be incorporated in models of wildlife habitat associations across broad geographical extents.

Spinning a laser web: predicting spider distributions using LiDAR

LiDAR remote sensing has been used to examine relationships between vertebrate diversity and environmental characteristics, but its application to invertebrates has been limited. Our objectives were to determine whether LiDAR-derived variables could be used to accurately describe single-species distributions and community characteristics of spiders in remote forested and mountainous terrain. We collected over 5300 spiders across multiple transects in the Bavarian National Park (Germany) using pitfall traps. We examined spider community characteristics (species richness, the Shannon index, the Simpson index, community composition, mean body size, and abundance) and single-species distribution and abundance with LiDAR variables and ground-based measurements. We used the R2 and partial R2 provided by variance partitioning to evaluate the predictive power of LiDAR-derived variables compared to ground measurements for each of the community characteristics. The total adjusted R2 for species richness, the Shannon index, community species composition, and body size had a range of 25-57%. LiDAR variables and ground measurements both contributed >80% to the total predictive power. For species composition, the explained variance was approximately 32%, which was significantly greater than expected by chance. The predictive power of LiDAR-derived variables was comparable or superior to that of the ground-based variables for examinations of single-species distributions, and it explained up to 55% of the variance. The predictability of species distributions was higher for species that had strong associations with shade in open-forest habitats, and this niche position has been well documented across the European continent for spider species. The similar statistical performance between LiDAR and ground-based measures at our field sites indicated that deriving spider community and species distribution information using LiDAR data can provide not only high predictive power at relatively low cost, but may also allow unprecedented mapping of community- and species-level spider information at scales ranging from stands to landscapes. Therefore, LiDAR is a viable tool to assist species-specific conservation as well as broader biodiversity planning efforts not only for a growing list of vertebrates, but for invertebrates as well.

Laser remote sensing of canopy habitat heterogeneity as a predictor of bird species richness in an eastern temperate forest, USA

Habitat heterogeneity has long been recognized as a fundamental variable indicative of species diversity, in terms of both richness and abundance. Satellite remote sensing data sets can be useful for quantifying habitat heterogeneity across a range of spatial scales. Past remote sensing analyses of species diversity have largely been limited to correlative studies based on the use of vegetation indices or derived land cover maps. A relatively new form of laser remote sensing (lidar) provides another means to acquire information on habitat heterogeneity. Here we examine the efficacy of lidar metrics of canopy structural diversity as predictors of bird species richness in the temperate forests of Maryland, USA. Canopy height, topography and the vertical distribution of canopy elements were derived from lidar imagery of the Patuxent National Wildlife Refuge and compared to bird survey data collected at referenced grid locations. The canopy vertical distribution information was consistently found to be the strongest predictor of species richness, and this was predicted best when stratified into guilds dominated by forest, scrub, suburban and wetland species. Similar lidar variables were selected as primary predictors across guilds. Generalized linear and additive models, as well as binary hierarchical regression trees produced similar results. The lidar metrics were also consistently better predictors than traditional remotely sensed variables such as canopy cover, indicating that lidar provides a valuable resource for biodiversity research applications.