Search

Trestima® is a computer vision-based smartphone application that utilises relascope theory to obtain estimates of forest attributes from smartphone photographs. The aim of this study was to investigate the accuracy of Trestima es - timation and evaluate whether it is sufficiently accurate for operational use in forestry. Our data consisted of 37 forest stands, encompassing 73.5 ha in southeastern Finland, where Trestima estimates were obtained by forestry professionals during their work. The results were compared with harvester data obtained from clear-cut stands. The number of photographs taken per stand ranged between 1-29 (average: 7.3; standard deviation: 5.0). The total amount of industrial roundwood harvested from the stands was 21,531 m3 and the average harvest removal per hectare was 282 m3 . The accuracy of Trestima estimation was relatively good when ≥ 10 photographs per stand were taken. In this case, the root mean square error percent (RMSE%) value associated with roundwood volume was 17.7%. When the number of photographs per stand was < 10, the accuracy of Trestima was much weaker (RMSE% 22.7- 55.3%). On average, Trestima underestimated harvested volumes in Scots pine (Pinus sylvestris L.) stands (Bias% 11.4-89.2), although the bias was smaller (Bias% –12.7-12.4) with Norway spruce (Picea abies [L.] Karst.) stands. The Trestima smartphone application is a possible option for traditional field measurements in operational forestry, provided that its usage instructions are strictly followed, which is not always the case in practice. [ABSTRACT FROM AUTHOR]

In this paper, we examine the transferability of airborne laser scanning (ALS) based models for individual-tree detection (ITD) from one ALS inventory area (A1) to two other areas (A2 and A3). All areas were located in eastern Finland less than 100 km from each other and were scanned using different ALS devices and parameters. The tree attributes of interest were diameter at breast height (Dbh), height (H), crown base height (Cbh), stem volume (V), and theoretical sawlog volume (Vlog) of Scots pine (Pinus sylvestris L.) with Dbh ≥ 16 cm. All trees were first segmented from the canopy height models, and various ALS metrics were derived for each segment. Then only the segments covering correctly detected pines were chosen for further inspection. The tree attributes were predicted using the k-nearest neighbor (k-NN) imputation. The results showed that the relative root mean square error (RMSE%) values increased for each attribute after the transfers. The RMSE% values were, for A1, A2, and A3, respectively: Dbh, 13.5%, 14.8%, and 18.1%; H, 3.2%, 5.9%, and 6.2%; Cbh, 13.3%, 15.3%, and 18.3%; V, 29.3%, 35.4%, and 39.1%; and Vlog, 38.2%, 54.4% and 51.8%. The observed values indicate that it may be possible to employ ALS-based tree-level k-NN models over different inventory areas without excessive reduction in accuracy, assuming that the tree species are known to be similar. [ABSTRACT FROM AUTHOR]

The objective of this study was to investigate the benefit of three different optical data sources (Sentinel-2, Landsat 8 and aerial images) to support airborne laser scanning (ALS) data in species-specific forest inventory. The data covered 633 sample plots in eastern Finland. We used nearest neighbor imputation for simultaneous prediction of Scots pine, Norway spruce and broadleaved species’ volume by species group. The variable selection was performed by means of simulated annealing of different data combinations. The results showed that, on average, all optical data sources improved the species-specific plot volume predictions. The improvement was always greatest for broadleaves. The species-specific root mean square errors were 64.3%, 61.5%, 58.1% and 57.9% for ALS, ALS+Landsat 8, ALS+Sentinel-2 and ALS+aerial image data combinations, respectively, and 54.2% for ALS with the channels of both aerial images and Sentinel-2. Compared to using just ALS and aerial images, adding the Sentinel’s second red edge and narrow near-infrared bands improved the separation of pine and spruce. Sentinel-2 outperformed Landsat 8 and was almost as good an option as aerial images. The results suggest that all optical data, from airborne or spaceborne sources, are useful when combined with ALS data in species-specific forest inventory. [ABSTRACT FROM AUTHOR]

We propose 3D triangulations of airborne Laser Scanning (ALS) point clouds as a new approach to derive 3D canopy structures and to estimate forest canopy effective LAI (LAI e ). Computational geometry and topological connectivity were employed to filter the triangulations to yield a quasi-optimal relationship with the field measured LAI e . The optimal filtering parameters were predicted based on ALS height metrics, emulating the production of maps of LAI e and canopy volume for large areas. The LAI e from triangulations was validated with field measured LAI e and compared with a reference LAI e calculated from ALS data using logarithmic model based on Beer’s law. Canopy transmittance was estimated using All Echo Cover Index (ACI), and the mean projection of unit foliage area (β) was obtained using no-intercept regression with field measured LAI e . We investigated the influence species and season on the triangulated LAI e and demonstrated the relationship between triangulated LAI e and canopy volume. Our data is from 115 forest plots located at the southern boreal forest area in Finland and for each plot three different ALS datasets were available to apply the triangulations. The triangulation approach was found applicable for both leaf-on and leaf-off datasets after initial calibration. Results showed the Root Mean Square Errors (RMSEs) between LAI e from triangulations and field measured values agreed the most using the highest pulse density data (RMSE = 0.63, the coefficient of determination (R 2 ) = 0.53). Yet, the LAI e calculated using ACI-index agreed better with the field measured LAI e (RMSE = 0.53 and R 2 = 0.70). The best models to predict the optimal alpha value contained the ACI-index, which indicates that within-crown transmittance is accounted by the triangulation approach. The cover indices may be recommended for retrieving LAI e only, but for applications which require more sophisticated information on canopy shape and volume, such as radiative transfer models, the triangulation approach may be preferred. [ABSTRACT FROM AUTHOR]

We present the first comparison of Sentinel-2A (S2) MSI (Multi-Spectral Instrument) and Landsat 8 (L8) OLI (Operational Land Imager) data in the retrieval of forest canopy cover (CC), effective canopy cover (ECC), and leaf area index (LAI). We used S2 and L8 images obtained from Suonenjoki, Finland on 17 and 22 August 2015, respectively. A combination of airborne lidar data and field plots was used to calculate CC, ECC and LAI for a set of 746 systematically placed lidar plots. Generalized additive models were used to link these variables with various types of spectral indices. Our results indicated that instead of using the native image resolution, the model accuracies were better for 60 m lidar plots comprising of nine (S2) or four (L8) image pixels. The best-case absolute root mean square errors of prediction (RMSEPs) (relative RMSEPs shown in parentheses) obtained from ten-fold cross validation of multivariate models for S2 CC, ECC and LAI were 0.126 (24.0%), 0.100 (20.8%), and 0.596 (19.6%), respectively. For L8, the corresponding RMSEPs were 0.128 (24.5%), 0.108 (22.4%), and 0.614 (20.2%). The marginally better performance of S2 models may be related to the 705 nm red edge band, which frequently occurred among the selected predictors. When testing indices that used bands available for both sensors, there were no systematic differences between S2 and L8. [ABSTRACT FROM AUTHOR]

Large area prediction of continuous field of tree cover i.e., canopy cover (CC) using Earth observation data is of high interest in practical forestry, ecology, and climate change mitigation activities. We report the accuracy of using Landsat images for CC prediction in boreal forests validated with field reference plots (N = 250) covering large variation in latitude, forest structure, species composition, and site type. We tested two statistical models suitable for estimating CC: the beta regression (BetaReg) and random forest (RanFor). Landsat-based predictors utilized include individual bands, spectral vegetation indices (SVI), and Tasseled cap (Tass) features. Additionally, we tested an alternative model based on spectral mixture analysis (SMA). Finally, we carried out a first validation in boreal forests of the recently published Landsat Tree Cover Continuous (TCC) global product. Results showed simple BetaReg with red band reflectance provided the highest prediction accuracy (leave-site-out RMSE CV 13.7%; R 2 CV 0.59; bias CV 0.5%). Spectral transformations into SVI and Tass did not improve accuracy. Including additional predictors did not significantly improve accuracy either. Nonlinear model RanFor did not outperform BetaReg. The alternative SMA model did not outperform the empirical models. However, empirical models cannot resolve the underestimation of high cover and overestimation of low cover. SMA prediction errors appeared less dependent on forest structure, while there seemed to be a potential for improvement by accounting for endmember variability of different tree species. Finally, using temporally concurrent observations, we showed the reasonably good accuracy of Landsat TCC product in boreal forests (RMSE 13.0%; R 2 0.53; bias −2.1%), however with a tendency to underestimate high cover. [ABSTRACT FROM AUTHOR]

Large-diameter trees are important for both ecological and economic reasons, but they have become increasingly rare. Thus, there is an interest in easily locating such trees, and for this purpose, the use of airborne laser scanning (ALS) seems suitable. Our objective was to assess the accuracy of area-based ALS estimation in predicting the number of large-diameter Scots pines ( Pinus sylvestris L.). A sample of 856 trees with a diameter >35 cm were measured from 1109 sample plots located in eastern Norway. We fitted negative binomial and zero-inflated negative binomial models for predicting large-diameter tree counts. ALS-derived and external variables were used as predictors when fitting the models. The accuracy was assessed based on the weighted kappa coefficient and cross validation. Our best model was based on three ALS height distribution variables, one horizontal ALS variable, and plot elevation. Its overall accuracy was 65.8% and the weighted kappa was 0.55. Although there was a clear relationship between the response and the proposed predictor variables, fairly large errors in the predicted large-diameter tree counts were common. [ABSTRACT FROM AUTHOR]

The estimation of individual biomass components within tree crowns, such as dead branches (DB), needles (NB), and branch biomass (BB), has received limited attention in the scientific literature despite their significant contribution to forest biomass. This study aimed to assess the potential of multispectral LiDAR data for estimating these biomass components in a multi-layered Abies borissi-regis forest. Destructive (i.e., 13) and non-destructive (i.e., 156) field measurements were collected from Abies borisii-regis trees to develop allometric equations for each crown biomass component and enrich the reference data with the non-destructively sampled trees. A set of machine learning regression algorithms, including random forest (RF), support vector regression (SVR) and Gaussian process (GP), were tested for individual-tree-level DB, NB and BB estimation using LiDAR-derived height and intensity metrics for different spectral channels (i.e., green, NIR and merged) as predictors. The results demonstrated that the RF algorithm achieved the best overall predictive performance for DB (RMSE% = 17.45% and R2 = 0.89), NB (RMSE% = 17.31% and R2 = 0.93) and BB (RMSE% = 24.09% and R2 = 0.85) using the green LiDAR channel. This study showed that the tested algorithms, particularly when utilizing the green channel, accurately estimated the crown biomass components of conifer trees, specifically fir. Overall, LiDAR data can provide accurate estimates of crown biomass in coniferous forests, and further exploration of this method's applicability in diverse forest structures and biomes is warranted. [ABSTRACT FROM AUTHOR]

Leaf area index (LAI) is one descriptor of forest canopy structure and can be linked to vegetation productivity, carbon cycling, and several other ecosystem services. Airborne lidar (light detection and ranging) provides proxies of canopy gap fraction (GF) in the near-vertical direction, which can be related to LAI using a logarithmic model derived from Beer’s Law. The approach has been successful in LAI mapping in boreal and temperate forests. In this study, we evaluated the logarithmic model and several GF proxies in tropical montane forests in southeastern Kenya. We used two discrete-return lidar datasets (max. scan angle ~16°) with different flying heights and pulse densities (5.4 and 2.6 pulses m–2). GF for the 0–15° zenith angle range (GF15) and effective LAI (Le) were estimated for 29 sample plots using digital hemispherical photography. Twenty-one plots were located in indigenous forests and eight plots in plantation forests. According to the results, GF15was best approximated by the proxies that included all canopy and ground return types (all echo cover index, ACI, root mean square error, RMSE = 0.050, bias = –0.003; Solberg’s cover index, SCI, RMSE = 0.057, bias = 0.002) although some saturation occurred when using data from the higher flight altitude. The results of theLemodelling propose that the logarithmic model needs to be fit separately for indigenous forest and plantations. Furthermore, the slope parameters of the models based on SCI suggest planophile (β≈ 1.6) and spherical (β≈ 2) leaf angle distribution for indigenous forests and plantations, respectively. We conclude that lidar cover indices based on all returns can estimate GF15in closed-canopy tropical forests but the detection of the smallest gaps can be limited by the scanner or scanning parameters. The application of the logarithmic model requires stratification in the structurally heterogeneous and multi-species forest areas asβshould be estimated separately for the different forest types. [ABSTRACT FROM AUTHOR]

Abstract: The prediction of tropical forest attributes using airborne laser scanning (ALS) is becoming attractive as an alternative to traditional field measurements. Area-based ALS inventories require a set of representative field plots from the study area, which may be difficult to obtain in tropical forests with limited accessibility. This study investigates the effect of sample-plot selection in Nepal, based on two accessibility factors: distance to road and degree of slope. The sparse Bayesian method was employed in the model to estimate above-ground biomass (AGB) with an independent validation dataset for model validation. Study findings showed that the sample plot distance and slope had a considerable effect on the accuracy of the AGB estimation, because the forest structure varied according to the level of accessibility. Thus, the field sample plots that are used in model construction should cover the full range of sample plot distances and slopes occurring within the area. [Copyright &y& Elsevier]

Forest canopy cover (C) is needed in forest area monitoring and for many ecological models. Airborne scanning lidar sensors can produce fairly accurateCestimates even without field training data. However, optical satellite images are more cost-efficient for large area inventories. Our objective was to use airborne lidar data to obtain accurate estimates ofCfor a set of sample plots in a boreal forest and to generalizeCfor a large area using a satellite image. The normalized difference vegetation index (NDVI) and reduced simple ratio (RSR) were calculated from the satellite image and used as predictors in the regressions. RSR, which combines information from the red, near-infrared, and shortwave infrared bands, provided the best performance in terms of absolute root mean square error (RMSE) (7.3%) in the training data. NDVI produced a markedly larger RMSE (10.0%). However, in an independent validation data set, RMSE increased (13.0–17.1%) because the systematic sample of validation plots contained more variation than the training plots. Our results are better than those reported earlier, which is probably explained by more consistentCestimates derived from the lidar. Our approach provides an efficient method for creatingCmaps for large areas. [ABSTRACT FROM AUTHOR]

Although simple geometrical shapes are commonly used to describe tree crowns, computational geometry enables calculation of the individual crown properties directly from airborne lidar point clouds. Our objective was to calculate crown volumes (CVs) using this technique and validate the results by comparing them with field-measured values and modelled ellipsoidal crowns. The CVs of standing trees were obtained by measuring the crown radii at different heights, integrating the obtained crown profiles as solids of revolution, and finally averaging the volumes obtained from the four separate profiles. With the lidar data, the CVs were extracted using 3D alpha shape and 3D convex hull techniques. Crown base heights (CBHs) were also estimated from the lidar data and used to exclude echoes from the understory, which was also done using field-based CBHs to exclude this error source. The results show that the field-measured CVs had a high correlation with lidar-based estimates (bestR2 = 0.83), but the lidar-based estimates were generally smaller than the field values. The best correspondence (root mean square difference (RMSD) = 45.0%, average difference = –24.7%) was obtained using the convex hull of the point data and field-measured CBH. The CBHs were consistently overestimated (RMSD = 37.3%; average difference = –20.0%), especially in spruces with long crowns. Thus using lidar-based CBH also increased the inaccuracy of the CV estimates. While the underestimation of CV is mainly explained by the inadequate number of echoes from the lower regions of the crowns, the CVs obtained from the lidar were better than those obtained with ellipsoids fitted by using general models for crown dimensions. The utility of the estimated CVs in the prediction of stem diameter is also demonstrated. [ABSTRACT FROM PUBLISHER]

Abstract: Remote sensing of forest canopy cover has been widely studied recently, but little attention has been paid to the quality of field validation data. Ecological literature has two different coverage metrics. Vertical canopy cover (VCC) is the vertical projection of tree crowns ignoring within-crown gaps. Angular canopy closure (ACC) is the proportion of covered sky at some angular range around the zenith, and can be measured with a field-of-view instrument, such as a camera. We compared field-measured VCC and ACC at 15° and 75° from the zenith to different LiDAR (Light Detection and Ranging) metrics, using several LiDAR data sets and comprehensive field data. The VCC was estimated to a high precision using a simple proportion of canopy points in first-return data. Confining to a maximum 15° scan zenith angle, the absolute root mean squared error (RMSE) was 3.7–7.0%, with an overestimation of 3.1–4.6%. We showed that grid-based methods are capable of reducing the inherent overestimation of VCC. The low scan angles and low power settings that are typically applied in topographic LiDARs are not suitable for ACC estimation as they measure in wrong geometry and cannot easily detect small within-crown gaps. However, ACC at 0–15° zenith angles could be estimated from LiDAR data with sufficient precision, using also the last returns (RMSE 8.1–11.3%, bias –6.1–+4.6%). The dependency of LiDAR metrics and ACC at 0–75° zenith angles was nonlinear and was modeled from laser pulse proportions with nonlinear regression with a best-case standard error of 4.1%. We also estimated leaf area index from the LiDAR metrics with linear regression with a standard error of 0.38. The results show that correlations between airborne laser metrics and different canopy field characteristics are very high if the field measurements are done with equivalent accuracy. [Copyright &y& Elsevier]

A new simple airborne method based on wide optics camera is developed for leaf area index (LAI) estimation in coniferous forests. The measurements are carried out in winter, when the forest floor is completely snow covered and thus acts as a light background for the hemispherical analysis of the images. The photos are taken automatically and stored on a laptop during the flights. The R2 value of the linear regression of the airborne and ground based LAI measurements was 0.89. [ABSTRACT FROM AUTHOR]

Angle count or relascope sampling has traditionally been used in forestry to estimate stand basal area. In this paper, we present an extension to the basal area relascope, the crown relascope, which differs from the normal relascope in that the relascope’s slot is very high and wide. We describe the theoretical basis of the instrument and present results from a field test in which a crown relascope with a basal area factor of 250 m2/ha (0.025) was used to estimate canopy cover of 73 sample plots in northern Finland. The crown relascope estimates had a root mean square difference of 9.3% and an average difference of –3.1% when compared with estimates obtained with the control method, line intersect sampling using the Cajanus tube. The results indicated that the crown relascope is a quick and fairly reliable instrument for canopy cover estimation, especially in relatively sparse forests where crown overlap is insignificant and visibility does not limit an efficient use of the instrument. [ABSTRACT FROM AUTHOR]

European aspen (Populus tremula L.) is a keystone species for biodiversity of boreal forests. Large-diameter aspens maintain the diversity of hundreds of species, many of which are threatened in Fennoscandia. Due to a low economic value and relatively sparse and scattered occurrence of aspen in boreal forests, there is a lack of information of the spatial and temporal distribution of aspen, which hampers efficient planning and implementation of sustainable forest management practices and conservation efforts. Our objective was to assess identification of European aspen at the individual tree level in a southern boreal forest using high-resolution photogrammetric point cloud (PPC) and multispectral (MSP) orthomosaics acquired with an unmanned aerial vehicle (UAV). The structure-from-motion approach was applied to generate RGB imagery-based PPC to be used for individual tree-crown delineation. Multispectral data were collected using two UAV cameras: Parrot Sequoia and MicaSense RedEdge-M. Tree-crown outlines were obtained from watershed segmentation of PPC data and intersected with multispectral mosaics to extract and calculate spectral metrics for individual trees. We assessed the role of spectral data features extracted from PPC and multispectral mosaics and a combination of it, using a machine learning classifier—Support Vector Machine (SVM) to perform two different classifications: discrimination of aspen from the other species combined into one class and classification of all four species (aspen, birch, pine, spruce) simultaneously. In the first scenario, the highest classification accuracy of 84% (F1-score) for aspen and overall accuracy of 90.1% was achieved using only RGB features from PPC, whereas in the second scenario, the highest classification accuracy of 86 % (F1-score) for aspen and overall accuracy of 83.3% was achieved using the combination of RGB and MSP features. The proposed method provides a new possibility for the rapid assessment of aspen occurrence to enable more efficient forest management as well as contribute to biodiversity monitoring and conservation efforts in boreal forests. [ABSTRACT FROM AUTHOR]

Very high point density laser scanning data from unmanned aerial systems (UAS) can be used to segment the trunks of individual trees. Such segmentation (e.g., individual trunk segmentation (ITS)) is useful, for example, in the estimation of diameter at breast height (DBH), which in turn is needed for the estimation of other tree and stand attributes, such as stem volume and diameter of the basal area median tree (DGM). In this paper, we assess the estimation of DBH directly from UAS LiDAR data in open and closed canopy conditions that represent the range of operational conditions encountered in boreal forests in Finland. We also compare trunk-based DBH estimates to corresponding estimates from individual tree crown segmentation (ITC) and fuse the results from the trunk- and crown-based estimates. The results showed that trunk segmentation performed slightly better than ITC in open canopy areas, whereas ITC performed better in closed canopy areas. The DBH prediction error was smaller for ITS (3.0 cm) than ITC (3.9 cm) when considering the trees that were recognized by both methods. We also conclude that a hybrid method, where both segmented tree trunks and tree crowns are fused, considerably increases the number of correctly segmented trees but does not decrease the prediction error associated with DGM compared to using either ITC or ITS individually. [ABSTRACT FROM AUTHOR]

• Bayesian inversion of PARAS model produced promising LAI retrieval accuracy. • Non-uniform priors improve inversion by counteracting optical saturation effects. • Landsat 8 OLI data showed better LAI retrieval accuracy than Sentinel-2 MSI. • We highlight benefits of uncertainty quantification in reflectance model inversion. The inversion of reflectance models is a generalizable tool to obtain estimates on forest biophysical parameters, such as leaf area index, with theoretically little information need from a study area, instead relying on the knowledge about physical processes in the forest radiation regime. The use of prior information can greatly improve the reflectance model inversion, however, the literature does not yet provide much information on the selection of priors and their influence on the inversion results. In this study, we used a Bayesian approach to invert the PARAS forest reflectance model and retrieve leaf area index from Sentinel-2 MSI and Landsat 8 OLI multispectral satellite images. The PARAS model is based on the theory of spectral invariants, which describes the influence of wavelength-independent parameters on forest radiative transfer. The Bayesian inversion approach is highly flexible, provides uncertainty quantification, and enables the explicit incorporation of prior knowledge into the inversion process. We found that the choice of prior information is crucial in inverting a forest reflectance model to predict leaf area index. Regularizing and informative priors for leaf area index strongly improved the predictions, relative to an uninformative prior, in that they counteracted the saturation effect of the optical signal occuring at high values for leaf area index. The predictions of leaf area index were more accurate for Landsat 8 than for Sentinel-2, due to potential inconsistencies in the visible bands of Sentinel-2 in our data, and the higher spectral resolution. [ABSTRACT FROM AUTHOR]

Multispectral light detection and ranging (LiDAR) instruments, such as Optech Titan, record intensities at multiple wavelengths and these intensities can be used for tree species prediction in the same way as multispectral image data. In this paper, our main objective was to compare the accuracy of tree species prediction in a boreal forest area using multispectral LiDAR, the 1064-nm wavelength channel (“unispectral LiDAR”), and unispectral LiDAR with auxiliary aerial image data. We also evaluated the effect of the widely used intensity range correction method. We classified the main tree species of field plots using linear discriminant analysis (LDA) and predicted the species-specific volume proportions (%) for Scots pine (Pinus sylvestris), Norway spruce (Picea abies), and broadleaved trees using the $k$ -nearest neighbor imputation. The effect of intensity correction on prediction errors for the dominant tree species was evaluated using optimal parameters derived from: 1) minimal intensity difference between flight lines; 2) parameters suggested by theory; and 3) uncorrected data. Although the range correction increased the classification accuracy slightly, it was observed to be ambiguous, and not consistent with theory for canopy echoes. Regardless, the intensity values were useful for the prediction of dominant tree species and species’ volume proportions. The results for the dominant tree species classification using multispectral LiDAR [overall accuracy (OA) 88.2%, kappa 0.79] were comparable to the use of unispectral LiDAR and aerial images (OA 89.1%, kappa 0.81). We conclude that the multispectral LiDAR may become a useful tool in operational species-specific forest inventories. [ABSTRACT FROM AUTHOR]

In-situ field measurements of sample plots are a major cost component in airborne laser scanning (ALS) based forest inventories. Field measurements on new inventory areas can be reduced by utilizing existing stand attribute models from former inventory areas. We constructed a nationwide model for stem volume, and examined seven different calibration scenarios using 22 inventory areas distributed evenly throughout Finland. These scenarios can be divided into three main categories: 1) calibration with additional predictor variables, 2) calibration with 200 geographically nearest sample plots, and 3) calibration with MS-NFI (Multi-source National Forest Inventory of Finland) volume at the target inventory area. Calibration with degree days, precipitation, and proportion of birch resulted in the greatest decrease in error rate of stem volume predictions. The mean of the root mean square errors (RMSE) among the 22 inventory areas decreased from 28.6% to 25.9%, and the standard deviation of RMSEs from 4.3% to 3.9% using three additional predictor variables. Correspondingly, the mean and standard deviation of absolute values of mean differences (|MD|) decreased from 8.3% to 5.6% and from 5.6% to 4.4%, respectively. All calibration scenarios decreased the error rate, especially the high |MDs| observed in the northern part of Finland. Calibration with sample plots from geographically nearest inventory areas was useful when there were sample plots that offered a good representation of the target area. MS-NFI based calibration was also a reasonable option if loggings and other inconsistencies between datasets could be detected and accounted for. [ABSTRACT FROM AUTHOR]

South East Asia region is a triple hotspot of carbon, biodiversity, and forest degradation. The latter is leading to heterogeneous forest landscapes with predominant and varying intensity of human modification. Optical Earth observation data offers the most feasible and accessible means of mapping and monitoring the forest cover heterogeneity in this region. Further, difficulties in ground data collection in this region present the need for more robust physically-based interpretation of satellite data which extends beyond simple empirical correlations of reflectance data with forest characteristics. We present an application of forest reflectance parameterization based on spectral invariants in lowland humid tropical forests of Malaysian Borneo. Our framework combined field measurements (hemispherical photos) of canopy structure, radiative transfer modelling at leaf (PROSPECT) and canopy scale (PARAS), and assessments using satellite data (Landsat-7 Enhanced Thematic Mapper Plus, ETM +). Field data were collected representing a gradient of forest degradation within a large-scale experimental landscape. Results showed limited ability of Landsat-class satellites to uncover the full gradient of forest degradation in this biome due to similarity in canopy structure as characterized by effective canopy cover, leaf area index, and angular profile of canopy gap fraction. Our findings, however, showed strong expression of leaf optical properties at canopy level. This supports the feasibility to remotely measure canopy biochemistry in dense humid tropical rainforests at medium spatial resolution using freely available Landsat-class satellites, and forthcoming hyperspectral satellite missions. [ABSTRACT FROM AUTHOR]

Understanding the influence of forest structure on forest albedo, and thus on the energy exchange between the forest and the atmosphere, is urgently needed in areas with large forest cover and active forest management. Fine resolution albedo retrievals enable quantifying the relationships between forest variables and albedo also in patchy landscapes, such as in the managed forests in Fennoscandia. In this study, field plot data, airborne laser scanning (ALS) data and high resolution satellite albedo retrievals from Landsat were used to investigate the main factors influencing forest albedo in Central Finland in midsummer. Tree species, forest structure and understory (ground) vegetation composition all influenced forest albedo. The tree species-specific models were estimated on a subpixel scale by utilizing information on the proportions of each species within a plot. Tree species considerably improved the albedo prediction when added to a model containing only a structural variable, whereas a further addition of the site fertility class as a proxy of understory vegetation composition only slightly improved the model. Albedo decreased with increasing volume of growing stock, but the decrease leveled off at high volumes. The albedo of plots with high volume was instead mainly governed by tree species and was the lowest for Norway spruce, intermediate for Scots pine and highest for broadleaved species. Norway spruce albedo decreased almost linearly with increasing mean tree height. ALS-derived canopy cover explained fairly well the variation in albedo in the visible region, but the total shortwave albedo was better predicted by ALS-derived tree height. [ABSTRACT FROM AUTHOR]

The terrestrial net ecosystem productivity (NEP) has increased during the past three decades, but the mechanisms responsible are still unclear. We analyzed 17 years (2001–2017) of eddy‐covariance measurements of NEP, evapotranspiration (ET) and light and water use efficiency from a boreal coniferous forest in Southern Finland for trends and inter‐annual variability (IAV). The forest was a mean annual carbon sink (252 [±42] gC m‐2a‐1), and NEP increased at rate +6.4–7.0 gC m‐2a‐1 (or ca. +2.5% a‐1) during the period. This was attributed to the increasing gross‐primary productivity GPP and occurred without detectable change in ET. The start of annual carbon uptake period was advanced by 0.7 d a‐1, and increase in GPP and NEP outside the main growing season contributed ca. one‐third and one‐fourth of the annual trend, respectively. Meteorological factors were responsible for the IAV of fluxes but did not explain the long‐term trends. The growing season GPP trend was strongest in ample light during the peak growing season. Using a multi‐layer ecosystem model, we showed that direct CO2 fertilization effect diminishes when moving from leaf to ecosystem, and only 30–40% of the observed ecosystem GPP increase could be attributed to CO2. The increasing trend in leaf‐area index (LAI), stimulated by forest thinning in 2002, was the main driver of the enhanced GPP and NEP of the mid‐rotation managed forest. It also compensated for the decrease of mean leaf stomatal conductance with increasing CO2 and LAI, explaining the apparent proportionality between observed GPP and CO2 trends. The results emphasize that attributing trends to their physical and physiological drivers is challenged by strong IAV, and uncertainty of LAI and species composition changes due to the dynamic flux footprint. The results enlighten the underlying mechanisms responsible for the increasing terrestrial carbon uptake in the boreal zone. [ABSTRACT FROM AUTHOR]

Most forest growth studies using airborne laser scanning (ALS) consider how the changes in forest attributes are observed in repeated ALS data acquisitions, but the prediction of future forest growth from ALS data is still a rarely discussed topic. This study examined the prediction of the periodic annual increment (PAI) of the width of tree rings over a period of 10 years. The requirement for this approach is that ALS data are acquired at the beginning of the growth period. This is followed by field measurements of growth by drilling after a given growth period. The PAI was modelled in terms of ALS metrics by using the principle of the area-based approach. The metrics related to intensity were particularly significant as predictors, whereas the effective leaf area index was not. The root-mean-square error (RMSE) of the predictions was slightly over 21%. Additional field information (soil type, management operations) improved the RMSE by 2.7 percentage units. [ABSTRACT FROM AUTHOR]

The article focuses on effects of numbers of observations and predictors for various model types on the performance of forest inventory with airborne laser scanning.

Abstract: In recent years, airborne LiDAR sensors have shown remarkable performance in the mapping of forest vegetation. This experimental study looks at LiDAR data at the scale of individual pulses to elucidate the sources behind interpulse variation in backscattering. Close-range photogrammetry was used for obtaining the canopy reference measurements at the ratio scale. The experiments illustrated different orientation techniques in the field, LiDAR acquisitions and photogrammetry in both leaf-on and leaf-off conditions, and two-waveform recording LiDAR sensors. The intrafootprint branch silhouettes in zenith-looking images, in which the camera, footprint, and LiDAR sensor were collinear, were extracted and contrasted with LiDAR backscattering. An enhanced planimetric match (refinement of strip matching) was achieved by shifting the pulses in a strip and searching for the maximal correlation between the silhouette and LiDAR intensity. The relative silhouette explained up to 80–90% of the interpulse variation. We tested whether accounting for the Gaussian spread of intrafootprint irradiance would improve the correlations, but the effect was blurred by small-scale geometric noise. Accounting for receiver gain variations in the Leica ALS60 sensor data strengthened the dependences. The size of the vegetation objects required for triggering a LiDAR observation was analyzed. We demonstrated the use of LiDAR pulses adjacent to canopy vegetation, which did not trigger a canopy echo, for canopy mapping. Pulses not triggering an echo constitute the complement to the actual canopy. We conclude that field photogrammetry is a useful tool for mapping forest canopies from below and that quantitative analysis is feasible even at the scale of single pulses for enhanced understanding of LiDAR observations from vegetation. [Copyright &y& Elsevier]

Abstract: Spectral invariants provide a novel approach for characterizing canopy structure in forest reflectance models and for mapping biophysical variables using satellite images. We applied a photon recollision probability (p) based forest reflectance model (PARAS) to retrieve leaf area index (LAI) from fine resolution SPOT HRVIR and Landsat ETM+ satellite data. First, PARAS was parameterized using an extensive database of LAI-2000 measurements from five conifer-dominated boreal forest sites in Finland, and mixtures of field-measured forest understory spectra. The selected vegetation indices (e.g. reduced simple ratio, RSR), neural networks and kNN method were used to retrieve effective LAI (L e ) based on reflectance model simulations. For comparison, we established empirical vegetation index-LAI regression models for our study sites. The empirical RSR–L e regression performed best when applied to an independent test site in southern Finland [RMSE 0.57 (24.2%)]. However, the difference to the best reflectance model based retrievals produced by neural networks was only marginal [RMSE 0.59 (25.1%)]. According to this study, the PARAS model provides a simple and flexible modelling tool for calibrating algorithms for LAI retrieval in conifer-dominated boreal forests. The advantage of PARAS is that it directly uses field measurements to parameterize canopy structure (LAI-2000, hemispherical photographs) and optical properties of foliage and understory. [Copyright &y& Elsevier]

Surface fuel load (SFL) constitutes one of the most significant fuel components and is used as an input variable in most fire behavior prediction systems. The aim of the present study was to investigate the potential of discrete-return multispectral Light Detection and Ranging (LiDAR) data to reliably predict SFL in a coniferous forest characterized by dense overstory and complex terrain. In particular, a linear regression analysis workflow was employed with the separate and combined use of LiDAR-derived structural and pulse intensity information for the load estimation of the total surface fuels and individual surface fuel types. Following a leave-one-out cross-validation (LOOCV) approach, the models developed from the different sets of predictor variables were compared in terms of their estimation accuracy. LOOCV indicated that the predictive models produced by the combined use of structural and intensity metrics significantly outperformed the models constructed with the individual sets of metrics, exhibiting an explained variance (R2) between 0.59 and 0.71 (relative Root Mean Square Error (RMSE) 19.3–37.6%). Overall, the results of this research showcase that both structural and intensity variables provided by multispectral LiDAR data are significant for surface fuel load estimation and can successfully contribute to effective pre-fire management, including fire risk assessment and behavior prediction in case of a fire event. [ABSTRACT FROM AUTHOR]

Accurate canopy base height (CBH) information is essential for forest and fire managers since it constitutes a key indicator of seedling growth, wood quality and forest health as well as a necessary input in fire behavior prediction systems such as FARSITE, FlamMap and BEHAVE. The present study focused on the potential of airborne LiDAR data analysis to estimate plot-level CBH in a dense uneven-aged structured forest on complex terrain. A comparative study of two widely employed methods was performed, namely the voxel-based approach and regression analysis, which revealed a clear outperformance of the latter. More specifically, the voxel-based CBH estimates were found to lack correlation with the reference data ( R 2 = 0.15 , r R M S E = 42.36 % ) while most CBH values were overestimated resulting in an r b i a s of − 17.52 % . On the contrary, cross-validation of the developed regression model showcased an R 2 , r R M S E and r b i a s of 0. 61 , 18.19 % and − 0.09 % respectively. Overall analysis of the results proved the voxel-based approach incapable of accurately estimating plot-level CBH due to vegetation and topographic heterogeneity of the forest environment, which however didn't affect the regression analysis performance. [ABSTRACT FROM AUTHOR]

The leaf area index (LAI) is a crucial structural parameter of forest canopies. Light Detection and Ranging (LiDAR) provides an alternative to passive optical sensors in the estimation of LAI from remotely sensed data. However, LiDAR-based LAI estimation typically relies on empirical models, and such methods can only be applied when the field-based LAI data are available. Compared with an empirical model, a physically-based model—e.g., the Beer–Lambert law based light extinction model—is more attractive due to its independent dataset with training. However, two challenges are encountered when applying the physically-based model to estimate LAI from discrete LiDAR data: i.e., deriving the gap fraction and the extinction coefficient from the LiDAR data. We solved the first problem by integrating LiDAR and hyperspectral data to transfer the LiDAR penetration ratio to the forest gap fraction. For the second problem, the extinction coefficient was estimated from tiled (1 km × 1 km) LiDAR data by nonlinearly optimizing the cost function of the angular LiDAR gap fraction and simulated gap fraction from the Beer–Lambert law model. A validation against LAI-2000 measurements showed that the estimates were significantly correlated to the reference LAI with an R2 of 0.66, a root mean square error (RMSE) of 0.60 and a relative RMSE of 0.15. We conclude that forest LAI can be directly estimated by the nonlinear optimization method utilizing the Beer–Lambert model and a spectrally corrected LiDAR penetration ratio. The significance of the proposed method is that it can produce reliable remotely sensed forest LAI from discrete LiDAR and spectral data when field-measured LAI are unavailable. [ABSTRACT FROM AUTHOR]

Airborne Light Detection and Ranging (LiDAR) information alone is insufficient for species-specific prediction of forest stand attributes, and therefore auxiliary optical image features (OIF) are commonly used to decrease the prediction errors associated with species-specific tree attributes. However, this requires collection and merging of two data sources, LiDAR and OIF, which increases the costs of the inventory. The recently introduced multispectral LiDAR (M-ALS) provides a potential single-sensor solution for obtaining species-specific information, as its multispectral intensity values resemble optical image data. Image point clouds (IPC) derived from aerial stereo images are another single-sensor option that provides both geometric and optical information. We compared two single-sensor options, M-ALS and IPC, with two LiDAR data sets (leaf-on and leaf-off) with auxiliary OIF, for the prediction of boreal tree species' volumes. In terms of root-mean-square error (RMSE) in the validation data, the LiDAR+OIF combination performed best (leaf-on RMSE: 33.3%; leaf-off RMSE 34.3%), followed by M-ALS + OIF (RMSE = 35.2%) and IPC + OIF (RMSE = 42.4%). The mean RMSE value associated with M-ALS increased to the same level (44.7%) as the IPC + OIF combination when optical image features were not included. Both IPC and M-ALS are potential single sensor solutions for forest inventories, but the use of both LiDAR and OIF provides the most accurate results. • Species-specific volumes predicted with multiple sensor configurations. • LiDAR with optical image features performed better than multispectral LiDAR alone. • IPC data performed markedly worse than LiDAR data in predicting total volume. • Single wavelength leaf-on LiDAR with image features performed the best. [ABSTRACT FROM AUTHOR]

Gap fraction measurements with digital a hemispherical camera were carried out in 2007–2016 in a Scots pine ( Pinus sylvestris L.) stand growing on transitional Sphagnum bog in Järvselja, Estonia. The measurements were done at the time when (1) the pines had maximum foliage, or (2) after autumn needle fall or before bud burst. Data for gap fraction calculation were extracted from camera raw files. For data recording, the sensor output signal maximum was kept according to the image brightness histogram within the interquartile range of the sensor dynamic range. A linear conversion method (LinearRatio) was used for image processing. The relative needle loss in autumn (20–30%) increased gap fraction by 0.02–0.05 at view zenith angles from 10° to 60° and decreased the mean plant area index from 1.86 to 1.68 after autumn litter fall. The measurement results were in good agreement with a theoretical gap fraction model simulation and with gap fraction estimates made from cover photos. The measurement and data processing protocol is appropriate for long-term monitoring in permanent sample plots. [ABSTRACT FROM AUTHOR]

The spectral invariants theory predicts that amount of radiation scattered by a canopy in a given wavelength is a simple function of the leaf albedo and a spectrally invariant parameter p. This study provides the first assessment of p parameter obtained from the Earth Reflector Type Index (ERTI) provided in Deep Space Climate Observatory (DSCOVR) Earth Polychromatic Imaging Camera (EPIC) Vegetation Earth System Data Record (VESDR) product. Our results suggest that ERTI parameter can possibly deliver information about photon recollision probability, albeit with a risk of lower accuracy, even for conditions that differ from the original assumptions in the spectral invariants theory (non-reflecting soil, dense canopies). We verified the expected link between ERTI p parameter in the EPIC VESDR product, foliage clumping, and canopy cover with empirical data. Importantly, this was done with study sites in Australia that included types of forests that are still relatively unexplored with spectral invariants theory. Our findings contribute to understanding product uncertainty and spatial representativeness of the in situ measurements from the Terrestrial Ecosystem Research Network (TERN) SuperSites in Australia. • First validation of ERTI p parameter from the EPIC VESDR product with in situ data. • ERTI p parameter linked to foliage clumping, canopy cover. • Spatially representative TERN sites, data suitable for satellite products validation. [ABSTRACT FROM AUTHOR]

During the growing season, the photosynthesis and growth of boreal forests are regulated by physiological responses to environmental factors. Physiological variations affect the spectral properties of leaves. Linking canopy-level spectral reflectance to leaf-level processes for monitoring forest seasonal physiology using satellite images is hindered by view and illumination effects and variations in canopy structure. To better understand the connection between the two structural levels, we used nine narrow-band vegetation indices (VIs) derived from Hyperion imagery to track the seasonal dynamics of boreal forest stands: the photochemical reflectance indices (PRI and \textPRI515) related to the xanthophyll cycle, the red edge (RE) index, the Maccioni (Macc) and the green normalized difference vegetation index related to chlorophyll concentration (Ca+b), the carotenoid simple ratio and Gitelson carotenoid concentration index related to carotenoid concentration (Cx+c), the normalized difference vegetation index related to fractional cover, and the plant senescence reflectance index related to the Cx+c/Ca+b ratio. As ground truth, we used measurements of exposed pine shoot light use efficiency (LUE) and photosynthesis. Over the study period (May to August), LUE and photosynthesis were best correlated with the chlorophyll VIs Macc and RE. Both indices also exhibited the lowest coefficient of variation in association with forest structure. PRI, on the other hand, was affected by canopy structure and observation geometry and was uncoupled from LUE during the growing season. Our findings demonstrate that the photosynthesis and productivity of boreal forests in the growing season are best tracked using VIs related to total pigment concentration (i.e., chlorophyll). [ABSTRACT FROM AUTHOR]

In this paper, we consider a computational method for detecting trees on the basis of airborne laser scanning (ALS) data. In the approach, locations, heights, and crown shapes of trees are tracked automatically by fitting multiple 3-D crown height models to ALS data of a field plot. The estimates are computed with an iterative reconstruction method based on Bayesian inversion paradigm. The formulation allows for utilizing prior information on tree shapes in the estimation. Here, the prior models are written based on field measurement data and allometric models for tree shapes. The feasibility of the approach is tested with ALS and field data from a managed boreal forest. The algorithm found 70.2% of the trees in the area, which is a clear improvement compared to a usual 2.5D crown delineation approach (53.1% of the trees detected). [ABSTRACT FROM AUTHOR]

Abstract: Clumping index, the measure of foliage grouping relative to a random distribution of leaves in space, is a key structural parameter of plant canopies that influences canopy radiation regimes and controls canopy photosynthesis and other land–atmosphere interactions. In this study, we retrieve the clumping index using the original 275m resolution data of the Multi-angle Imaging SpectroRadiometer (MISR) instrument over a set of sites representing diverse biomes and different canopy structures. Also for the first time, the MISR derived clumping index values are directly validated with both in-situ vertical profiles and seasonal trajectories of clumping index. Our results illustrate that MISR data with 275m allow clumping index estimates at much more pertinent scales (both spatial and temporal) than previous maps from Polarization and Directionality of Earth Reflectances (POLDER) and Moderate Resolution Imaging Spectroradiometer (MODIS) for modeling local carbon and energy fluxes. [Copyright &y& Elsevier]

Stem biomass is a fundamental component of the global carbon cycle that is essential for forest productivity estimation. Over the last few decades, Light Detection and Ranging (LiDAR) has proven to be a useful tool for accurate carbon stock and biomass estimation in various biomes. The aim of this study was to investigate the potential of multispectral LiDAR data for the reliable estimation of single-tree total and barkless stem biomass (TSB and BSB) in an uneven-aged structured forest with complex topography. Destructive and non-destructive field measurements were collected for a total of 67 dominant and co-dominant Abies borisii-regis trees located in a mountainous area in Greece. Subsequently, two allometric equations were constructed to enrich the reference data with non-destructively sampled trees. Five different regression algorithms were tested for single-tree BSB and TSB estimation using height (height percentiles and bicentiles, max and average height) and intensity (skewness, standard deviation and average intensity) LiDAR-derived metrics: Generalized Linear Models (GLMs), Gaussian Process (GP), Random Forest (RF), Support Vector Regression (SVR) and Extreme Gradient Boosting (XGBoost). The results showcased that the RF algorithm provided the best overall predictive performance in both BSB (i.e., RMSE = 175.76 kg and R2 = 0.78) and TSB (i.e., RMSE = 211.16 kg and R2 = 0.65) cases. Our work demonstrates that BSB can be estimated with moderate to high accuracy using all the tested algorithms, contrary to the TSB, where only three algorithms (RF, SVR and GP) can adequately provide accurate TSB predictions due to bark irregularities along the stems. Overall, the multispectral LiDAR data provide accurate stem biomass estimates, the general applicability of which should be further tested in different biomes and ecosystems. [ABSTRACT FROM AUTHOR]

Increasing use of airborne light detection and ranging (LiDAR) in forest inventories offers new possibilities to apply process-based forest models (PBM) in practice. We present a new approach, where a simplified PBM is run using inputs derived from the LiDAR data. The PBM was built by combining several existing models together, and it was tested with 52 Scots pine ( Pinus sylvestris L.) dominated sample plots in Finland with the LiDAR (PBM_LIDAR) and field (PMB_FIELD) inputs. The results were compared with empirical growth predictions (EM_FIELD) and field reference growth. LiDAR-based stand variables (mean height of tree and crown base and leaf area index) were, on average, well in line with the field measurements. Basal area growth was slightly underestimated with the PBM_LIDAR (bias 4.1%, root mean square prediction error (RMSPE, 26.7%) and overestimated with the PBM_FIELD (bias -10.2%, RMSPE 33.3%), the EM_FIELD being the least biased (bias -1.9%, RMSPE of 24.6%). The bias varied with stand age and fertility. The dependence on field reference growth was highest with EM_FIELD and PBM_LIDAR ( R2 = 0.47 and 0.34, respectively), and lowest with PBM_FIELD ( R2 = 0.18). Despite several development needs, the approach is promising for easy incorporation of canopy and weather data into forest growth predictions. [ABSTRACT FROM AUTHOR]

Abstract: Medium resolution leaf area index (LAI) products have usually been validated only at the peak of the growing season for boreal forest areas. In this study, we evaluated seasonal variations in MODIS Collection 5 (C5) LAI products for an evergreen coniferous forest area in southern Finland. A time series of field measurements of LAI and nine fine resolution satellite images (SPOT4 HRVIR, EO-1 Hyperion) for the snow free period in 2010 were used for developing empirical regression models for LAI estimation. We used two different reference LAI: the first one corresponded to an effective overstory LAI, and the second one was a clumping-corrected LAI that included both over- and understory vegetation. According to the results, MODIS LAI retrievals were closer to the effective LAI in the spring and autumn, when saturation of MODIS LAI main algorithm did not occur. On the other hand, in the summer, when saturation occurred, MODIS LAI was closer to the total LAI. Therefore, as a result of high LAI retrieved under the condition of saturation, the seasonal course of LAI was exaggerated in comparison to the reference LAI time series. Furthermore, MODIS LAI retrievals were unrealistically small before the leaf burst if considering the large proportion of evergreen coniferous forest in the study area. The results emphasize the need for multi-seasonal validation data sets in order to better understand the performance of LAI retrieval algorithms for evergreen coniferous forests throughout the snow free period. [Copyright &y& Elsevier]

Abstract: Methods for analyzing foliage nonrandomness by means of the TRAC instrument, digital hemispheric photography, and a gap fraction model are assessed at two RAMI (RAdiation transfer Model Intercomparison) mature stands in Järvselja, Estonia. The six different methods involve calculation of the canopy element clumping index, at scales coarser than that of a shoot. The major aim was to define the merits and limitations of the various methods. We conclude the gap size distribution and beyond-shoot clumping is very stable across the stands for the solar zenith angle range from 30° to 60°. Estimates based on the gap size distribution and the combination of gap size and logarithm methods performed the best while compared with an independent gap fraction model. We clarify the effect of the assumed leaf inclination angle distribution on gap size distribution and differences between estimates of beyond-shoot clumping. We show that the modified, gap-size distribution based method of Chen and Cihlar can provide reliable beyond-shoot clumping estimates without any a priori assumptions about the total gap fraction, segment length or the leaf inclination angle distribution. We also illustrate the changes in element clumping with measurement height. The compiled data extend the original parameter dataset to be used in the next phase of RAMI for different benchmark tests and reflectance modeling experiments, and contribute toward systematic validation efforts of radiative transfer models, operational algorithms, and field instruments, as promoted by the Committee on Earth Observation Satellites (CEOS). [ABSTRACT FROM AUTHOR]

Drone applications are becoming increasingly common in the arena of forest management and forest inventories. In particular, the use of photogrammetrically derived drone-based image point clouds (DIPC) in individual tree detection has become popular. Use of an area-based approach (ABA) in small areas has also been considered. However, in-situ field measurements of sample plots substantially increase the cost of small area forest inventories. Therefore, we examined whether small-scale forest management inventories could be carried out without local field measurements. We used nationwide and regional ABA models for stem volumes fitted with airborne laser scanning (ALS) data to predict stem volumes using corresponding metrics calculated from DIPC data. The stem volumes were predicted at the cell level (15 × 15 m) and aggregated to test plots (30 × 30 m). Height metrics for the dominant tree layer from the DIPC data showed strong correlations with similar metrics computed from the ALS data. The ALS-based models applied with DIPC metrics performed well, especially if the ABA model was fitted in the same geographical area (regional model) and the inventory units were disaggregated to coniferous and deciduous dominated stands using auxiliary information from Multi-source National Forest Inventory data (root mean square error at 30 × 30 m level was 13.1%). The corresponding root mean square error associated with the nationwide ABA model was 20.0% with an overestimation (mean difference 9.6%). • Airborne laser scanning based volume models were applied to drone image point clouds. • Stem volumes were predicted with low error rates without new field measurements. • Pre-classification to coniferous and deciduous dominated stands improved the results. [ABSTRACT FROM AUTHOR]