During December 2020, a crowdsourcing campaign to understand what has been driving tropical forest loss during the past decade was undertaken. For 2 weeks, 58 participants from several countries ...reviewed almost 115 K unique locations in the tropics, identifying drivers of forest loss (derived from the Global Forest Watch map) between 2008 and 2019. Previous studies have produced global maps of drivers of forest loss, but the current campaign increased the resolution and the sample size across the tropics to provide a more accurate mapping of crucial factors leading to forest loss. The data were collected using the Geo-Wiki platform ( www.geo-wiki.org ) where the participants were asked to select the predominant and secondary forest loss drivers amongst a list of potential factors indicating evidence of visible human impact such as roads, trails, or buildings. The data described here are openly available and can be employed to produce updated maps of tropical drivers of forest loss, which in turn can be used to support policy makers in their decision-making and inform the public.
The sustainable management of urban green areas requires clear and efficient protocols for measuring the biometric properties of tree vegetation. Specifically, operational in situ sampling solutions ...are essential to inventory forked (multi-stemmed) trees. This study aimed to assess the efficiency of two different sampling protocols for mean tree diameter at breast height (DBH) measurement of forked urban trees. The protocols were tested on a dataset of 76 forked trees, each having more than three stems and sampled in urban areas of Kyiv, Ukraine. First, we tested the efficiency of mean tree DBH estimations using measurements of randomly selected one, two, or three stems (random sampling, or RSM). Second, we examined different combinations of the thinnest, thickest, and average stems (identified visually) for each tree to estimate mean tree DBH (targeted sampling, or TSM). The distributions of mean tree DBH and root mean square errors (RMSE) were utilized to compare the utility of the two approaches. The TSM of three stems (the thinnest, thickest, and average) provided the highest accuracy of mean tree DBH estimation (RMSE% = 6.3% of the mean), compared to the RSM (RMSE% = 12.1%). The TSM of the four thickest stems demonstrated the overestimation of mean tree DBH for forked trees with five or more stems. Accurate mean tree DBH estimates can be derived with negligible systematic errors applying the RSM over a large number of measured trees. However, these estimates will not likely match the measurements from previous inventories due to random stem selection. We recommend using the TSM with measuring three specific stems as a balanced solution in terms of estimation accuracy, bias, and time costs.
Several global high-resolution built-up surface products have emerged over the last five years, taking full advantage of open sources of satellite data such as Landsat and Sentinel. However, these ...data sets require validation that is independent of the producers of these products. To fill this gap, we designed a validation sample set of 50 K locations using a stratified sampling approach independent of any existing global built-up surface products. We launched a crowdsourcing campaign using Geo-Wiki ( https://www.geo-wiki.org/ ) to visually interpret this sample set for built-up surfaces using very high-resolution satellite images as a source of reference data for labelling the samples, with a minimum of five validations per sample location. Data were collected for 10 m sub-pixels in an 80 × 80 m grid to allow for geo-registration errors as well as the application of different validation modes including exact pixel matching to majority or percentage agreement. The data set presented in this paper is suitable for the validation and inter-comparison of multiple products of built-up areas.
Spatially explicit and consistent mapping of forest biomass is one of the key tasks towards full and appropriate accounting of carbon budgets and productivity potentials at different scales. Landsat ...imagery coupled with terrestrial-based data and processed using modern machine learning techniques is a suitable data source for mapping of forest components such as deadwood. Using relationships between deadwood biomass and growing stock volume, here we indirectly map this ecosystem compartment within the study area in northern Ukraine. Several machine learning techniques were applied: Random Forest (RF) for the land cover and tree species classification task,
-Nearest Neighbours (
-NN) and Gradient Boosting Machines (GBM) for the deadwood imputation purpose. Land cover (81.9%) and tree species classification (78.9%) were performed with a relatively high level of overall accuracy. Outputs of deadwood biomass mapping using
-NN and GBM matched quite well (8.4 ± 2.3 t·ha
(17% of the mean) vs. 8.1 ± 1.7 t·ha
(16% of the mean), respectively mean ±
deadwood biomass stock), indicating a strong potential of ensemble boosters to predict forest biomass in a spatially explicit manner. The main challenges met in the study were related to the limitations of available ground-based data, thus showing the need for national statistical inventory implications in Ukraine.
Thirty-five years after the accident, large forest areas in the Chernobyl Exclusion Zone still contain huge amounts of radionuclides released from the Chernobyl Nuclear Power Plant Unit 4 in April ...1986. An assessment of the radiological and radioecological consequences of persistent radioactive contamination and development of remediation strategies for Chernobyl forests imply acquiring comprehensive data on their contamination levels and dynamics of biomass inventories. The most accurate forest inventory data can be obtained in ground timber cruises. However, such cruises in radioactive contaminated forest ecosystems in the Chernobyl Exclusion Zone result in radiation exposures of the personnel involved, which means the need for development of the remote sensing methods. The purpose of this study is to analyze the applicability and limitations of the photogrammetric method for the remote large-scale monitoring of aboveground biomass inventories. Based on field measurements, we estimated the biomass inventories in 31 Scots pine stands including both artificial plantations and natural populations. The stands differed significantly in age (from a few years in natural populations to 115 years in the oldest plantation), productivity (from 0.4 to 19.8 kg m−2), mean height (from 4.1 to 36 m), and other parameters. Photogrammetric data were obtained from the same stands using unmanned aerial vehicle (UAV). These data were then processed using two approaches to derive the canopy height model (CHM) parameters which were tested for correlation with the aboveground biomass inventories. In the first approach, we found that the inventories correlated well with the mean value of CHM of the site (R2 = 0.79). In the second approach, the total aboveground biomass was approximated by a function of the average height of trees detected at the site and the total crown projection area (R2 = 0.78). Among other local parameters, the total crown projection area was identified as the major factor impacting the accuracy of the aboveground biomass inventory estimates from the UAV survey data in both approaches. In the dense stands with the high total crown projections areas (more than 0.90), the average relative deviations of the UAV-based aboveground biomass estimates from the results of the field measurements were close to 0, which means the adequate accuracy of the UAV surveys data for radioecological monitoring purposes. The relative deviations of the UAV-based estimates in both approaches increased in the stands consisting of separated groups of trees, which indicates potential limitation of the approaches and need for their further development.
•Applicability of aerial surveys for biomass estimate in Chernobyl forests was tested.•Canopy height models of the studied stands were derived from photogrammetric datasets.•Biomass inventories correlated well with the average values of canopy height models.•Biomass inventories also correlated with the parameters of individual trees.•Accuracy of biomass estimates was high in the dense forest stands.
Following the nuclear disaster of 1986, forests have established throughout the abandoned agricultural landscapes within Chernobyl Exclusion Zone (ChEZ). However, they are yet to be monitored ...properly. Their biometrical parameters need a robust assessment considering climate change mitigation potential and wildfire-induced risks. To predict basal area (BA) and growing stock volume (GSV) of these forests using spatially explicit approach, we utilized Sentinel-2 satellite data and three types of machine learning models (k-Nearest Neighbors (k-NN), Random Forest (RF) and Gradient Boosting Machine (GBM)). Root mean square error among all models ranged between 5.2 m
2
ha
−1
(49% of the mean) and 7.2 m
2
ha
−1
(71% of the mean), derived for BA by the GBM and k-NN models, respectively. While total and mean estimates of forest attributes were quite similar within an entire ChEZ, GBM approach outperformed other methods by predicting GSV more precisely when compared to local reference data. At the same time, k-NN approach has shown better performance in terms of preserving the initial empirical distribution and semivariation patterns. We concluded that k-NN method should be used for the spatial predictions of forest attributes, however, with a specific focus given on the training data set quality and profound model validation.
Questions
Nuclear power is under increasing consideration in many countries because it is a low‐carbon strategy to satisfy growing energy demands. Yet, the long‐term environmental impacts of nuclear ...accidents remain unclear. Here we asked how ionizing radiation affects tree regeneration and forest development after the Chernobyl nuclear accident. We hypothesized that high levels of 137Cs contamination in the soil: (a) inhibit tree establishment; (b) accelerate structural development (i.e., facilitation of an early differentiation of tree sizes); while (c) prolonging the dominance of early‐seral deciduous communities (because of an elevated susceptibility of conifers to ionizing radiation).
Location
Chernobyl Exclusion Zone, Ukraine.
Methods
We sampled 103 plots on former agricultural lands in the Chernobyl Exclusion Zone that were abandoned after the accident in 1986.
Results
Contamination had no significant effect on the stem density of forests established on former agricultural lands (p = 0.769). Structural development was not accelerated by radioactive contamination (p > 0.191), but we did find weak indications that the presence of tree regeneration was reduced by high radiation levels (p = 0.054). Tree species composition did not vary significantly with contamination (p = 0.250). Individual Scots pine trees did, however, experience a considerably higher proportion of deformed stems when contamination levels were high (p = 0.009).
Conclusions
Our analyses confirm negative effects of radioactive contamination on the individual tree health of Scots pine, yet early stand development in the Chernobyl Exclusion Zone was largely insensitive to different levels of radiation. As wildfires threaten to remobilize and redistribute radionuclides stored in the growing forests of the Chernobyl Exclusion Zone, our findings have potential implications for human health. We conclude that forest dynamics is a key element for assessing the long‐term risk at nuclear accident sites and requires intensified research and monitoring.
We examined how forest development on abandoned agricultural fields was affected by radioactive contamination in the Chernobyl Exclusion Zone. Early stand development in the Chernobyl Exclusion Zone was largely insensitive to different levels of radiation. Yet contamination above background levels increased the probability of growth deformations in Pinus sylvestris individuals and reduced tree regeneration.
Abstract
Stem taper equations are crucial for forest management allowing to reliably estimate merchantable wood volume. Their main benefit is the ability to predict stem diameters at a certain height ...of the stem. Ukraine has recently adopted European Union standards for round wood classification, which prompted the necessity to model stem taper and updates all reference data to conform with the new standards. This study is a systematic attempt to develop a set of taper equations for the most common forest tree species in the Polissia and Forest steppe of Ukraine. For this purpose, we used a data set of 1994 sample trees representing eight tree species collected on 238 sample plots. The Kozak A. (2004, My last words on taper equations. For. Chron. 80, 507–515) model was chosen to fit the taper equations. To characterize the variability in stem shape among tree species, mixed-effect models were calibrated for this equation. In this model, random-effect parameters were selected based on their coefficients of variation through a bootstrapping process. This is a novel feature we suggest for the process of calibrating taper models. The Kozak A. (2004, My last words on taper equations. For. Chron. 80, 507–515) equation showed a good performance in predicting diameters outside bark and estimating the total stem volume. Our mixed-effect modelling approach accurately characterizes the variation in stem form for different tree species based on adequately chosen random-effect parameters. The stem volumes derived from the developed taper models were compared with existing volume equations outputs (divergence up to 0.5 per cent). A deviation up to 5 per cent was found between the values of fitted and observed cylindrical form factors for the studied tree species. We expect that our taper equations will complement the future steps towards the development of reliable merchantable volume distribution models for the main tree species in the forests of flat land Ukraine, thus, contributing to transparency, reliability and sustainability of forest management and markets in Europe.
A wide range of UAV systems used for forest research requires unified approaches to data collection. The research aims to determine the optimal parameters for UAV data collection to obtain accurate ...information about stands, considering the cost of resources for its collection. The process of collecting remote sensing data consisted of nine combinations divided into three levels of overlap and three levels of spatial resolution (survey altitude) and changing the degree of filtering of a dense point cloud during image processing. Individual tree detectingin the stand was performed using the R programming language and the ForestTools package. The results of the assessment of the dependence of the radius of tree crowns on their height were used to set the parameters of the variable filter function for finding local maxima for Scots pine stands. Errors in the identification of treetops were estimated using the F-score. The identified heights were compared with the field data of the ground survey. The proportion of classified digital elevation model DEM in the dense point cloud was reduced from the total area of the test site using images of 4.1 cm/pix spatial resolution (150 m survey altitude). The study presents the results of assessing the impact of spatial resolution of optical images collected from UAVs and their overlap on the results of measurements of stands parameters. It is determined that a photogrammetric survey with input images with a longitudinal overlap of less than 90% is not appropriate for the study of forest areas due to the impossibility of aligning all images. The results of the assessment of tree accounting in the stand showed that it is most appropriate to use images with a spatial resolution of up to 3.3 cm/pix (120 m survey altitude), otherwise, the proportion of missed treetops increases. Reducing the spatial resolution of remote sensing data leads to an increase in errors in determining the height of individual trees, and the average heights of the experimental plots had the same trend. Given the combination of the assessed factors, it is not recommended to use images with a spatial resolution of more than 3.3 cm/pix for forestry research due to increased errors in the individual tree detection and tree height determination. The results obtained can be used to select data collection parameters for research on Scots pine stands to assess their growing stock and phytomass
This study investigated the precision of measuring the height of trees using different methods. The paper evaluates the possibilities of using the stereophotogrammetric method to determine tree ...height indicators using unmanned aerial vehicles (UAV) in the conditions of a mature pine stand. The study compares the results of measuring the height of Scots pine trees with altimeters and height indicators determined from remote sensing data obtained using UAVs. In total, the study investigated six diverse ways to measure the height of growing trees. Experimental data on the height of the model trees were collected by three different altimeters (hand-held ground instruments) and the Phantom 4 Pro UAV. The use of UAVs involved optical capture and data collection using on-board equipment. Methods for determining the height of trees based on the results of processing data collected by quadcopter attachments were used. Specifically, the authors of this paper used the method of measuring the height of trees from a point cloud based on one-way vertical survey of model trees and calculating a digital crown height model (CHM) based on aerial photography of horizontal spans over a tree stand. The results of mathematical analysis of the conducted studies demonstrate the highest precision of the method using CHM to determine the height of growing trees. The value of the average random error in measuring the height of model trees using CHM was under 2%. The next most precise method of determining tree height was the TruPulse 360B laser-optical device, which demonstrated the highest precision among height meters. The use of the TruPulse 360b for ground-based measurements and the CHM method (based on UAV optical imaging data) yielded better results that meet the height precision standards for industrial inventory. Methods for determining the height of trees based on optical survey data from UAVs can be used for survey, inventory, forest management, and other works related to forestry and monitoring changes in forest ecosystems.