Forest biomass is an essential indicator for monitoring the Earth’s ecosystems and climate. It is a critical input to greenhouse gas accounting, estimation of carbon losses and forest degradation, assessment of renewable energy potential, and for developing climate change mitigation policies such as REDD+, among others. Wall-to-wall mapping of aboveground biomass (AGB) is now possible with satellite remote sensing (RS). However, RS methods require extant, up-to-date, reliable, representative and comparable in situ data for calibration and validation. Here, we present the Forest Observation System (FOS) initiative, an international cooperation to establish and maintain a global in situ forest biomass database. AGB and canopy height estimates with their associated uncertainties are derived at a 0.25 ha scale from field measurements made in permanent research plots across the world’s forests. All plot estimates are geolocated and have a size that allows for direct comparison with many RS measurements. The FOS offers the potential to improve the accuracy of RS-based biomass products while developing new synergies between the RS and ground-based ecosystem research communities.

Since 24 February 2022, Ukraine has experienced full-scale military aggression initiated by the Russian Federation. The war has had a major negative impact on vegetation cover of war-affected regions. We explored interactions between pre-war forest management and the impacts of military activities in three of the most forested Ukrainian areas of interest (AOI), affected by the war. These were forests lying between Kharkiv and Luhansk cities (AOI ‘East’), forests along the Dnipro River delta (AOI ‘Kherson’), and those of the Chornobyl Exclusion Zone (AOI CEZ). We used Sentinel satellite imagery to create damaged forest cover masks for the year 2022. We mapped forests with elevated fire hazard, which was defined as a degree of exposure to the fire-supporting land use (mostly an agricultural land, a common source of ignitions in Ukraine). We evaluated the forest disturbance rate in 2022, as compared to pre-war rates. We documented significant increases in non-stand replacing disturbances (low severity fires and non-fire disturbances) for all three of the AOIs. Damaged forest cover varied among the AOIs (24,180 ± 4,715 ha, or 9.3% ± 1.8% in the ‘East’ AOI; 7,293 ± 1,925 ha, or 15.7% ± 4.1% in the ‘Kherson’ AOI; 7,116 ± 1,274 ha, or 5.0% ± 0.9% in the CEZ AOI). Among the forests damaged in 2022, the ‘Kherson’ AOI will likely have the highest proportion of an area with elevated fire hazard in the coming decades, as compared to other regions (89% vs. 70% in the ‘East’ and CEZ AOIs respectively). Future fire risks and extensive war-related disturbance of forest cover call for forest management to develop strategies explicitly addressing these factors.

Climate change continues to threaten forests and their ecosystem services while substantially altering natural disturbance regimes. Land cover changes and consequent management entail discrepancies in carbon sequestration provided by forest ecosystems and its accounting. Currently there is a lack of sufficient and harmonized data for Ukraine that can be used for the robust and spatially explicit assessment of forest provisioning and regulation of ecosystem services. In the frame of this research, we established an experimental polygon (area 45 km²) in Northern Ukraine aiming at estimating main forest carbon stocks and fluxes and determining the impact caused by natural disturbances and harvest for the study period of 2010–2015. Coupled field inventory and remote sensing data (RapidEye image for 2010 and SPOT 6 image for 2015) were used. Land cover classification and estimation of biomass and carbon pools were carried out using Random Forest and k-Nearest Neighbors (k-NN) method, respectively. Remote sensing data indicates a ca. 16% increase of carbon stock, while ground-based computations have shown only a ca. 1% increase. Net carbon fluxes for the study period are relatively even: 5.4 Gg C·year⁻¹ and 5.6 Gg C C·year⁻¹ for field and remote sensing data, respectively. Stand-replacing wildfires, as well as insect outbreaks and wind damage followed by salvage logging, and timber harvest have caused 21% of carbon emissions among all C sources within the experimental polygon during the study period. Hence, remote sensing data and non-parametric methods coupled with field data can serve as reliable tools for the precise estimation of forest carbon cycles on a regional spatial scale. However, featured land cover changes lead to unexpected biases in consistent assessment of forest biophysical parameters, while current management practices neglect natural forest dynamics and amplify negative impact of disturbances on ecosystem services.

For 34 years since the 1986 nuclear disaster, the Chernobyl Exclusion Zone (ChEZ) landscapes have been protected with very limited human interventions. Natural afforestation has largely occurred throughout the abandoned farmlands, while natural disturbance regimes, which dominantly include wildfires, have become more frequent and severe in the last years. Here, we utilize the dense time series of Landsat satellite imagery (1986–2020) processed by using the temporal segmentation algorithm LandTrendr in order to derive a robust land cover and forest mask product for the ChEZ. Additionally, we carried out an analysis of land cover transitions on the former farmlands. The Random Forest classification model developed here has achieved overall accuracies of 80% (using training data for 2017) and 89% on a binary “forest/non-forest” validation (using data from 1988). The total forest cover area within the ChEZ has increased from 41% (in 1986) to 59% (in 2020). This forest gain can be explained by the afforestation that has occurred in abandoned farmlands, which compensates for forest cover losses due to large fire events in 1992, 2015–2016, and 2020. Most transitions from open landscapes to dense forest cover occurred after the year 2000 and are possibly linked to past forest management practices. We conclude that a consistent forest strategy, with the aid of remote monitoring, is required to efficiently manage new forests in the ChEZ in order to retain their ecosystem functions and to ensure sustainable habitats.

Wildfires in the Chornobyl Exclusion Zone (CEZ) and other radioactively contaminated areas threaten human health and well-being with the potential to resuspend radionuclides. Wildfire behavior simulation is a necessary tool to examine the efficiency of fuel treatments in the CEZ, but it requires systematically updated maps of fuel types and canopy metrics. The objective of this study was to demonstrate an effective approach for mapping fuel types, canopy height (CH), and canopy cover (CC) in territories contaminated by radionuclides using Landsat time series (LTS) and Global Ecosystem Dynamics Investigation (GEDI) LiDAR observations. We combined LTS and GEDI data to map fuel types and canopy metrics used in wildfire simulations within the CEZ. Our classification model showed an adequate overall accuracy (75%) in mapping land covers and associated fuel types. The phenology metrics extracted from LTS reliably distinguished spectrally similar vegetation types (such as grasslands and croplands) which exhibit different flammability through the year. We also predicted a suite of relative heights metrics and CC at Landsat 30-m pixel level (R² = 0.23-0.26) using the nearest neighbor technique. The imputed maps adequately captured the dynamics of CH and CC in the CEZ after recent large wildfires occurred in 2015, 2020, and 2022. Thus, we illustrate a LTS processing approach to produce wall-to-wall maps of canopy characteristics that are important for wildfire simulations. We conclude that continuous updating of land cover and canopy fuel data is crucial to ensure relevant fire management of radioactively contaminated landscapes and support local decision-making.

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 m² ha⁻¹ (49% of the mean) and 7.2 m² 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.