Explore the vast range of titles available.

Forests are the planet’s main CO2 filtering agent as well as important economical, environmental and social assets. Climate change is exerting an increased stress, resulting in a need for improved research methodologies to study their health, composition or evolution. Traditionally, information about forests has been collected using expensive and work-intensive field inventories, but in recent years unoccupied autonomous vehicles (UAVs) have become very popular as they represent a simple and inexpensive way to gather high resolution data of large forested areas. In addition to this trend, deep learning (DL) has also been gaining much attention in the field of forestry as a way to include the knowledge of forestry experts into automatic software pipelines tackling problems such as tree detection or tree health/species classification. Among the many sensors that UAVs can carry, RGB cameras are fast, cost-effective and allow for straightforward data interpretation. This has resulted in a large increase in the amount of UAV-acquired RGB data available for forest studies. In this review, we focus on studies that use DL and RGB images gathered by UAVs to solve practical forestry research problems. We summarize the existing studies, provide a detailed analysis of their strengths paired with a critical assessment on common methodological problems and include other information, such as available public data and code resources that we believe can be useful for researchers that want to start working in this area. We structure our discussion using three main families of forestry problems: (1) individual Tree Detection, (2) tree Species Classification, and (3) forest Anomaly Detection (forest fires and insect Infestation).

Vegetation biodiversity in mountainous regions is controlled by altitudinal gradients and their corresponding microclimate. Higher temperatures, shorter snow cover periods, and high variability in the precipitation regime might lead to changes in vegetation distribution in mountains all over the world. In this study, we evaluate vegetation distribution along an altitudinal gradient (1334–1667 m.a.s.l.) in the Zao Mountains, northeastern Japan, by means of alpha diversity indices, including species richness, the Shannon index, and the Simpson index. In order to assess vegetation species and their characteristics along the mountain slope selected, fourteen 50 m × 50 m plots were selected at different altitudes and scanned with RGB cameras attached to Unmanned Aerial Vehicles (UAVs). Image analysis revealed the presence of 12 dominant tree and shrub species of which the number of individuals and heights were validated with fieldwork ground truth data. The results showed a significant variability in species richness along the altitudinal gradient. Species richness ranged from 7 to 11 out of a total of 12 species. Notably, species such as Fagus crenata, despite their low individual numbers, dominated the canopy area. In contrast, shrub species like Quercus crispula and Acer tschonoskii had high individual numbers but covered smaller canopy areas. Tree height correlated well with canopy areas, both representing tree size, which has a strong relationship with species diversity indices. Species such as F. crenata, Q. crispula, Cornus controversa, and others have an established range of altitudinal distribution. At high altitudes (1524–1653 m), the average shrubs’ height is less than 4 m, and the presence of Abies mariesii is negligible because of high mortality rates caused by a severe bark beetle attack. These results highlight the complex interactions between species abundance, canopy area, and altitude, providing valuable insights into vegetation distribution in mountainous regions. However, species diversity indices vary slightly and show some unusually low values without a clear pattern. Overall, these indices are higher at lower altitudes, peak at mid-elevations, and decrease at higher elevations in the study area. Vegetation diversity indices did not show a clear downward trend with altitude but depicted a vegetation composition at different altitudes as controlled by their surrounding environment. Finally, UAVs showed their significant potential for conducting large-scale vegetation surveys reliably and in a short time, with low costs and low manpower.

Unmanned Aerial Vehicles (UAV) are becoming an essential tool for evaluating the status and the changes in forest ecosystems. This is especially important in Japan due to the sheer magnitude and complexity of the forest area, made up mostly of natural mixed broadleaf deciduous forests. Additionally, Deep Learning (DL) is becoming more popular for forestry applications because it allows for the inclusion of expert human knowledge into the automatic image processing pipeline. In this paper we study and quantify issues related to the use of DL with our own UAV-acquired images in forestry applications such as: the effect of Transfer Learning (TL) and the Deep Learning architecture chosen or whether a simple patch-based framework may produce results in different practical problems. We use two different Deep Learning architectures (ResNet50 and UNet), two in-house datasets (winter and coastal forest) and focus on two separate problem formalizations (Multi-Label Patch or MLP classification and semantic segmentation). Our results show that Transfer Learning is necessary to obtain satisfactory outcome in the problem of MLP classification of deciduous vs evergreen trees in the winter orthomosaic dataset (with a 9.78% improvement from no transfer learning to transfer learning from a a general-purpose dataset). We also observe a further 2.7% improvement when Transfer Learning is performed from a dataset that is closer to our type of images. Finally, we demonstrate the applicability of the patch-based framework with the ResNet50 architecture in a different and complex example: Detection of the invasive broadleaf deciduous black locust (Robinia pseudoacacia) in an evergreen coniferous black pine (Pinus thunbergii) coastal forest typical of Japan. In this case we detect images containing the invasive species with a 75% of True Positives (TP) and 9% False Positives (FP) while the detection of native trees was 95% TP and 10% FP.

Insect outbreaks are a recurrent natural phenomenon in forest ecosystems expected to increase due to climate change. Recent advances in Unmanned Aerial Vehicles (UAV) and Deep Learning (DL) Networks provide us with tools to monitor them. In this study we used nine orthomosaics and normalized Digital Surface Models (nDSM) to detect and classify healthy and sick Maries fir trees as well as deciduous trees. This study aims at automatically classifying treetops by means of a novel computer vision treetops detection algorithm and the adaptation of existing DL architectures. Considering detection alone, the accuracy results showed 85.70% success. In terms of detection and classification, we were able to detect/classify correctly 78.59% of all tree classes (39.64% for sick fir). However, with data augmentation, detection/classification percentage of the sick fir class rose to 73.01% at the cost of the result accuracy of all tree classes that dropped 63.57%. The implementation of UAV, computer vision and DL techniques contribute to the development of a new approach to evaluate the impact of insect outbreaks in forest.

Understanding how forests recover after severe disturbances is essential for developing effective management strategies that promote stable forest regeneration. Disturbances are particularly significant in transition zones such as treelines, which are highly sensitive to climate change. In the subalpine treeline ecotone of Zao Mountains (northeastern Japan), a severe double-pest infestation devastated the Abies mariesii forest, triggering a treeline retreat of nearly 400 m. Prior to the infestation, the stand density was estimated at 3135 (based on the sum of standing living and dead standing trees and fallen trees detected in the orthomosaics generated for the year 2019). Of these, 3023 were standing trees (of which 2787 were dead). By 2025, the number of standing trees had declined to 2472 (18.2% reduction), a significant development for Abies seedlings, which appeared to establish in decaying fallen logs. In order to evaluate whether this disturbance has permanent or temporary effects, high-resolution unmanned aerial vehicle (UAV) imagery was collected annually over the study area, resulting in six orthomosaics from 2019 to 2025 (2020 data unavailable) for continuous and precise forest monitoring. Analysis of the monitored area revealed that in the 6.9 ha study site, entirely covered by 1.2–1.3 m tall sasa vegetation (Sasa kurilensis), the number of new young trees increased from 60 (2019) to 119 in 2025. These younger trees were mainly located near surviving mature trees. Sasa vegetation did not exert a negative effect on regeneration but instead appeared to function as a strong wind protection, facilitating Abies seedling growth. In conclusion, fallen logs and sasa vegetation appeared to have a positive effect on fir regeneration as suggested by the increasing number of young trees observed over time. The bark beetle outbreak functioned as a forest stand-replacing disturbance, where the subalpine fir forest at the treeline is expected to regenerate naturally within the coming decades.

Invasive blueberry species endanger the sensitive environment of wetlands and protection laws call for management measures. Therefore, methods are needed to identify blueberry bushes, locate them, and characterise their distribution and properties with a minimum of disturbance. UAVs (Unmanned Aerial Vehicles) and image analysis have become important tools for classification and detection approaches. In this study, techniques, such as GIS (Geographical Information Systems) and deep learning, were combined in order to detect invasive blueberry species in wetland environments. Images that were collected by UAV were used to produce orthomosaics, which were analysed to produce maps of blueberry location, distribution, and spread in each study site, as well as bush height and area information. Deep learning networks were used with transfer learning and unfrozen weights in order to automatically detect blueberry bushes reaching True Positive Values (TPV) of 93.83% and an Overall Accuracy (OA) of 98.83%. A refinement of the result masks reached a Dice of 0.624. This study provides an efficient and effective methodology to study wetlands while using different techniques.

Deep Learning (DL) has become popular due to its ease of use and accuracy, with Transfer Learning (TL) effectively reducing the number of images needed to solve environmental problems. However, this approach has some limitations which we set out to explore: Our goal is to detect the presence of an invasive blueberry species in aerial images of wetlands. This is a key problem in ecosystem protection which is also challenging in terms of DL due to the severe imbalance present in the data. Results for the ResNet50 network show a high classification accuracy while largely ignoring the blueberry class, rendering these results of limited practical interest to detect that specific class. Moreover, by using loss function weighting and data augmentation results more akin to our practical application, our goals can be obtained. Our experiments regarding TL show that ImageNet weights do not produce satisfactory results when only the final layer of the network is trained. Furthermore, only minor gains are obtained compared with random weights when the whole network is retrained. Finally, in a study of state-of-the-art DL architectures best results were obtained by the ResNeXt architecture with 93.75 True Positive Rate and 98.11 accuracy for the Blueberry class with ResNet50, Densenet, and wideResNet obtaining close results.

In recent decades, forests have experienced an increasing trend in the number of pest outbreaks worldwide, apparently driven by strong annual variability in precipitation, higher air temperatures, and strong winds. Pest outbreaks have negative ecological, economic, and environmental impacts on forest ecosystems, such as reduced biodiversity, carbon sequestration, and overall forest health. Traditional monitoring methods of these disturbances, while accurate, are time-consuming and limited in scope. Remote sensing, particularly UAV (Unmanned Aerial Vehicle)-based technologies, offers a precise and cost effective alternative for monitoring forest health. This study evaluates the temporal and spatial progression of bark beetle damage in a fir-dominated forest in the Zao Mountains, Japan, using UAV RGB imagery and DL (Deep Learning) models (YOLO - You Only Look Ones), over a four-year period (2021–2024). Trees were classified into six health categories: Healthy, Light Damage, Medium Damage, Heavy Damage, Dead, and Fallen. The results revealed a significant decline in healthy trees, from 67.4% in 2021 to 25.6% in 2024, with a corresponding increase in damaged and dead trees. Light damage emerged as a potential early indicator of forest health decline. The DL model achieved an accuracy of 74.9% to 82.8%. The results showed the effectiveness of DL in detecting severe damage but highlighted that challenges in distinguishing between healthy and lightly damaged trees still remain. The study highlights the potential of UAV-based remote sensing and DL for monitoring forest health, providing valuable insights for targeted management interventions. However, further refinement of the classification methods is needed to improve accuracy, particularly in the precise detection of tree health categories. This approach offers a scalable solution for monitoring forest health in similar ecosystems in other subalpine areas of Japan and the world.

Snow, especially in mountainous regions, plays a major role acting as a quasi-reservoir, as it gradually releases fresh water during the melting season and thereby fills rivers, lakes, and groundwater aquifers. For vegetation and irrigation, the timing of the snowmelt is crucial. Therefore, it is necessary to understand how snowmelt varies under different local conditions. While differences in slope aspect and vegetation (individually) were linked to differences in snow accumulation and ablation, this study connects the two and describes their influence on the soil moisture response to snowmelt. This research focuses on the catchment of the “Brunnenkopfhütte” (BKH) in Bavaria, southern Germany, where an automatic weather station (AWS) has operated since 2016. In addition, soil temperature and moisture monitoring systems in the surrounding area on a south aspect slope on an open field (SO), on a south aspect slope in the forest (SF), and a north aspect slope in the forest (NF) have operated since 2020. On snow-free days in winter, the soil temperature at the SF site was on average 1 °C lower than on the open site. At the NF site, this soil temperature difference increased to 2.3 °C. At the same time, for a 1 °C increase in the air temperature, the soil temperature increases by 0.35 °C at the NF site. In addition, at this site, snow cover disappeared approximately one week later than on the south aspect slopes. Snow cover at the SF site disappeared even earlier than at the SO site. Finally, a significant difference in the soil moisture response was found between the sites. While the vegetation cover dampens the magnitude of the soil moisture increases, at the NF site, no sharp increases in soil moisture were observed.

The effects of the present global climate change appear more pronounced in high latitudes and alpine regions. Transitions zones, such as the southern fringe of the boreal region in northern Mongolia, are expected to experience drastic changes as a result. This area is dry and cold with forests forming only on the north-facing slopes of hills and grasslands distributing on the south-facing slopes, making it difficult for continuous forests to exist. However, in the Hovsgol Lake Basin, there is a vast continuous pure forest of Siberian larch (Larix sibirica). In other words, the lake water thawing/freezing process may have created a unique climatic environment that differs with the climate of the adjacent Darhad Basin, where no lake exists. Thus, in order to compare the effect of the thawing/freezing dynamics of lake water and the active layer on the thermal regime at each basin, respectively, temperatures were simultaneously measured. The Darhad Basin has similar latitude, topography, area, and elevation conditions. As expected, the presence of the lake affected the annual temperature amplitude, as it was 60% of that in the Darhad Basin. The difference in the seasonal freeze–thaw cycles of the lake and the active layer caused a significant difference in the thermal regime, especially in winter.