Non-destructive estimation of individual tree biomass: Allometric models, terrestrial and UAV laser scanning

Remote Sensing of Environment, Journal Year: 2022, Volume and Issue: 280, P. 113180 - 113180

Published: Aug. 5, 2022

Calibration and validation of aboveground biomass (AGB) products retrieved from satellite-borne sensors require accurate AGB estimates across hectare scales (1 to 100 ha). Recent studies recommend making use non-destructive terrestrial laser scanning (TLS) based techniques for individual tree estimation that provide unbiased predictors. However, applying these large sites landscapes remains logistically challenging. Unoccupied aerial vehicle (UAV-LS) has the potential address this through collection high density point clouds many hectares, but on data been challenging so far, especially in dense tropical canopies. In study, we investigated how TLS UAV-LS can be used purpose by testing different modelling strategies with availability framework requirements. The study included four forested three biomes: temperate, wet tropical, savanna. At each site, coincident campaigns were conducted. Diameter at breast height (DBH) estimated clouds. Individual was ≥170 trees per site quantitative structure (QSM), treated as best available, estimate absence direct, destructive measurements. automatically segmented using a shortest-path algorithm full 3D cloud. Predictions evaluated terms root mean square error (RMSE) population bias, latter being absolute difference between total sample QSM predicted AGB. application global allometric scaling models (ASM) local scale modalities, i.e., field-inventory light detection ranging LiDAR metrics, resulted prediction errors range reported studies, relatively bias. adjustment factors should considered translate modalities. When calibrating models, DBH confirmed strong predictor AGB, useful when field inventories. combination derived metrics non-parametric generally produced RMSE, very low bias ≤5% starting 55 training samples. hectares reduced fieldwork time. Overall, contributes exploitation scale, relevant calibration space-borne missions targeting estimation.

Language: Английский

---

From UAV to PlanetScope: Upscaling fractional cover of an invasive species Rosa rugosa

Journal of Environmental Management, Journal Year: 2023, Volume and Issue: 336, P. 117693 - 117693

Published: March 11, 2023

Invasive plant species pose a direct threat to biodiversity and ecosystem services. Among these, Rosa rugosa has had severe impact on Baltic coastal ecosystems in recent decades. Accurate mapping monitoring tools are essential quantify the location spatial extent of invasive support eradication programs. In this paper we combined RGB images obtained using an Unoccupied Aerial Vehicle, with multispectral PlanetScope map R. at seven locations along Estonian coastline. We used RGB-based vegetation indices 3D canopy metrics combination random forest algorithm thickets, obtaining high accuracies (Sensitivity = 0.92, specificity 0.96). then presence/absence maps as training dataset predict fractional cover based derived from constellation Extreme Gradient Boosting (XGBoost). The XGBoost yielded prediction (RMSE 0.11, R2 0.70). An in-depth accuracy assessment site-specific validations revealed notable differences between study sites (highest 0.74, lowest 0.03). attribute these various stages invasion density thickets. conclusion, UAV is cost-effective method highly heterogeneous ecosystems. propose approach valuable tool extend local geographical scope assessments into wider areas regional evaluations.

Language: Английский

---

Innovative and practical tools for monitoring and assessing biodiversity status and impacts of multiple human pressures in marine systems

Environmental Monitoring and Assessment, Journal Year: 2024, Volume and Issue: 196(8)

Published: July 4, 2024

Language: Английский

---

Object-Based Wetland Vegetation Classification Using Multi-Feature Selection of Unoccupied Aerial Vehicle RGB Imagery

Remote Sensing, Journal Year: 2021, Volume and Issue: 13(23), P. 4910 - 4910

Published: Dec. 3, 2021

Wetland vegetation is an important component of wetland ecosystems and plays a crucial role in the ecological functions environments. Accurate distribution mapping dynamic change monitoring are essential for conservation restoration. The development unoccupied aerial vehicles (UAVs) provides efficient economic platform classification. In this study, we evaluated feasibility RGB imagery obtained from DJI Mavic Pro classification at species level, with specific application to Honghu, which listed as international importance. A total ten object-based image analysis (OBIA) scenarios were designed assess contribution five machine learning algorithms accuracy, including Bayes, K-nearest neighbor (KNN), support vector (SVM), decision tree (DT), random forest (RF), multi-feature combinations feature selection implemented by recursive elimination algorithm (RFE). overall accuracy kappa coefficient compared determine optimal method. main results follows: (1) RF showed best performance among algorithms, 89.76% 0.88 when using 53 features (including spectral (RGB bands), height information, indices, texture features, geometric features) (2) model constructed only poor results, 73.66% 0.70. By adding VIs, construct layer layer, was improved 8.78%, 3.41%, 2.93%, 0.98%, respectively, demonstrating importance combinations. (3) different types not equal, information most classification, followed indices. (4) RFE effectively reduced number original 36, generating subset based on result (RF-RFE) had scenarios, provided 90.73%, 0.97% higher than without selection. illustrate that combination UAV-based OBIA approach straightforward, yet powerful, high-precision spite limited information. Compared satellite data or UAVs equipped other sensors, cameras more cost convenient mapping.

Language: Английский

---

Accurate mapping of Brazil nut trees (Bertholletia excelsa) in Amazonian forests using WorldView-3 satellite images and convolutional neural networks

Ecological Informatics, Journal Year: 2021, Volume and Issue: 63, P. 101302 - 101302

Published: April 16, 2021

Language: Английский

---

Global application of an unoccupied aerial vehicle photogrammetry protocol for predicting aboveground biomass in non‐forest ecosystems

Remote Sensing in Ecology and Conservation, Journal Year: 2021, Volume and Issue: 8(1), P. 57 - 71

Published: July 7, 2021

Non-forest ecosystems, dominated by shrubs, grasses and herbaceous plants, provide ecosystem services including carbon sequestration forage for grazing, are highly sensitive to climatic changes. Yet these ecosystems poorly represented in remotely sensed biomass products undersampled situ monitoring. Current global change threats emphasize the need new tools capture non-forest at appropriate scales. Here we developed deployed a protocol photogrammetric height using unoccupied aerial vehicle (UAV) images test its capability delivering standardized measurements of across globally distributed field experiment. We assessed whether canopy inferred from UAV photogrammetry allows prediction aboveground (AGB) low-stature plant species conducting 38 surveys over 741 harvested plots sample 50 species. found mean was strongly predictive AGB species, with median adjusted R2 0.87 (ranging 0.46 0.99) error leave-one-out cross-validation 3.9%. Biomass per-unit-of-height similar within but different among, functional types. that reconstructions were wind speed not sun elevation during surveys. demonstrated our approach produced generalizable growth forms environmental settings yielded accuracies as good those obtained approaches. demonstrate can deliver accurate estimates wide range dynamic heterogeneous ecosystems. Many academic land management institutions have technical capacity deploy approaches extents 1-10 ha-1. Photogrammetric could much-needed information required calibrate validate vegetation models satellite-derived essential understand vulnerable understudied non-forested around globe.

Language: Английский

---

Drone imagery to create a common understanding of landscapes

Landscape and Urban Planning, Journal Year: 2022, Volume and Issue: 228, P. 104571 - 104571

Published: Sept. 15, 2022

Language: Английский

---

A Survey on Monitoring of Wild Animals during Fires Using Drones

Fire, Journal Year: 2022, Volume and Issue: 5(3), P. 60 - 60

Published: April 29, 2022

Forest fires occur for natural and anthropogenic reasons affect the distribution, structure, functioning of terrestrial ecosystems worldwide. Monitoring their impacts on is an essential prerequisite effectively managing this widespread environmental problem. With development information technologies, unmanned aerial vehicles (drones) are becoming increasingly important in remote monitoring environment. One main applications drone technology related to nature observation wild animals. Unmanned thought be best solution detecting forest fires. There methods wildfires using drones with fire- and/or smoke-detection equipment. This review aims study possibility large animals during It was established that order use monitor even small groups fires, effective sensing technologies critical temperature conditions required, which can provided not only by sensors used, but also adapted software image recognition.

Language: Английский

---

Evidence on the efficacy of small unoccupied aircraft systems (UAS) as a survey tool for North American terrestrial, vertebrate animals: a systematic map

Environmental Evidence, Journal Year: 2023, Volume and Issue: 12(1)

Published: Feb. 13, 2023

Small unoccupied aircraft systems (UAS) are replacing or supplementing occupied and ground-based surveys in animal monitoring due to improved sensors, efficiency, costs, logistical benefits. Numerous UAS sensors available have been used various methods. However, justification for selection methods not typically offered published literature. Furthermore, existing reviews do adequately cover past current applications monitoring, nor their associated UAS/sensor characteristics environmental considerations. We present a systematic map that collects consolidates evidence pertaining of animals.

Language: Английский

---

A feature enhancement framework for landslide detection

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2023, Volume and Issue: 124, P. 103521 - 103521

Published: Oct. 27, 2023

Accurate landslide detection is essential for disaster mitigation and relief. In this study, we develop a feature enhancement framework that integrates attention multiscale mechanisms with U-Net (AMU-Net) detection. The has four steps. First, the module in convolutional block enhances response of landslides when extracting high-level representations. Second, skip connection captures more contextual information concatenating fine coarse features. Third, architecture encodes mapping decodes semantics Fourth, shifted window was applied to enhance receptive field pixels prediction process, which reduced errors boundaries. Besides, explored effect random split regional methods on model training. upper reach Jinsha River, data unoccupied aerial vehicle (UAV) images digital surface (DSM) were prepared design experiments considers disparities between UAV satellite remote sensing. controlled reported mean Intersection over Union (mIoU) proposed AMU-Net achieved 0.797, 2% higher than other models. Furthermore, visualized maps revealed method can effectively restrain irrelevant responses backgrounds capture features from various fields. Comparative studies all above proved superiority

Language: Английский

---