A Sentinel-2 machine learning dataset for tree species classification in Germany

Earth system science data, Journal Year: 2025, Volume and Issue: 17(2), P. 351 - 367

Published: Feb. 3, 2025

Abstract. We present a machine learning dataset for tree species classification in Sentinel-2 satellite image time series of bottom-of-atmosphere reflectance. It is geared towards training classifiers but less suitable validating the resulting maps. The based on German National Forest Inventory 2012 as well analysis-ready imagery computed using Framework Operational Radiometric Correction Environmental monitoring (FORCE) processing pipeline. From data, we extracted positions, filtered 387 775 trees upper canopy layer, and automatically corresponding reflectance from L2A images. These are labeled with species, which allows pixel-wise tasks. Furthermore, provide auxiliary information such approximate position, year possible disturbance events, or diameter at breast height. Temporally, spans years July 2015 to end October 2022, approx. 75.3 million data points 48 3 groups 13.8 observations non-tree backgrounds. Spatially, it covers whole Germany. available following DOI (Freudenberg et al., 2024): https://doi.org/10.3220/DATA20240402122351-0.

Language: Английский

---

From single trees to country-wide maps: Modeling mortality rates in Germany based on the Crown Condition Survey

Forest Ecology and Management, Journal Year: 2024, Volume and Issue: 568, P. 122081 - 122081

Published: July 5, 2024

Most years in the period from 2018 to 2022 have been exceptionally dry Central Europe. In Germany's forests, this long-lasting drought has caused unprecedented tree mortality. Systematic ground-based surveys, such as annual Crown Condition Survey, provide information on vitality status of different species and their mortality rates. However, models are needed be able map spatial patterns for each based cause-effect relationships derived field observations. study, logistic regression were used identify most important drivers Germany. For purpose, dead surviving trees Survey combined with a large set potential predictor variables domains climate, topography, soil, land cover deposition. After feature selection, evaluated using area under curve (AUC) statistic. Norway spruce (Picea abies; AUC = 0.9) showed by far greatest increase mortality, country-wide average observed predicted rates approaching almost 10% per year 2020 2022, much higher at regional level. Much was explained climatic water balance driest summer previous years. The other main also clear responses conditions. case European beech (Fagus sylvatica; 0.94) Pedunculate Sessile oak (Quercus robur petraea; 0.88), peaks time series stayed below 1%. these broadleaved species, more dependent range site conditions, i.e., soil topography. Scots pine (Pinus sylvestris; 0.76), which rate peaked 1.2% 2020, given could explain only lesser degree than species. prediction produce maps species-specific temporal 100-m resolution, covering all 1998 2022. visualize over time. regions western central Germany, seriously affected dieback can clearly identified. presented risk assessment, forest planning, providing decision support practitioners.

Language: Английский

---

Assessing experimental silvicultural treatments enhancing structural complexity in a central European forest – BEAST time‐series analysis based on Sentinel‐1 and Sentinel‐2

Remote Sensing in Ecology and Conservation, Journal Year: 2024, Volume and Issue: 10(4), P. 531 - 550

Published: April 3, 2024

Abstract Assessing the dynamics of forest structure complexity is a critical task in times global warming, biodiversity loss and increasing disturbances order to ensure resilience forests. Recent studies on emphasize essential functions deadwood accumulation diversification light conditions for enhancement structural complexity. The implementation an experimental patch‐network managed broad‐leaved forests within Germany enables standardized analysis various aggregated distributed treatments characterized through diverse structures. To monitor enhanced as seasonal trend components, dense time‐series from high spatial resolution imagery Sentinel‐1 (Synthetic‐Aperture Radar, SAR) Sentinel‐2 (multispectral) are analyzed decomposition models (BEAST, Bayesian Estimator Abrupt change, Seasonal change Trend). Based several statistics comprehensive catalog spectral indices, metrics ( n = 84) 903) calculated at patch‐level. Metrics best identifying treatment event assessed by point dates probability scores. Heterogeneity VH NMDI (Normalized Multi‐band Drought Index) capture most accurately, with clear advantages identification treatments. In addition, structures downed or no can be characterized, well more complex standing structures, such snags habitat trees. conclude, complementary sensors have potential assess complexities, thus supporting continuous monitoring habitats functioning over time.

Language: Английский

---

Identification of damage severity in Fraxinus excelsior L. trees caused by ash dieback using multisensory and multitemporal UAV imagery

Forest Ecology and Management, Journal Year: 2025, Volume and Issue: 585, P. 122660 - 122660

Published: March 22, 2025

Language: Английский

---

Potential of Earth Observation to Assess the Impact of Climate Change and Extreme Weather Events in Temperate Forests—A Review

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(12), P. 2224 - 2224

Published: June 19, 2024

Temperate forests are particularly exposed to climate change and the associated increase in weather extremes. Droughts, storms, late frosts, floods, heavy snowfalls, or changing climatic conditions such as rising temperatures more erratic precipitation having an increasing impact on forests. There is urgent need better assess impacts of extreme events (EWEs) temperate Remote sensing can be used map at multiple spatial, temporal, spectral resolutions low cost. Different approaches forest assessment offer promising methods for a broad analysis EWEs. In this review, we examine potential Earth observation assessing EWEs by reviewing 126 scientific papers published between 1 January 2014 31 2024. This study provides comprehensive overview sensors utilized, spatial temporal resolution studies, their distribution, thematic focus various abiotic drivers resulting responses. The indicates that multispectral, non-high-resolution timeseries were employed most frequently. A predominant proportion studies droughts. all instances EWEs, dieback prevailing response, whereas trends, phenology shifts account largest share response categories. detailed in-depth differentiation implies area-wide have so far barely distinguished effects different species level.

Language: Английский

---

Detailed validation of large-scale Sentinel-2-based forest disturbance maps across Germany

Forestry An International Journal of Forest Research, Journal Year: 2024, Volume and Issue: unknown

Published: July 11, 2024

Abstract Monitoring forest areas with satellite data has become a vital tool to derive information on disturbances in European forests at large scales. An extensive validation of generated maps is essential evaluate their potential and limitations detecting various disturbance patterns. Here, we present the results for four study Germany using Sentinel-2 from 2018 2022. We apply time series filtering method map annual larger than 0.1 ha based spectral clustering change magnitude. The presented part research design precursor national German monitoring system. In this context, are used estimate affected timber volume related economic losses. To better understand thematic accuracies reliability area estimates, performed an independent product 20 sets embedded our comprising total 11 019 sample points. collected reference datasets expert interpretation high-resolution aerial imagery, including dominant tree species, cause, severity level. Our achieves overall accuracy 99.1 ± 0.1% separating disturbed undisturbed forest. This mainly indicative forest, as that class covers 97.2% area. For class, user’s 84.4 2.0% producer’s 85.1 3.4% similar indicate estimated accurately. However, 2022, observe overestimation extent, which attribute high drought stress year leading false detections, especially around edges. varies widely among seems severity, patch size. User’s range 31.0 8.4% 98.8 1.3%, while 60.5 37.3% 100.0 0.0% across sets. These variations highlight single local set not representative region diversity patterns, such Germany. emphasizes need assess large-scale products many different possible, cover sizes, severities, causes.

Language: Английский

---

Individual tree detection and crown delineation in the Harz National Park from 2009 to 2022 using mask R–CNN and aerial imagery

ISPRS Open Journal of Photogrammetry and Remote Sensing, Journal Year: 2024, Volume and Issue: 13, P. 100071 - 100071

Published: July 5, 2024

Forest diebacks pose a major threat to global ecosystems. Identifying and mapping both living dead trees is crucial for understanding the causes implementing effective management strategies. This study explores efficacy of Mask R–CNN automated forest dieback monitoring. The method detects individual trees, delineates their crowns, classifies them as alive or dead. We evaluated approach using aerial imagery canopy height models in Harz Mountains, Germany, region severely affected by dieback. To assess model's ability track changes over time, we applied it images from three separate flight campaigns (2009, 2016, 2022). evaluation considered variations acquisition dates, cameras, post-processing techniques, image tilting. were analyzed based on detected trees' number, spatial distribution, height. A comprehensive accuracy assessment demonstrated R–CNN's robust performance, with precision scores ranging 0.80 0.88 F1-scores 0.91. These results confirm generalize across diverse conditions. While minor observed between 2009 period 2016 2022 witnessed substantial dieback, 64.57% loss trees. Notably, taller appeared be particularly affected. highlights potential valuable tool It enables efficient detection, delineation, classification providing data informed practices.

Language: Английский

---

Forest disturbance detection in Central Europe using transformers and Sentinel-2 time series

Remote Sensing of Environment, Journal Year: 2024, Volume and Issue: 315, P. 114475 - 114475

Published: Oct. 24, 2024

Language: Английский

---

Remote Sensing of Forests in Bavaria: A Review

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(10), P. 1805 - 1805

Published: May 20, 2024

In recent decades, climatic pressures have altered the forested landscape of Bavaria. Widespread loss trees has unevenly impacted entire state, which 37% is covered by forests (5% more than national average). 2018 and 2019—due in large part to drought subsequent insect infestations—more tree-covered areas were lost Bavaria any other German state. Moreover, annual crown condition survey revealed a decreasing trend tree vitality since 1998. We conducted systematic literature review regarding remote sensing total, 146 scientific articles published between 2008 2023. While 88 studies took place Bavarian Forest National Park, only five publications whole Outside park, remaining 2.5 million hectares forest are understudied. The most commonly studied topics related bark beetle infestations (24 papers); however, few papers focused on drivers infestations. majority utilized airborne data, while utilizing spaceborne data multispectral; types under-utilized- particularly thermal, lidar, hyperspectral. recommend future both spatially broaden investigations state or scale increase temporal acquisitions together with contemporaneous situ data. Especially understudied response climate, catastrophic disturbances, regrowth species composition, phenological timing, sector management. utilization forestry uptake results among stakeholders remains challenge compared heavily European countries. An integral economy tourism sector, also vital for climate regulation via atmospheric carbon reduction land surface cooling. Therefore, monitoring centrally important attaining resilient productive forests.

Language: Английский

---

Unravelling key factors of forest condition: Multidimensional assessment in Mediterranean pine ecosystems

Forest Ecology and Management, Journal Year: 2024, Volume and Issue: 578, P. 122487 - 122487

Published: Dec. 30, 2024

Language: Английский

---