Dr. Lars Waser

Rates of tree mortality are increasing globally, with implications for forests and climate. Yet, how and why these trends vary globally remain unknown. Developing a comprehensive assessment of global tree mortality will require systematically integrating data from ground-based long-term forest monitoring with large-scale remote sensing. We surveyed the metadata from 466 865 forest monitoring plots across 89 countries and five continents using questionnaires and discuss the potential to use these to estimate tree mortality trends globally. Our survey shows that the area monitored has increased steadily since 1960, but we also identify many regions with limited ground-based information on tree mortality. The integration of existing ground-based forest inventories with remote sensing and modelling can potentially fill those gaps, but this requires development of technical solutions and agreements that enable seamless flows of information from the field to global assessments of tree mortality. A truly global monitoring effort should promote fair and equitable collaborations, transferring funding to and empowering scientists from less wealthy regions. Increasing interest in forests as a natural climate solution, the advancement of new technologies and world-wide connectivity means that now a global monitoring system of tree mortality is not just urgently needed but also possible.

Forest edges represent the transition zone (ecotone) between the forest interior and the surrounding open land. Due to their great ecological importance, the value of assessing the structure of forest edges has been recognized. In Switzerland, for example, forest edge structure is assessed during field surveys of the Swiss National Forest Inventory (NFI). However, these assessments are time consuming and limited to sample plots. Publicly available countrywide airborne laser scanning (ALS) data, in contrast, offers possibilities to retrieve forest edge structure information over large spatial extents. In this study, we derived five metrics from ALS point clouds, namely the canopy height variability; ratios of the areas of the three edge components, i.e. shrub belt, shelterbelt and forest layer; the sky-view fraction; the shelterbelt slope; and the front density of the forest edge. These metrics describe the three-dimensional edge structure and therefore could enhance existing NFI edge metrics, which focus on two-dimensional structure characteristics. An expert assessment of the ALS edge metrics demonstrated the ability of the defined set of edge metrics to capture ecologically relevant and indicative characteristics of forest edges. Understanding this relationship between edge metrics and ecological functions is a prerequisite if ALS metrics are to be integrated into NFIs. We subsequently used the ALS edge metrics to group 284 forest edge into three classes with respect to their structural complexity using k-means clustering. The results indicated that the structual complexity was low for 173, medium for 46 and high for 65 forest edges, respectively. Applied to countrywide ALS data sets, our approach allows to retrieve area-wide, spatially continuous information on the forest edge conditions, and, if multitemporal ALS data is available, to monitor the development of the forest edges.

Despite decades of development, the uptake of remote sensing-based information products in the forestry sector is still lagging behind in central and southern Europe. This may partly relate to a mismatch of the developed remote sensing products and the requirements of potential users. Here, we present the results of a questionnaire survey in which we questioned 355 forest practitioners from eight central and southern European countries. We aimed to learn about forest practitioners' technical requirements for four remote sensing-based information products, including information on tree species, canopy height, wood volume/biomass, and forest disturbances. We asked for practitioners’ preferences with respect to thematic and spatial detail as well as the maximal acceptable error and the temporal frequency with which the information layers would be needed. We then examined whether the education, age, and professional background affect the requirements. Preferences with respect to spatial and thematic detail were comparably diverse while more homogenous patterns could be observed for demands with respect to errors and temporal frequency. Our results indicate that for some information products such as canopy height maps, existing remote sensing technology, and workflows can match all demands of practitioners. Remotely sensed information on forest disturbances partly fulfils the demands of the practitioners while for products related to tree species and wood volume/biomass the level of thematic detail and the accuracy of the products demanded by practitioners in central and southern Europe is not yet fully matched. We found no statistically significant differences between the demographic groups examined. The findings of this study improve our understanding of matches and mismatches of the technical requirements of practitioners for remote sensing-based information products.

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

Vermehrt auftretenden Waldveränderungen begegnet man mit einer Vielzahl an fernerkundungsbasierten Produkten wie Waldschadenskarten. Für die Nutzung dieser Produkte ist es wichtig, deren Anwendbarkeit für einen bestimmten Verwendungszweck richtig einschätzen zu können. Dabei helfen Angaben zur Genauigkeit, die aus der Validierung mit Referenzdaten abgeleitet werden. Welche gängigen Validierungsverfahren und Genauigkeitsmaßzahlen genutzt werden, was sie aussagen und warum sie essenziell sind, wird in diesem Artikel erklärt.

The tree species composition (TSC) reflects a forest's tree species diversity and is relevant for forest planning, biodiversity conservation, and forest resources management. Yet, accurate information on tree species composition at landscape scale is largely missing, especially for mixed forests and remote areas. One reason being that mapping tree species is time-consuming, and costly, especially in mixed forests and remote areas. Here we develop a robust method for mapping TSC in a mixed temperate forest. Based on forest inventory plots and considering the frequency of dominant tree species in the inventory dataset, five species groups were defined: pure oriental beech, mixed oriental beech, pure common hornbeam, mixed common hornbeam, and mixed deciduous. The classification is based on three-year time series data of Landsat-8 (L8) and Sentinel-2 (S2) derived spectral-temporal features (STMs) and vegetation indices within the long-term, seasonal, and monthly time scales. Model performances of three Machine Learning (ML) algorithms, Random Forest (RF), Support Vector Machine (SVM), and Classification and Regression Tree (CART) were compared and revealed different classification accuracies (overall accuracies (OAs) between ∼ 70 % and 86 %). Highest OA was obtained using SVM regardless of the classification dataset (STMs and satellite time series). The comparisons between different time scales indicated that with both L8 and S2 time series the seasonal STMs produced higher accuracies than monthly and long-term STMs with S2 outperforming L8 across all time scales and with all tested ML algorithms. We conclude that the freely available satellite time series, spectral-temporal features, and ML algorithms are favourable for accurate TSC mapping.

Mangrove forests, as an essential component of the coastal zones in tropical and subtropical areas, provide a wide range of goods and ecosystem services that play a vital role in ecology. Mangroves are globally threatened, disappearing, and degraded. Consequently, knowledge on mangroves distribution and change is important for effective conservation and making protection policies. Developing remote sensing data and classification methods have proven to be suitable tools for mapping mangrove forests over a regional scale. Here, we scrutinized and compared the performance of pixel-based and object-based methods under Support Vector Machine (SVM) and Random Forest (RF) algorithms in mapping a mangrove ecosystem into four main classes (Mangrove tree, mudflat, water, and sand spit) using resampled and fused Sentinel-2 images. Additionally, landscape metrics were used to identify the differences between spatial patterns obtained from different classification methods. Results showed that pixel-based classifications were influenced heavily by the effect of salt and pepper noise, whereas in object-based classifications, boundaries of land use land cover (LULC) polygons were smoother and visually more appealing. Object-based classifications, with an excellent level of kappa, distinguished mudflat and sand spit from each other and from mangrove better than the pixel-based classifications which obtained a fair-to-good level of kappa. RF and SVM performed differently under comparable circumstances. The results of landscape metrics comparison presented that the classification methods can be affected on quantifying area and size metrics. Although the results supported the idea that fused Sentinel images may provide better results in mangrove LULC classification, further research needs to develop and evaluate various image fusion approaches to make use of all Sentinel's fine resolution images. Our results on the mapping of mangrove ecosystems can contribute to the improvement of management and conservation strategies for these ecosystems being impacted by human activities.

Regelmässig erhobene, aktuelle Informationen zu verschiedenen Waldparametern auf landesweiter Ebene erfuhren in den letzten Jahren eine steigende praktische und wissenschaftliche Relevanz. Neueste Technologien aus der Fernerkundung und dem maschinellen Lernen haben eine neue Ära eingeläutet und ermöglichen eine genaue, reproduzierbare und flächige Erstellung der gewünschten Datensätze. Sie dienen als zuverlässige Basis zur Beschreibung des heutigen Zustandes sowie der zukünftigen Entwicklung des Waldes.
Im folgenden Artikel werden die im Rahmen des Schweiz. Landesforstinventars (LFI) landesweit verfügbaren Datensätze Vegetationshöhenmodell, Waldmischungsgrad, Gebüschwald und Holzvorrat vorgestellt. Es wird aufgezeigt, was heutzutage mit Fernerkundungs- und Modellierungstechniken möglich ist und wo aber auch die Grenzen liegen. Neben einer kurzen Beschreibung der verschiedenen Datensätze, wird ein spezielles Augenmerk auf die Interpretation der Resultate und deren Anwendung gelegt. Anhand des Waldmischungsgrads LFI wird der Unterschied zwischen Modellgenauigkeit und Produktegenauigkeit erklärt. Ein wichtiger Punkt ist die korrekte und solide Validierung, mit unabhängigen Daten und auch Vor-Ort-Begehungen. Zudem wird aufgezeigt, dass ein Austausch zwischen der Praxis und Forschung wichtiger denn je ist.

In recent years, up-to-date and regularly provided information on relevant forest parameters on countrywide level have experienced an increasing demand from both research and practice. Cutting edge remote sensing technologies and data sets have opened a new era by providing reliable, reproducible, and wall-to-wall forest parameter data sets. They serve as basis to assess the current state of forest ecosystems and their future developments and add to the existing NFI estimates by providing spatially explicit information.
This article introduces the four countrywide available forest parameter data sets vegetation height model, dominant leaf type, shrub forest and growing stock which were produced in the framework of the Swiss National Forest Inventory (NFI).
Besides a detailed description of these approaches, their limitations are critically discussed. In particular, the focus is laid on a correct interpretation and application of these data sets, the need of a proper validation, and the need of a regular exchange between research and practitioners.

Fernerkundungsdaten wie Luft- und Satellitenbilder können zur großräumigen Erfassung von geschädigten Waldflächen eingesetzt werden. Doch welche technischen Systeme und Verfahren sind praxisreif und liefern Waldbesitzern, Forstbetrieben und Verwaltungen rechtzeitig mit ausreichender Genauigkeit und bei vertretbaren Kosten die benötigten Informationen?

Essential forest ecosystem services can be assessed by better understanding the diversity of vegetation, specifically those of Mediterranean region. A species level classification of maquis would be useful in understanding vegetation structure and dynamics, which would be an indicator of degradation or succession in the region. Although remote sensing was regularly used for classification in the region, maquis are simply represented as one to three categories based on density or height. To fill this gap, we test the capability of Sentinel-2 imagery, together with selected ancillary variables, for an accurate mapping of the dominant maquis formations. We applied Recursive Feature Selection procedure and used a Random Forest classifier. The algorithm is tested using ground truth collected from site and reached 78% and 93% overall accuracy at species level and physiognomic level, respectively. Our results suggest species level characterization of dominant maquis is possible with Sentinel-2 spatial resolution.

Mangrove ecosystems are of high ecological value and occur in tropical and subtropical coastal areas. They consist of salt-tolerant woody plants and are highly complex and productive.
Here, a novel approach was developed to map disturbances and evaluate the ecological status of Iranian mangroves based on multispectral WorldView-2 and Sentinel-2 and remote sensing techniques. A land cover map was generated that serves as basis to assess changes in the composition and structure of mangrove ecosystems. The spatial disturbance index (SDI) was calculated based on landscape metrics and principal component analysis (PCA) in order to detect mangroves degraded on Qeshm Island, Iran. The approach comprised: (I) selection of landscape metrics, (II) definition of the hexagon cell scale, (III) completion of a PCA of the standardized values ​​of eight selected metrics, and (IV) modeling of the SDI based on eigenvalues and a loading factor. For validation, the variance inflation factor (VIF) was used. The results indicate that the best set of key metrics to quantify SDI comprises the eight metrics patch density (PD), total edge (TE), mean patch size (MPS), mean shape index (MSI), Euclidean nearest neighbor distance (ENND), contagion (CONT), splitting index (SPLIT), and Shannon’s diversity index (SHDI). The first three components of the PCA explained in total 87% (47%, 27%, and 13%, respectively) of the variance of information distribution, demonstrating the effectiveness of the PCA. Then, based on the disturbance index, five spatial levels, ranging from very low to very high disturbances, were identified. This approach enabled to detect the zonation of disturbance that has occurred in the region and supports the sustainable management of mangrove ecosystems and future planning efforts, thus making their conservation more realistic.

Historical black-and-white (B&W) aerial images have been recognized as an important source of information for assessing vegetation dynamics. However, the use of these images is limited by the lack of multispectral information, as well as by their varying quality. It is therefore important to study and develop methods that are capable of automatic and accurate classification of these B&W images while reducing the need for tedious manual work. The goal of this study was to assess changes over 30 years in woody vegetation cover along alpine treeline ecotones using B&W aerial images from two time points. A convolutional neural networks model was firstly set up based on three structure classes calculated from Airborne Laser Scanning data using the B&W aerial images from 2010. Then, the model was improved by active addition of training samples of those that were wrongly predicted from historical B&W aerial images from 1980. A comparison with visual image interpretation revealed generally high agreement for the class "dense forest" and lower agreement for the class "group of trees". The study illustrates that vegetation changes at the treeline ecotone can be detected in order to assess areawide long-term vegetation dynamics at a fine spatial resolution.

Nation-wide Sentinel-2 mosaics were used with National Forest Inventory (NFI) plot data for modelling and subsequent mapping of spruce-, pine-, and deciduous-dominated forest in Norway at a 16 m × 16 m resolution. The accuracies of the best model ranged between 74% for spruce and 87% for deciduous forest. An overall accuracy of 90% was found on stand level using independent data from more than 42 000 stands. Errors mostly resulting from a forest mask reduced the model accuracies by ~10%. The produced map was subsequently used to generate model-assisted (MA) and poststratified (PS) estimates of species-specific forest area. At the national level, efficiencies of the estimates increased by 20% to 50% for MA and up to 90% for PS. Greater minimum numbers of observations constrained the use of PS. For MA estimates of municipalities, efficiencies improved by up to a factor of 8 but were sometimes also less than 1. PS estimates were always equally as or more precise than direct and MA estimates but were applicable in fewer municipalities. The tree species prediction map is part of the Norwegian forest resource map and is used, among others, to improve maps of other variables of interest such as timber volume and biomass.

The constellation of two Sentinel-1 satellites provides an unprecedented coverage of Synthetic Aperture Radar (SAR) data at high spatial (20 m) and temporal (2 to 6 days over Europe) resolution. The availability of dense time series enables the analysis of the SAR temporal signatures and exploitation of these signatures for classification purposes. Frequent backscatter observations allow derivation of temporally filtered time series that reinforce the effect of changes in vegetation phenology by limiting the influence of short-term changes related to environmental conditions. Recent studies have already shown the potential of multitemporal Sentinel-1 data for forest mapping, forest type classification (coniferous or broadleaved forest) as well as for derivation of phenological variables at local to national scales. In the present study, we tested the viability of a recently published multi-temporal SAR classification method for continental scale forest mapping by applying it over Europe and evaluating the derived forest type and tree cover density maps against the European-wide Copernicus High Resolution Layers (HRL) forest datasets and national-scale forest maps from twelve countries. The comparison with the Copernicus HRL datasets revealed high correspondence over the majority of the European continent with overall accuracies of 86.1% and 73.2% for the forest/non-forest and forest type maps, respectively, and a Pearson correlation coefficient of 0.83 for tree cover density map. Moreover, the evaluation of both datasets against the national forest maps showed that the obtained accuracies of Sentinel-1 forest maps are almost within range of the HRL datasets. The Sentinel-1 forest/non-forest and forest type maps obtained average overall accuracies of 88.2% and 82.7%, respectively, as compared to 90.0% and 87.2% obtained by the Copernicus HRL datasets. This result is especially promising due to the facts that these maps can be produced with a high degree of automation and that only a single year of Sentinel-1 data is required as opposed to the Copernicus HRL forest datasets that are updated every three years.

Although trees growing outside forests are less perceived in relation to those inside forests, they are also valuable and serve various functions. The biophysical thresholds of Trees Outside Forests (TOF) were recently defined and standardised by the Global Forest Resources Assessments of the United Nations Food and Agricultural Organization (UNFAO-FRA). With these definitions, standardised mapping of TOF resources became feasible at the national and international level. Here, we aimed to fill the spatial information gap of TOF resources at the national scale with an automated mapping approach based on the UNFAO-FRA definition. The approach was carried out in ArcGIS using adapted UNFAO-FRA biophysical thresholds with routinely acquired countrywide remote sensing data for the whole of Switzerland: a Vegetation Height Model (VHM), a Topographic Landscape Model (TLM Regio) land cover map, and a Forest Mask. Results were validated using stereo-image interpretation data of the Swiss National Forest Inventory (NFI), which verified 95%, 55% and 75% of overall, producer’s and user’s accuracy, respectively. Of the five forest production regions in Switzerland, the highest accuracy was achieved in the Central Plateau and the lowest in the Alps, with accuracy generally decreasing with increasing elevation. Omission and commission errors were highly correlated with the vertical and horizontal accuracy of the VHM, and the applied biophysical thresholds caused both types of error. The final TOF map produced with our approach is at the countrywide scale, is superior to existing TOF information, meets UNFAO-FRA management and reporting needs, and enables the derivation of TOF’s biomass, carbon sequestration potential and species distribution for the whole country.

The continual increase of shrub forest in the Swiss Alps over the past few decades impacts biodiversity, forest succession and the protective function of forests. Therefore, up-to-date and area-wide information on its distribution is of great interest. To detect the shrub forest areas for the whole of Switzerland (41,285 km²), we developed an approach that uses Random Forest (RF), bias correction techniques and data from multiple remote sensing sources. Manual aerial orthoimage interpretation of shrub forest areas was conducted in a non-probabilistic way to derive initial training data. The multi-sensor and open access predictor data included digital terrain and vegetation height models obtained from Airborne Laser Scanning (ALS) and stereo-imagery, as well as Synthetic Aperture Radar (SAR) backscatter from Sentinel-1 and multispectral imagery from Sentinel-2. To mitigate the expected bias due to the training data sampling strategy, two techniques using RF probability estimates were tested to improve mapping accuracy. 1) an iterative and semi-automated active learning technique was used to generate further training data and 2) threshold-moving related object growing was applied. Both techniques facilitated the production of a shrub forest map for the whole of Switzerland at a spatial resolution of 10 m. An accuracy assessment was performed using independent data covering 7640 regularly distributed National Forest Inventory (NFI) plots. We observed the influence of the bias correction techniques and found higher accuracies after each performed iteration. The Mean Absolute Error (MAE) for the predicted shrub forest proportion was reduced from 6.04% to 2.68% while achieving a Mean Bias Error (MBE) of close to 0. The present study underscores the potential of combining multi-sensor data with bias correction techniques to provide cost-effective and accurate countrywide detection of shrub forest. Moreover, the map complements currently available NFI plot sample point data.

Künstliche Intelligenz, Deep learning, hochaufgelöste Satellitenbilder, automatisierte Auswertungen, landesweite Datensätze etc. sind alles Begriffe, mit denen viele im Bereich Wald- und Forstwissenschaften schon einmal konfrontiert worden sind. Was steckt aber nun genau dahinter bzw. wird bereits angewandt? Im folgenden Artikel wird anhand des neuen «Waldmischungsgrades LFI» aufgezeigt, was fernerkundungs-technisch alles möglich ist und wo eben auch die Grenzen liegen. Dabei wird auch deutlich, dass ein Austausch zwischen der Praxis und der Forschung wichtiger denn je ist.

Countrywide winter and summer Sentinel-1 (S1) backscatter data, cloud-free summer Sentinel-2 (S2) images, an Airborne Laser Scanning (ALS)-based Digital Terrain Model (DTM) and a forest mask were used to model and subsequently map Dominant Leaf Type (DLT) with the thematic classes broadleaved and coniferous trees for the whole of Switzerland. A novel workflow was developed that is robust, cost-efficient and highly automated using reference data from aerial image interpretation. Two machine learning approaches based on Random Forest (RF) and deep learning (UNET) for the whole country with three sets of predictor variables were applied. 24 subareas based on aspect and slope categories were applied to explore effects of the complex mountainous topography on model performances. The reference data split into training, validation and test data sets was spatially stratified using a 25 km regular grid. Model accuracies of both RF and UNET were generally highest with Kappa (K) around 0.95 when predictors were included from both S1/S2 and the topographic variables aspect, elevation and slope from the DTM. While only slightly lower accuracies were obtained when using S2 and DTM data, lowest accuracies were obtained when only predictors from S1 and DTM were included, with RF performing worse than UNET. While on countrywide level RF and UNET performed overall similarly, substantial differences in model performances, i.e. higher variances and lower accuracies, were found in subareas with northwest to northeast orientations. The combined use of S1/S2 and DTM predictors mitigated these problems related to topography and shadows and was therefore superior to the single use of S1 and DTM or S2 and DTM data. The comparison with independent National Forest Inventory (NFI) plot data demonstrated precisions of K around 0.6 in the predictions of DLT and indicated a trend of increasing deviations in mixed forests. A comparison with the Copernicus High Resolution Layer (HRL) DLT 2018 revealed overall higher map accuracies with the exception of pure broadleaved forest. Although, spatial patterns of DTL were overall similar, UNET performed better than RF in areas with a distinct DLT on forest stand level, with the largest differences occurring when only S1 and DTM data was used. In contrast, predictions obtained from RF were more accurate in mixed stands. This study goes beyond the case study level and meets the requirements of countrywide data sets, in particular regarding repeatability, updating, costs and characteristics of training data sets. The 10 m countrywide DLT maps add complementary and spatially explicit information to the existing NFI estimates and are thus highly relevant for forestry practice and other related fields.