Field Reference Data Research Articles

Computation of prediction intervals for forest aboveground biomass predictions using generalized linear models in a large-extent boreal forest region

Abstract Remotely sensed data have an important application for estimation of forest variables, e.g. height, volume, and aboveground biomass (AGB). The increased use of remotely sensed data implemented along with model-based inference has shown improved efficiency in prediction and mapping of such forest variables. In this study, plot-level airborne laser scanning data and Swedish National Forest Inventory field reference data were used to predict AGB using generalized linear models (GLMs) assuming Gamma and Tweedie distributions for the field observed AGB. The GLMs were selected considering the convenience of not correcting transformation bias as it is required in other regression models with transformed response variable. To overcome the challenge in providing reliable uncertainty estimates for the estimated forest variable map products at individual pixel-scale, we focused on computing 95% prediction intervals (PIs) for Gamma and Tweedie GLMs with a square root link function. The relative uncertainties were computed as the ratio between the half-width of the PIs and the predicted AGBs. The AGB-airborne laser scanning models were developed with root mean square error values of 22.6 Mgha−1 (26%) and 21.7 Mgha−1 (25%), respectively, for the Gamma and Tweedie GLMs. Two methods were applied to compute PIs for the Gamma GLM, one using the R package ‘ciTools’ and another derived through asymptotic theory. It was found that the 95% PIs computed using ‘ciTools’ had the most accurate coverage probability in comparison to the other method. An extended version of these PIs was also utilized for the Tweedie GLMs. The range of PIs associated with the prediction of AGB were narrower for lower predicted AGB values compared with the length of higher predicted AGB values. Comparing the two fitted models, the Gamma GLM showed lower relative uncertainties for the lower range of predicted AGBs, whereas the Tweedie GLM showed lower relative uncertainties for the higher range of predicted AGBs. Overall, the Tweedie GLM provided a better model fit for AGB predictions.

Predicting Potato Diseases in Smallholder Agricultural Areas of Nigeria Using Machine Learning and Remote Sensing-Based Climate Data

Crop disease management is crucial for sustainable food production. Although farmers in Nigeria continue to apply broad-spectrum fungicides to potato, potato diseases are still on the rise. Machine-learning methods have recently become more common as part of epidemiological early warning systems. They provide vital information on data–disease relations that may be useful for predisease management. In this study, we build on machine-learning methods to develop spatial early warning tools for Nigeria, using the Jos Plateau as a test case. Both remote sensing meteorological and field data were used to (i) predict disease incidence using field reference data and a random forest (RF) classifier and to (ii) identify local conditions conducive to potato diseases using multi-criteria classification (MCC) based on machine-learning results. The results of the RF for 2019, 2020, and 2021 showed similar spatial characteristics, whereas the MCC varied significantly. Both models predicted that between 72 and 96% of potato fields would be infested. The MCC model further revealed that spatiotemporal frequencies of vulnerability in June can serve as the indicator that informs degrees of infestation. A 5-day vulnerability window used in the context of the MCC proved to be the most useful tool for developing an efficient spraying regime, based on a combination of temperature, rainfall, and relative humidity thresholds. As a result, we were able to develop an operational early-warning system for potato disease in the tropical highlands of Africa. In particular, we introduced spatial risks, creating a more sustainable early-warning approach. [Formula: see text] Copyright © 2024 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license .

Generating synthetic laser scanning data of forests by combining forest inventory information, a tree point cloud database and an open-source laser scanning simulator

Abstract Airborne laser scanning (ALS) data are routinely used to estimate and map structure-related forest inventory variables. The further development, refinement and evaluation of methods to derive forest inventory variables from ALS data require extensive datasets of forest stand information on an individual tree-level and corresponding ALS data. A cost-efficient method to obtain such datasets is the combination of virtual forest stands with a laser scanning simulator. We present an approach to simulate ALS data of forest stands by combining forest inventory information, a tree point cloud database and the laser scanning simulation framework HELIOS++. ALS data of six 1-ha plots were simulated and compared to real ALS data of these plots. The synthetic 3D representations of the forest stands were composed of real laser scanning point clouds of individual trees that were acquired by an uncrewed aerial vehicle (UAV), and, for comparison, simplified tree models with cylindrical stems and spheroidal crowns. The simulated ALS point clouds of the six plots were compared with the real point clouds based on canopy cover, height distribution of returns and several other point cloud metrics. In addition, the performance of biomass models trained using these synthetic data was evaluated. The comparison revealed that, in general, both the real tree models and the simplified tree models can be used to generate synthetic data. The results differed for the different study sites and depending on whether all returns or only first returns were considered. The measure of canopy cover was better represented by the data of the simplified tree models, whereas the height distribution of the returns was – for most of the study sites – better represented by the real tree model data. Training biomass models with metrics derived from the real tree model data led to an overestimation of biomass, while using metrics of the simplified tree model data resulted in an underestimation of biomass. Still, the accuracy of models trained with simulated data was only slightly lower compared to models trained with real ALS data. Our results suggest that the presented approach can be used to generate ALS data that are sufficiently realistic for many applications. The synthetic data may be used to develop new or refine existing ALS-based forest inventory methods, to systematically investigate the relationship between point cloud metrics and forest inventory variables and to analyse how this relationship is affected by laser scanning acquisition settings and field reference data characteristics.

Detecting the presence of natural forests using airborne laser scanning data

Centuries of forest exploitation have caused significant loss of natural forests in Europe, leading to a decline in populations for many species. To prevent further loss in biodiversity, the Norwegian government has set a target of protecting 10% of the forested area. However, recent data from the National Forest Inventory (NFI) reveals that less than 2% of Norway's forested area consists of natural forests. To identify forests with high conservation value, we used vertical and horizontal variables derived from airborne laser scanning (ALS) data, along with NFI plot measurements. Our study aimed to predict the presence of natural forests across three counties in southeastern Norway, using three different definitions: pristine, near-natural, and semi-natural forests. Natural forests are scarce, and their underrepresentation in field reference data can compromise the accuracy of the predictions. To address this, we assessed the potential gain of including additional field data specifically targeting natural forests to achieve a better balance in the dataset. Additionally, we examined the impact of stratifying the data by dominant tree species on the performance of the models. Our results revealed that semi-natural forests were the most accurately predicted, followed by near-natural and pristine forests, with Matthews correlation coefficient values of 0.32, 0.24, and 0.17, respectively. Including additional field data did not improve the predictions. However, stratification by species improved the accuracy of predictions for near-natural and semi-natural forests, while reducing accuracy for pristine forests. The use of horizontal variables did not improve the predictions. Our study demonstrates the potential of ALS data in identifying forests with high conservation value. It provides a basis for further research on the use of ALS data for the detection and conservation of natural forests, offering valuable insights to guide future forest preservation efforts.

Quantify and account for field reference errors in forest remote sensing studies

Field inventoried data are often used as references (ground truth) in forest remote sensing studies. However, the reference values are affected by various kinds of errors, which tend to make the reported accuracies of the remote sensing-based predictions worse than they are. The more accurate the remote sensing techniques are becoming, the more pronounced this problem will be. This paper addresses the impact of uncertainties in field reference data due to measurement errors, model errors, and position errors when evaluating the accuracy of biomass predictions from airborne laser scanning at plot level. We present novel theoretical analysis methods that take the interactions of the error sources into account. Further, an error characterization model (ECM) is used to describe the error structure of the remote sensing-based predictions, and we show how the parameters of the ECM can be adjusted when field references contain errors. We also show how root mean square error (RMSE) estimates can be adjusted. Based on data from Scandinavian forests, we conclude that the field reference errors have an impact on the remote sensing-based predictions. By accounting for these errors the RMSE of the remote sensing-based predictions was reduced by 6–18%. The most influential sources of error in the field references were found to be the residual errors of the allometric biomass model and the field plot position errors. Together, these two sources accounted for 97% of the variance while measurement errors and biomass model parameter uncertainties were negligible in our study.

Multi-Sensor Remote Sensing of Intertidal Flat Habitats for Migratory Shorebird Conservation

Many species of shorebirds migrate long distances from their overwintering grounds in the southern hemisphere to breeding grounds in the northern hemisphere. The coastal intertidal zone, consisting of sand and mud flats exposed at low tide and covered at high tide, is heavily used as a migratory stopover or overwintering habitat. Understanding the spatial distribution of sediment types at these stopover sites is a critical step for understanding habitat use by shorebird species. Due to their importance as overwintering and stopover habitat for the imperiled western Atlantic subpopulation of the shorebird, the red knot (Calidris canutus rufa), as well as other migratory shorebirds, the northern coast of Brazil between Pará and Maranhão, and Bahía Lomas in northern Tierra del Fuego, Chile, were selected for further investigation as to the applicability of remotely sensed characterization of the intertidal flat habitats. Examination of the Landsat 8 multispectral reflectance and Sentinel-1 SAR backscatter reveals that sand and mud represent endmembers at opposite ends of a continuous gradient in feature space. While remotely sensed data can be used to discriminate between mud and sand intertidal types, the spectral relationships varied between the two very different geographic locations. The inclusion of both multispectral and radar sensing imagery can lead to important insights about the physical properties of the sediment that would be omitted by using one data source alone. Spectral unmixing techniques in Google Earth Engine were used to map the intertidal zone into general sediment classes spanning the gradient (i.e., mud, sandy mud, muddy sand, and sand). Comparison of the mapped outputs with field reference data suggests that mapping of mud- vs. sand-dominated areas can be accomplished with reasonable accuracy (overall accuracy of 75%).

Landsat Data Based Prediction of Loblolly Pine Plantation Attributes in Western Gulf Region, USA

The suitability of using Landsat sensor variables to predict key stand attributes, including stand average dominant/codominant tree height (HT), mean diameter at breast height (DBH), the number of trees per hectare (NT), basal area per hectare (BA), and stand density index (SDI), of intensively managed loblolly pine plantations in the Western Gulf Region at the plot/stand level was assessed. In total, thirty Landsat sensor variables including six original bands, three vegetation indices, three Tasseled Cap transformed indices, and eighteen texture measure variables were used as predictors. Field data of 125 permanent plots located across east Texas and western Louisiana were used as reference data. Individual trees of those plots were measured at plot establishment (referred to as the first cycle measurement; average about 4.5 years old) and remeasured in three-year intervals (the second cycle measurement at approximately seven years old and the third cycle measurement at approximately 10 years old). Thus, field reference data represent stand development from open- (first cycle) to closed-canopy (third cycle). Models to predict stand HT, DBH, NT, BA, and SDI were developed by cycle using multiple linear regression (MLR) and also random forests (RF) methods. Results indicated that the first cycle stands HT, DBH, BA, and SDI were well predicted using the Landsat sensor variables with R2 > 0.7 and low RMSEs. These relationships weakened with stand age, although still moderate with R2 being around 0.45 for the second cycle measurement and became practically useless (R2 < 0.30) for the third cycle measurement. For NT, no meaningful models were achieved regardless of the measurement cycle. The MLR and RF models were comparable in accuracy and had similar key predictors. Overall, the shortwave infrared bands, red band, and wetness index were the most important predictors, but their dominance declined with the cycle. Texture measure variables were relatively less important but a trend of increasing their importance with cycle was noted. Results show promise for operationally predicting stand variables for young pine plantations, an age class that typically presents significant challenges using conventional forest measurement methodologies. Potential methods to further improve model accuracy and how to use the results within the context of pine plantation management planning in the region were discussed.

Aboveground biomass mapping by integrating ICESat-2, SENTINEL-1, SENTINEL-2, ALOS2/PALSAR2, and topographic information in Mediterranean forests

ABSTRACT The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) provides an extraordinary opportunity to support global large-scale forest carbon mapping, but further research is needed in order to obtain wall-to-wall forest aboveground biomass (AGB) maps with this technology. The effects of vegetation structure on the performance of canopy height and AGB modeling using ICESat-2 photon-counting light detection and ranging (LiDAR) data in Mediterranean forest areas have not been previously studied in the literature. In this study, we combined recent ICESat-2 vegetation (ATL08) data, Airborne Laser Scanning (ALS)- and field-based estimates, and a multi-sensor earth observation composite for extrapolation of AGB estimates and AGB mapping. A diverse gradient of forest Mediterranean ecosystems, distributed over 19,744.15 km2 of forest area in the region of Extremadura (Spain), with different species and structural complexity forming 5 different forest types (3 Quercus spp. dominated and 2 Pinus spp. dominated forests), was used to (i) evaluate the precision of ICESat-2 canopy height estimations, (ii) develop ICESat-2-based AGB models, and (iii) generate a spatially continuous prediction of AGB by using data from the satellite missions Sentinel-1 (S1), Sentinel-2 (S2), Phased Array L-band Synthetic Aperture Radar (ALOS2/PALSAR2), and Shuttle Radar Topography Mission (SRTM). First, ALS- and ICESat-2-derived metrics that best described canopy height (p98 and rh98, respectively) were compared at the ATL08 segment level. Second, ALS-based AGB values were derived at the ATL08 segment scale. Third, ALS-based AGB estimates at the ICESat-2 segment level were used as dependent variables to fit ICESat-2-based AGB models. Fourth, a multi-sensor approach was then implemented to predict ICESat-2-derived AGB, by means of a Random Forest (RF) modeling technique, with predictors retrieved from S1, S2, ALOS2/PALSAR2, and SRTM. Finally, RF was used to generate wall-to-wall AGB maps that were compared with field-, ALS- and ICESat-2-based observations. The agreement between the ALS- and ICESat-2-derived metrics related to the canopy height distribution was higher for Pinus spp. forest than for the Quercus spp-dominated forests. The ICESat-2-based AGB models yielded model efficiency (Mef) values between 0.56 and 0.80, with a RMSE ranging from 7.76 to 17.71 Mg ha−1 and rRMSE from 19.04 to 55.21%. The multi-sensor RF models provided the following results when compared with the ICESat-2- and ALS-based AGB observations: R 2 values of 0.63 and 0.64, and RMSE values of 11.10 Mg ha−1(rRMSE = 28.15%) and 12.28 Mg ha−1 (rRMSE = 31.45%), respectively, and an approximately unbiased result (0.03 Mg ha−1 and 0.09 Mg ha−1). When applied to the field-based validation data set (4th Spanish National Forest Inventory (SNFI-4) plots = 508), the RF-derived AGB model showed a relatively lower predictive capacity (R 2 = 0.45), a higher RMSE value (25.88 Mg ha−1) and slightly biased results (−1.47 Mg ha−1), especially for larger field-derived AGB intervals. The results of this study serve to provide an initial quantitative assessment of the ICESat-2 ATL08 data for large-scale AGB estimation. The findings suggest that a multi-sensor approach may be feasible for extrapolating ICESat-2-derived AGB estimates over areas where field or ALS reference data are not available.

Direct and automatic measurements of stem curve and volume using a high-resolution airborne laser scanning system

Today, high-quality reference tree measurements, including the position, diameter, height and volume, are cumbersome and slow to carry out, but highly needed for forest inventories based on airborne laser scanning. Mobile laser scanning technologies hold the promise for collecting reference data for forest inventories with an extremely high efficiency. Perhaps, the most efficient approach for reference data collection would be to mount a high-resolution laser scanning system on board an airborne vehicle flying at a low altitude above the forest canopy since this would allow recording reference samples of individual trees with the speed of flight. To demonstrate the potential of this technology, we mounted an in-house developed HeliALS-DW laser scanning system on board a helicopter and collected point cloud data in a boreal forest on three test sites containing a total of 1469 trees. The obtained point clouds incorporated sufficiently many high-quality stem hits for estimating the stem curves and stem volumes of individual trees since the point clouds had a relatively high point density of 2200–3800 echoes/m2, and the scanner had been tilted by 15° from the nadir to increase the possibility of recording stem hits. To automatically estimate the diameters at breast height (DBH) and stem curves of individual trees, we used algorithms designed to tolerate moderate drifts in the trajectory of the laser scanner. Furthermore, the stem volumes of individual trees were computed by using the estimated stem curves and tree heights without any allometric models. Using the proposed methods, we were able to estimate the stem curves with a root-mean-square error (RMSE) of 1.7–2.6 cm (6–9%) while detecting 42–71% of the trees. The RMSE of stem volume estimates was 0.1–0.15 m3 (12–21%). We also showed that the tree detection rate could be improved up to 87–96% for trees with a DBH exceeding 20 cm if slightly larger average errors for the stem attributes were allowed. Our results pave the way for using high-resolution airborne laser scanning for field reference data collection by conducting direct measurements of tree stems with a high efficiency.

Individual tree detection and estimation of stem attributes with mobile laser scanning along boreal forest roads

The collection of field-reference data is a key task in remote sensing-based forest inventories. However, traditional methods of collection demand extensive personnel resources. Thus, field-reference data collection would benefit from more automated methods. In this study, we proposed a method for individual tree detection (ITD) and stem attribute estimation based on a car-mounted mobile laser scanner (MLS) operating along forest roads. We assessed its performance in six ranges with increasing mean distance from the roadside. We used a Riegl VUX-1LR sensor operating with high repetition rate, thus providing detailed cross sections of the stems. The algorithm we propose was designed for this sensor configuration, identifying the cross sections (or arcs) in the point cloud and aggregating those into single trees. Furthermore, we estimated diameter at breast height (DBH), stem profiles, and stem volume for each detected tree. The accuracy of ITD, DBH, and stem volume estimates varied with the trees’ distance from the road. In general, the proximity to the sensor of branches 0–10 m from the road caused commission errors in ITD and over estimation of stem attributes in this zone. At 50–60 m from roadside, stems were often occluded by branches, causing omissions and underestimation of stem attributes in this area. ITD’s precision and sensitivity varied from 82.8% to 100% and 62.7% to 96.7%, respectively. The RMSE of DBH estimates ranged from 1.81 cm (6.38%) to 4.84 cm (16.9%). Stem volume estimates had RMSEs ranging from 0.0800 m3 (10.1%) to 0.190 m3 (25.7%), depending on the distance to the sensor. The average proportion of detected reference volume was highly affected by the performance of ITD in the different zones. This proportion was highest from 0 to 10 m (113%), a zone that concentrated most ITD commission errors, and lowest from 50 to 60 m (66.6%), mostly due to the omission errors in this area. In the other zones, the RMSE ranged from 87.5% to 98.5%. These accuracies are in line with those obtained by other state-of-the-art MLS and terrestrial laser scanner (TLS) methods. The car-mounted MLS system used has the potential to collect data efficiently in large-scale inventories, being able to scan approximately 80 ha of forests per day depending on the survey setup. This data collection method could be used to increase the amount of field-reference data available in remote sensing-based forest inventories, improve models for area-based estimations, and support precision forestry development.

Land cover classification of treeline ecotones along a 1100 km latitudinal transect using spectral‐ and three‐dimensional information from UAV‐based aerial imagery

AbstractThe alpine treeline ecotone is expected to move upwards in elevation with global warming. Thus, mapping treeline ecotones is crucial in monitoring potential changes. Previous remote sensing studies have focused on the usage of satellites and aircrafts for mapping the treeline ecotone. However, treeline ecotones can be highly heterogenous, and thus the use of imagery with higher spatial resolution should be investigated. We evaluate the potential of using unmanned aerial vehicles (UAVs) for the collection of ultra‐high spatial resolution imagery for mapping treeline ecotone land covers. We acquired imagery and field reference data from 32 treeline ecotone sites along a 1100 km latitudinal gradient in Norway (60–69°N). Before classification, we performed a superpixel segmentation of the UAV‐derived orthomosaics and assigned land cover classes to segments: rock, water, snow, shadow, wetland, tree‐covered area and five classes within the ridge‐snowbed gradient. We calculated features providing spectral, textural, three‐dimensional vegetation structure, topographical and shape information for the classification. To evaluate the influence of acquisition time during the growing season and geographical variations, we performed four sets of classifications: global, seasonal‐based, geographical regional‐based and seasonal‐regional‐based. We found no differences in overall accuracy (OA) between the different classifications, and the global model with observations irrespective of data acquisition timing and geographical region had an OA of 73%. When accounting for similarities between closely related classes along the ridge‐snowbed gradient, the accuracy increased to 92.6%. We found spectral features related to visible, red‐edge and near‐infrared bands to be the most important to predict treeline ecotone land cover classes. Our results show that the use of UAVs is efficient in mapping treeline ecotones, and that data can be acquired irrespective of timing within a growing season and geographical region to get accurate land cover maps. This can overcome constraints of a short field‐season or low‐resolution remote sensing data.

Accelerating the simulation of annual bifacial illumination of real photovoltaic systems with ray tracing

SummaryAccurate modeling of bifacial illumination is critical to improve the prediction of the energy yield of bifacial solar systems. Monte Carlo ray tracing is the most powerful tool to accomplish this task. In this work, we accelerate Monte Carlo ray tracing of large solar systems by nearly 90%. Our model achieves root-mean-square error values of 7.9% and 37.2% for the front and rear irradiance compared against single-axis tracking field reference data, respectively. The rear irradiance modeling error decreases to 18.9% if suspected snow periods are excluded. Crucially, our full system simulations show that surrounding ground surfaces affect the rear irradiance deep into the system. Therefore, unit system simulations cannot necessarily ignore the influence of the perimeter of large installations to accurately estimate annual yield. Large-scale simulations involving high-performance supercomputing were necessary to investigate these effects accurately, calibrate our simplified models, and validate our results against experimental measurements.

Reliable Estimates of Merchantable Timber Volume from Terrestrial Laser Scanning

Simple and accurate determination of merchantable tree height is needed for accurate estimations of merchantable volume. Conventional field methods of forest inventory can lead to biased estimates of tree height and diameter, especially in complex forest structures. Terrestrial laser scanner (TLS) data can be used to determine merchantable height and diameter at different heights with high accuracy and detail. This study focuses on the use of the random sampling consensus method (RANSAC) for generating the length and diameter of logs to estimate merchantable volume at the tree level using Huber’s formula. For this study, we used two plots; plot A contained deciduous trees and plot B consisted of conifers. Our results demonstrated that the TLS-based outputs for stem modelling using the RANSAC method performed very well with low bias (0.02 for deciduous and 0.01 for conifers) and a high degree of accuracy (97.73% for deciduous and 96.14% for conifers). We also found a high correlation between the proposed method and log length (−0.814 for plot A and −0.698 for plot B), which is an important finding because this information can be used to determine the optimum log properties required for analyzing stem curvature changes at different heights. Furthermore, the results of this study provide insight into the applicability and ergonomics during data collection from forest inventories solely from terrestrial laser scanning, thus reducing the need for field reference data.

Under-Canopy UAV Laser Scanning Providing Canopy Height and Stem Volume Accurately

The automation of forest field reference data collection has been an intensive research objective for laser scanning scientists ever since the invention of terrestrial laser scanning more than two decades ago. In this study, we demonstrated that an under-canopy UAV laser scanning system utilizing a rotating laser scanner can alone provide accurate estimates of canopy height and stem volume for the majority of trees in a boreal forest. We mounted a rotating laser scanner based on a Velodyne VLP-16 sensor onboard a manually piloted UAV. The UAV was commanded with the help of a live video feed from the onboard camera. Since the system was based on a rotating laser scanner providing varying view angles, all important elements such as treetops, branches, trunks, and ground could be recorded with laser hits. In an experiment including two different forest structures, namely sparse and obstructed canopy, we showed that our system can measure the heights of individual trees with a bias of −20 cm and a standard error of 40 cm in the sparse forest and with a bias of −65 cm and a standard error of 1 m in the obstructed forest. The accuracy of the obtained tree height estimates was equivalent to airborne above-canopy UAV surveys conducted in similar forest conditions or even at the same sites. The higher underestimation and higher inaccuracy in the obstructed site can be attributed to three trees with a height exceeding 25 m and the reduced point density of these tree tops due to occlusion and the limited ranging capacity of the scanner. Additionally, we used our system to estimate the stem volumes of individual trees with a standard error at the level of 10%. This level of error is equivalent to the error obtained when merging above-canopy UAV laser scanner data with terrestrial point cloud data. The results show that we do not necessarily need a combination of terrestrial point clouds and point clouds collected using above-canopy UAV systems in order to accurately estimate the heights and the volumes of individual trees in reference data collection.

Monitoring Invasive Plant Species Using Hyperspectral Remote Sensing Data

The species richness and biodiversity of vegetation in Hungary are increasingly threatened by invasive plant species brought in from other continents and foreign ecosystems. These invasive plant species have spread aggressively in the natural and semi-natural habitats of Europe. Common milkweed (Asclepias syriaca) is one of the species that pose the greatest ecological menace. Therefore, the primary purpose of the present study is to map and monitor the spread of common milkweed, the most common invasive plant species in Europe. Furthermore, the possibilities to detect and validate this special invasive plant by analyzing hyperspectral remote sensing data were investigated. In combination with field reference data, high-resolution hyperspectral aerial images acquired by an unmanned aerial vehicle (UAV) platform in 138 spectral bands in areas infected by common milkweed were examined. Then, support vector machine (SVM) and artificial neural network (ANN) classification algorithms were applied to the highly accurate field reference data. As a result, common milkweed individuals were distinguished in hyperspectral images, achieving an overall accuracy of 92.95% in the case of supervised SVM classification. Using the ANN model, an overall accuracy of 99.61% was achieved. To evaluate the proposed approach, two experimental tests were conducted, and in both cases, we managed to distinguish the individual specimens within the large variety of spreading invasive species in a study area of 2 ha, based on centimeter spatial resolution hyperspectral UAV imagery.

Combining TanDEM-X and Sentinel-2 for large-area species-wise prediction of forest biomass and volume

In this study, data from the satellite sensors TanDEM-X and Sentinel-2 were combined with national field inventory data to predict forest above-ground biomass (AGB) and stem volume (VOL) over a large area in Sweden. The data sources were evaluated both separately and in combination. The study area covers approximately 20,000,000 ha and corresponds to about 70% of the Swedish forest area. The study area was divided into tiles of 2.5 × 2.5 km2, which were processed sequentially. The field plots were inventoried on 7 m and 10 m circular plots by the Swedish National Forest Inventory, and plot AGB and VOL at the year of the satellite data were estimated based on a 10-year period of field data. The AGB and VOL were modelled using the k nearest neighbor (kNN) algorithm, with k = 5 neighbors. The combined use of two data sources with different scene extents enabled the generation of seamless AGB and VOL maps. Moreover, the kNN algorithm provided the VOL divided per tree species, which was used for classification of the dominant tree species at stand-level. The overall accuracy for the dominant tree species classification was 77%. The predicted AGB and VOL rasters were evaluated using 549 field inventoried forest stands distributed over Sweden. The RMSE for the predictions based on both data sources were 31.4 t/ha (29.1%) for AGB, and 59.0 m3/ha (30.2%) for VOL. By estimating and removing the variance due to sampling (the stand values were estimated from sample plots), the RMSE was improved to 18.0 t/ha (16.6%). The evaluated approach of using kNN was suitable for estimating forest variables from a combination of different satellite sensors, provided sufficient field reference data are available. The TanDEM-X data were most important for the AGB and VOL predictions, while Sentinel-2 data were essential to map the tree species.

Drone data for decision making in regeneration forests: from raw data to actionable insights1

In this study, we aim at developing ways to directly translate raw drone data into actionable insights, thus enabling us to make management decisions directly from drone data. Drone photogrammetric data and data analytics were used to model stand-level immediate tending need and cost in regeneration forests. Field reference data were used to train and validate a logistic model for the binary classification of immediate tending need and a multiple linear regression model to predict the cost to perform the tending operation. The performance of the models derived from drone data was compared to models utilizing the following alternative data sources: airborne laser scanning data (ALS), prior information from forest management plans (Prior) and the combination of drone +Prior and ALS +Prior. The use of drone data and prior information outperformed the remaining alternatives in terms of classification of tending needs, whereas drone data alone resulted in the most accurate cost models. Our results are encouraging for further use of drones in the operational management of regeneration forests and show that drone data and data analytics are useful for deriving actionable insights.

Comparison of Backpack, Handheld, Under-Canopy UAV, and Above-Canopy UAV Laser Scanning for Field Reference Data Collection in Boreal Forests

In this work, we compared six emerging mobile laser scanning (MLS) technologies for field reference data collection at the individual tree level in boreal forest conditions. The systems under study were an in-house developed AKHKA-R3 backpack laser scanner, a handheld Zeb-Horizon laser scanner, an under-canopy UAV (Unmanned Aircraft Vehicle) laser scanning system, and three above-canopy UAV laser scanning systems providing point clouds with varying point densities. To assess the performance of the methods for automated measurements of diameter at breast height (DBH), stem curve, tree height and stem volume, we utilized all of the six systems to collect point cloud data on two 32 m-by-32 m test sites classified as sparse (n = 42 trees) and obstructed (n = 43 trees). To analyze the data collected with the two ground-based MLS systems and the under-canopy UAV system, we used a workflow based on our recent work featuring simultaneous localization and mapping (SLAM) technology, a stem arc detection algorithm, and an iterative arc matching algorithm. This workflow enabled us to obtain accurate stem diameter estimates from the point cloud data despite a small but relevant time-dependent drift in the SLAM-corrected trajectory of the scanner. We found out that the ground-based MLS systems and the under-canopy UAV system could be used to measure the stem diameter (DBH) with a root mean square error (RMSE) of 2–8%, whereas the stem curve measurements had an RMSE of 2–15% that depended on the system and the measurement height. Furthermore, the backpack and handheld scanners could be employed for sufficiently accurate tree height measurements (RMSE = 2–10%) in order to estimate the stem volumes of individual trees with an RMSE of approximately 10%. A similar accuracy was obtained when combining stem curves estimated with the under-canopy UAV system and tree heights extracted with an above-canopy flying laser scanning unit. Importantly, the volume estimation error of these three MLS systems was found to be of the same level as the error corresponding to manual field measurements on the two test sites. To analyze point cloud data collected with the three above-canopy flying UAV systems, we used a random forest model trained on field reference data collected from nearby plots. Using the random forest model, we were able to estimate the DBH of individual trees with an RMSE of 10–20%, the tree height with an RMSE of 2–8%, and the stem volume with an RMSE of 20–50%. Our results indicate that ground-based and under-canopy MLS systems provide a promising approach for field reference data collection at the individual tree level, whereas the accuracy of above-canopy UAV laser scanning systems is not yet sufficient for predicting stem attributes of individual trees for field reference data with a high accuracy.

FIRST RESULTS OF THE LEM BENCHMARK DATABASE FOR AGRICULTURAL APPLICATIONS

Abstract. Applying remote sensing technology to map and monitor agriculture and its impacts can greatly contribute for the proper development of this activity, promoting efficient food, fiber and energy production. For that, not only remote sensing images are needed, but also ground truth information, which is a key factor for the development and improvement of methodologies using remote sensing data. While a variety of images are current available, inclusive cost-free images, field reference data is scarcer. For agricultural applications, especially in tropical regions such as Brazil, where the agriculture is very dynamic and diverse (recent agricultural frontiers, crop rotations, multiple cropping systems, several management practices, etc.), and cultivated over a vast territory, this task is not trivial. One way of boosting the researches in agricultural remote sensing is to stimulate people to share their data, and to foster different groups to use the same dataset, so distinct methods can be properly compared. In this context, our group created the LEM Benchmark Database (a project funded by the ISPRS Scientific Initiative project - 2017) from the Luiz Eduardo Magalhães (LEM) municipality, Bahia State, Brazil. The database contains a set of pre-processed multitemporal satellite images (Landsat-8/OLI, Sentinel-2/MSI and SAR band-C Sentinel-1) and shapefiles of agricultural fields with their correspondent monthly land use classes, covering the period of one Brazilian crop year (2017–2018). In this paper we present the first results obtained with this database.

Evaluating the potential of full‐waveform lidar for mapping pan‐tropical tree species richness

AbstractAimMapping tree species richness across the tropics is of great interest for effective conservation and biodiversity management. In this study, we evaluated the potential of full‐waveform lidar data for mapping tree species richness across the tropics by relating measurements of vertical canopy structure, as a proxy for the occupation of vertical niche space, to tree species richness.LocationTropics.Time periodPresent.Major taxa studiedTrees.MethodsFirst, we evaluated the characteristics of vertical canopy structure across 15 study sites using (simulated) large‐footprint full‐waveform lidar data (22 m diameter) and related these findings to in‐situ tree species information. Then, we developed structure–richness models at the local (within 25–50 ha plots), regional (biogeographical regions) and pan‐tropical scale at three spatial resolutions (1.0, 0.25 and 0.0625 ha) using Poisson regression.ResultsThe results showed a weak structure–richness relationship at the local scale. At the regional scale (within a biogeographical region) a stronger relationship between canopy structure and tree species richness across different tropical forest types was found, for example across Central Africa and in South America [R2 ranging from .44–.56, root mean squared difference as a percentage of the mean (RMSD%) ranging between 23–61%]. Modelling the relationship pan‐tropically, across four continents, 39% of the variation in tree species richness could be explained with canopy structure alone (R2 = .39 and RMSD% = 43%, 0.25‐ha resolution).Main conclusionsOur results may serve as a basis for the future development of a set of structure–richness models to map high resolution tree species richness using vertical canopy structure information from the Global Ecosystem Dynamics Investigation (GEDI). The value of this effort would be enhanced by access to a larger set of field reference data for all tropical regions. Future research could also support the use of GEDI data in frameworks using environmental and spectral information for modelling tree species richness across the tropics.