Contrasting photoprotective responses of forest trees revealed using PRI light responses sampled with airborne imaging spectrometry.

The Photochemical Reflectance Index (PRI) provides an optical indicator of photosynthetic light-use efficiency, photoprotection, and stress in plants. Although PRI can be applied in remote sensing, its interpretation depends on irradiance, which is hard to obtain from satellite or airborne imagery. To quantify forest photoprotective responses remotely, we developed a framework for modeling and interpreting PRI-light responses of individual trees and species using airborne imaging spectrometry coupled with georeferenced forest inventory data from a temperate broad-leaved forest. We derived an irradiance proxy, used hierarchical modeling to analyze PRI-light responses, and developed a framework of physiological interpretations of model parameters as facultative and constitutive components of photoprotection. Photochemical Reflectance Index declined with illumination, and PRI-light relationships varied with landscape position and among tree crowns and species. More sun-exposed foliage had lower intercepts and slopes of the relationship, indicating greater constitutive, but less facultative, photoprotection. We show that tree photoprotective strategies can be quantified at multiple scales using airborne hyperspectral data in structurally complex forests. Our findings and approach have important implications for the remote sensing of forest stress by offering a new way to assess functional diversity through dynamic differences in photoprotection and photosynthetic downregulation and providing previsual indicators of forest stress.

Integrating thermal infrared stream temperature imagery and spatial stream network models to understand natural spatial thermal variability in streams.

Under a warmer future climate, thermal refuges could facilitate the persistence of species relying on cold-water habitat. Often these refuges are small and easily missed or smoothed out by averaging in models. Thermal infrared (TIR) imagery can provide empirical water surface temperatures that capture these features at a high spatial resolution (<1 m) and over tens of kilometers. Our study examined how TIR data could be used along with spatial stream network (SSN) models to characterize thermal regimes spatially in the Middle Fork John Day (MFJD) River mainstem (Oregon, USA). We characterized thermal variation in seven TIR longitudinal temperature profiles along the MFJD mainstem and compared them with SSN model predictions of stream temperature (for the same time periods as the TIR profiles). TIR profiles identified reaches of the MFJD mainstem with consistently cooler temperatures across years that were not consistently captured by the SSN prediction models. SSN predictions along the mainstem identified ~80% of the 1-km reach scale temperature warming or cooling trends observed in the TIR profiles. We assessed whether landscape features (e.g., tributary junctions, valley confinement, geomorphic reach classifications) could explain the fine-scale thermal heterogeneity in the TIR profiles (after accounting for the reach-scale temperature variability predicted by the SSN model) by fitting SSN models using the TIR profile observation points. Only the distance to the nearest upstream tributary was identified as a statistically significant landscape feature for explaining some of the thermal variability in the TIR profile data. When combined, TIR data and SSN models provide a data-rich evaluation of stream temperature captured in TIR imagery and a spatially extensive prediction of the network thermal diversity from the outlet to the headwaters.

Multi-temporal assessment of grassland α- and ß-diversity using hyperspectral imaging.

While more and more studies are exploring the application of remote sensing in assessing biodiversity for different ecosystems, most consider biodiversity at one point in time. Using several remote-sensing-based metrics, we asked how well remote sensing can detect biodiversity (both α- and ß-diversity) in a prairie grassland across time using airborne hyperspectral data collected in two successive years (2017 and 2018) and at different periods in the growing season (2018). The ability to detect biodiversity using "spectral diversity" and "spectral species" types indeed varied significantly over a 2-yr timespan. Toward the end of the growing season in 2018, the relationship between field- and remote-sensing-based α- and ß-diversity weakened compared to data collected from the same season in the previous year. This contrasting pattern between the two years was likely influenced by prescribed fire, altered weather, and the resulting shifting species composition and phenology. These findings indicate that direct detection of α- and ß-diversity in grasslands should be multi-temporal when possible and should consider the effect of disturbances, climate variables, and phenology. We demonstrate an essential role for airborne platforms in developing a global biodiversity monitoring system involving forthcoming space-borne hyperspectral sensors.

A Flight Direction Design Method for Airborne Spectral Imaging Considering the Anisotropy Reflectance of the Target in Rugged Terrain.

An excellent mission plan is the prerequisite for the acquisition of high quality airborne hyperspectral images which are useful for environmental research, mining etc. In order to minimize the radiance non-uniformity caused by the anisotropic reflectance of targets, the flight direction is mostly designed on the solar azimuth or 180° from it for whiskbroom and pushbroom imagers. However, the radiance to the observer is determined not only by the reflectance of the target, but also by the terrain, the illumination direction and the observation direction. So, the flight direction which is defined to minimize radiance non-uniformity might change with the terrain. In order to find the best flight direction for rugged terrain, we firstly analyze the causes of the effect of terrain on radiation non-uniformity based on the radiative transfer process. Then, the flight direction design method is proposed for composite sloping terrain. Tested by digital and physical simulation experiments, the radiance non-uniformity is minimized when the aircraft flies in the designated direction. Finally, a workflow for flight direction planning and optimizing is summarized, considering the flight mission planning techniques and the workflow of remote sensing missions.

Development of a Cost-Effective Airborne Remote Sensing System for Coastal Monitoring.

Coastal lands and nearshore marine areas are productive and rapidly changing places. However, these areas face many environmental challenges related to climate change and human-induced impacts. Space-borne remote sensing systems may be restricted in monitoring these areas because of their spatial and temporal resolutions. In situ measurements are also constrained from accessing the area and obtaining wide-coverage data. In these respects, airborne remote sensing sensors could be the most appropriate tools for monitoring these coastal areas. In this study, a cost-effective airborne remote sensing system with synthetic aperture radar and thermal infrared sensors was implemented to survey coastal areas. Calibration techniques and geophysical model algorithms were developed for the airborne system to observe the topography of intertidal flats, coastal sea surface current, sea surface temperature, and submarine groundwater discharge.

Automatic detection of standing dead trees based on improved YOLOv7 from airborne remote sensing imagery.

Detecting and localizing standing dead trees (SDTs) is crucial for effective forest management and conservation. Due to challenges posed by mountainous terrain and road conditions, conducting a swift and comprehensive survey of SDTs through traditional manual inventory methods is considerably difficult. In recent years, advancements in deep learning and remote sensing technology have facilitated real-time and efficient detection of dead trees. Nevertheless, challenges persist in identifying individual dead trees in airborne remote sensing images, attributed to factors such as small target size, mutual occlusion and complex backgrounds. These aspects collectively contribute to the increased difficulty of detecting dead trees at a single-tree scale. To address this issue, the paper introduces an improved You Only Look Once version 7 (YOLOv7) model that incorporates the Simple Parameter-Free Attention Module (SimAM), an unparameterized attention mechanism. This improvement aims to enhance the network's feature extraction capabilities and increase the model's sensitivity to small target dead trees. To validate the superiority of SimAM_YOLOv7, we compared it with four widely adopted attention mechanisms. Additionally, a method to enhance model robustness is presented, involving the replacement of the Complete Intersection over Union (CIoU) loss in the original YOLOv7 model with the Wise-IoU (WIoU) loss function. Following these, we evaluated detection accuracy using a self-developed dataset of SDTs in forests. The results indicate that the improved YOLOv7 model can effectively identify dead trees in airborne remote sensing images, achieving precision, recall and mAP@0.5 values of 94.31%, 93.13% and 98.03%, respectively. These values are 3.67%, 2.28% and 1.56% higher than those of the original YOLOv7 model. This improvement model provides a convenient solution for forest management.

Data science competition for cross-site individual tree species identification from airborne remote sensing data.

Data on individual tree crowns from remote sensing have the potential to advance forest ecology by providing information about forest composition and structure with a continuous spatial coverage over large spatial extents. Classifying individual trees to their taxonomic species over large regions from remote sensing data is challenging. Methods to classify individual species are often accurate for common species, but perform poorly for less common species and when applied to new sites. We ran a data science competition to help identify effective methods for the task of classification of individual crowns to species identity. The competition included data from three sites to assess each methods' ability to generalize patterns across two sites simultaneously and apply methods to an untrained site. Three different metrics were used to assess and compare model performance. Six teams participated, representing four countries and nine individuals. The highest performing method from a previous competition in 2017 was applied and used as a baseline to understand advancements and changes in successful methods. The best species classification method was based on a two-stage fully connected neural network that significantly outperformed the baseline random forest and gradient boosting ensemble methods. All methods generalized well by showing relatively strong performance on the trained sites (accuracy = 0.46-0.55, macro F1 = 0.09-0.32, cross entropy loss = 2.4-9.2), but generally failed to transfer effectively to the untrained site (accuracy = 0.07-0.32, macro F1 = 0.02-0.18, cross entropy loss = 2.8-16.3). Classification performance was influenced by the number of samples with species labels available for training, with most methods predicting common species at the training sites well (maximum F1 score of 0.86) relative to the uncommon species where none were predicted. Classification errors were most common between species in the same genus and different species that occur in the same habitat. Most methods performed better than the baseline in detecting if a species was not in the training data by predicting an untrained mixed-species class, especially in the untrained site. This work has highlighted that data science competitions can encourage advancement of methods, particularly by bringing in new people from outside the focal discipline, and by providing an open dataset and evaluation criteria from which participants can learn.

Fusion neural networks for plant classification: learning to combine RGB, hyperspectral, and lidar data.

Airborne remote sensing offers unprecedented opportunities to efficiently monitor vegetation, but methods to delineate and classify individual plant species using the collected data are still actively being developed and improved. The Integrating Data science with Trees and Remote Sensing (IDTReeS) plant identification competition openly invited scientists to create and compare individual tree mapping methods. Participants were tasked with training taxon identification algorithms based on two sites, to then transfer their methods to a third unseen site, using field-based plant observations in combination with airborne remote sensing image data products from the National Ecological Observatory Network (NEON). These data were captured by a high resolution digital camera sensitive to red, green, blue (RGB) light, hyperspectral imaging spectrometer spanning the visible to shortwave infrared wavelengths, and lidar systems to capture the spectral and structural properties of vegetation. As participants in the IDTReeS competition, we developed a two-stage deep learning approach to integrate NEON remote sensing data from all three sensors and classify individual plant species and genera. The first stage was a convolutional neural network that generates taxon probabilities from RGB images, and the second stage was a fusion neural network that "learns" how to combine these probabilities with hyperspectral and lidar data. Our two-stage approach leverages the ability of neural networks to flexibly and automatically extract descriptive features from complex image data with high dimensionality. Our method achieved an overall classification accuracy of 0.51 based on the training set, and 0.32 based on the test set which contained data from an unseen site with unknown taxa classes. Although transferability of classification algorithms to unseen sites with unknown species and genus classes proved to be a challenging task, developing methods with openly available NEON data that will be collected in a standardized format for 30 years allows for continual improvements and major gains for members of the computational ecology community. We outline promising directions related to data preparation and processing techniques for further investigation, and provide our code to contribute to open reproducible science efforts.

Adaptive capacity in the foundation tree species Populus fremontii: implications for resilience to climate change and non-native species invasion in the American Southwest.

Populus fremontii (Fremont cottonwood) is recognized as one of the most important foundation tree species in the southwestern USA and northern Mexico because of its ability to structure communities across multiple trophic levels, drive ecosystem processes and influence biodiversity via genetic-based functional trait variation. However, the areal extent of P. fremontii cover has declined dramatically over the last century due to the effects of surface water diversions, non-native species invasions and more recently climate change. Consequently, P. fremontii gallery forests are considered amongst the most threatened forest types in North America. In this paper, we unify four conceptual areas of genes to ecosystems research related to P. fremontii's capacity to survive or even thrive under current and future environmental conditions: (i) hydraulic function related to canopy thermal regulation during heat waves; (ii) mycorrhizal mutualists in relation to resiliency to climate change and invasion by the non-native tree/shrub, Tamarix; (iii) phenotypic plasticity as a mechanism for coping with rapid changes in climate; and (iv) hybridization between P. fremontii and other closely related Populus species where enhanced vigour of hybrids may preserve the foundational capacity of Populus in the face of environmental change. We also discuss opportunities to scale these conceptual areas from genes to the ecosystem level via remote sensing. We anticipate that the exploration of these conceptual areas of research will facilitate solutions to climate change with a foundation species that is recognized as being critically important for biodiversity conservation and could serve as a model for adaptive management of arid regions in the southwestern USA and around the world.

Performance of three reflectance calibration methods for airborne hyperspectral spectrometer data.

In this study, the performances and accuracies of three methods for converting airborne hyperspectral spectrometer data to reflectance factors were characterized and compared. The "reflectance mode (RM)" method, which calibrates a spectrometer against a white reference panel prior to mounting on an aircraft, resulted in spectral reflectance retrievals that were biased and distorted. The magnitudes of these bias errors and distortions varied significantly, depending on time of day and length of the flight campaign. The "linear-interpolation (LI)" method, which converts airborne spectrometer data by taking a ratio of linearly-interpolated reference values from the preflight and post-flight reference panel readings, resulted in precise, but inaccurate reflectance retrievals. These reflectance spectra were not distorted, but were subject to bias errors of varying magnitudes dependent on the flight duration length. The "continuous panel (CP)" method uses a multi-band radiometer to obtain continuous measurements over a reference panel throughout the flight campaign, in order to adjust the magnitudes of the linear-interpolated reference values from the preflight and post-flight reference panel readings. Airborne hyperspectral reflectance retrievals obtained using this method were found to be the most accurate and reliable reflectance calibration method. The performances of the CP method in retrieving accurate reflectance factors were consistent throughout time of day and for various flight durations. Based on the dataset analyzed in this study, the uncertainty of the CP method has been estimated to be 0.0025 ± 0.0005 reflectance units for the wavelength regions not affected by atmospheric absorptions. The RM method can produce reasonable results only for a very short-term flight (e.g., < 15 minutes) conducted around a local solar noon. The flight duration should be kept shorter than 30 minutes for the LI method to produce results with reasonable accuracies. An important advantage of the CP method is that the method can be used for long-duration flight campaigns (e.g., 1-2 hours). Although this study focused on reflectance calibration of airborne spectrometer data, the methods evaluated in this study and the results obtained are directly applicable to ground spectrometer measurements.

A data science challenge for converting airborne remote sensing data into ecological information.

Ecology has reached the point where data science competitions, in which multiple groups solve the same problem using the same data by different methods, will be productive for advancing quantitative methods for tasks such as species identification from remote sensing images. We ran a competition to help improve three tasks that are central to converting images into information on individual trees: (1) crown segmentation, for identifying the location and size of individual trees; (2) alignment, to match ground truthed trees with remote sensing; and (3) species classification of individual trees. Six teams (composed of 16 individual participants) submitted predictions for one or more tasks. The crown segmentation task proved to be the most challenging, with the highest-performing algorithm yielding only 34% overlap between remotely sensed crowns and the ground truthed trees. However, most algorithms performed better on large trees. For the alignment task, an algorithm based on minimizing the difference, in terms of both position and tree size, between ground truthed and remotely sensed crowns yielded a perfect alignment. In hindsight, this task was over simplified by only including targeted trees instead of all possible remotely sensed crowns. Several algorithms performed well for species classification, with the highest-performing algorithm correctly classifying 92% of individuals and performing well on both common and rare species. Comparisons of results across algorithms provided a number of insights for improving the overall accuracy in extracting ecological information from remote sensing. Our experience suggests that this kind of competition can benefit methods development in ecology and biology more broadly.

Remote sensing tree classification with a multilayer perceptron.

To accelerate scientific progress on remote tree classification-as well as biodiversity and ecology sampling-The National Institute of Science and Technology created a community-based competition where scientists were invited to contribute informatics methods for classifying tree species and genus using crown-level images of trees. We classified tree species and genus at the pixel level using hyperspectral and LiDAR observations. We compared three algorithms that have been implemented extensively across a broad range of research applications: support vector machines, random forests, and multilayer perceptron. At the pixel level, the multilayer perceptron algorithm classified species or genus with high accuracy (92.7% and 95.9%, respectively) on the training data and performed better than the other two algorithms (85.8-93.5%). This indicates promise for the use of the multilayer perceptron (MLP) algorithm for tree-species classification based on hyperspectral and LiDAR observations and coincides with a growing body of research in which neural network-based algorithms outperform other types of classification algorithm for machine vision. To aggregate patterns across the images, we used an ensemble approach that averages the pixel-level outputs of the MLP algorithm to classify species at the crown level. The average accuracy of these classifications on the test set was 68.8% for the nine species.

ASPIS, A Flexible Multispectral System for Airborne Remote Sensing Environmental Applications.

Airborne multispectral and hyperspectral remote sensing is a powerful tool for environmental monitoring applications. In this paper we describe a new system (ASPIS) composed by a 4-CCD spectral sensor, a thermal IR camera and a laser altimeter that is mounted on a flexible Sky-Arrow airplane. A test application of the multispectral sensor to estimate durum wheat quality is also presented.

Detection of single graves by airborne hyperspectral imaging.

Airborne hyperspectral imaging (HSI) was assessed as a potential tool to locate single grave sites. While airborne HSI has shown to be useful to locate mass graves, it is expected the location of single graves would be an order of magnitude more difficult due to the smaller size and reduced mass of the targets. Two clearings were evaluated (through a blind test) as potential sites for containing at least one set of buried remains. At no time prior to submitting the locations of the potential burial sites from the HSI were the actual locations of the sites released or shared with anyone from the analysis team. The two HSI sensors onboard the aircraft span the range of 408-2524nm. A range of indicators that exploit the narrow spectral and spatial resolutions of the two complimentary HSI sensors onboard the aircraft were calculated. Based on the co-occurrence of anomalous pixels within the expected range of the indicators three potential areas conforming to our underlying assumptions of the expected spectral responses (and spatial area) were determined. After submission of the predicted burial locations it was revealed that two of the targets were located within GPS error (10m) of the true burial locations. Furthermore, due to the history of the TPOF site for burial work, investigation of the third target is being considered in the near future. The results clearly demonstrate promise for hyperspectral imaging to aid in the detection of buried remains, however further work is required before these results can justifiably be used in routine scenarios.