Fusing Sentinel-1 and -2 to Model GEDI-Derived Vegetation Structure Characteristics in GEE for the Paraguayan Chaco

Vegetation structure is a key component in assessing habitat quality for wildlife and carbon storage capacity of forests. Studies conducted at global scale demonstrate the increasing pressure of the agricultural frontier on tropical forest, endangering their continuity and biodiversity within. The Paraguayan Chaco has been identified as one of the regions with the highest rate of deforestation in South America. Uninterrupted deforestation activities over the last 30 years have resulted in the loss of 27% of its original cover. The present study focuses on the assessment of vegetation structure characteristics for the complete Paraguayan Chaco by fusing Sentinel-1, -2 and novel spaceborne Light Detection and Ranging (LiDAR) samples from the Global Ecosystem Dynamics Investigation (GEDI). The large study area (240,000 km²) calls for a workflow in the cloud computing environment of Google Earth Engine (GEE) which efficiently processes the multi-temporal and multi-sensor data sets for extrapolation in a tile-based random forest (RF) regression model. GEDI-derived attributes of vegetation structure are available since December 2019, opening novel research perspectives to assess vegetation structure composition in remote areas and at large-scale. Therefore, the combination of global mapping missions, such as Landsat and Sentinel, are predestined to be combined with GEDI data, in order to identify priority areas for nature conservation. Nevertheless, a comprehensive assessment of the vegetation structure of the Paraguayan Chaco has not been conducted yet. For that reason, the present methodology was developed to generate the first high-resolution maps (10 m) of canopy height, total canopy cover, Plant-Area-Index and Foliage-Height-Diversity-Index. The complex ecosystems of the Paraguayan Chaco ranging from arid to humid climates can be described by canopy height values from 1.8 to 17.6 m and canopy covers from sparse to dense (total canopy cover: 0 to 78.1%). Model accuracy according to median R² amounts to 64.0% for canopy height, 61.4% for total canopy cover, 50.6% for Plant-Area-Index and 48.0% for Foliage-Height-Diversity-Index. The generated maps of vegetation structure should promote environmental-sound land use and conservation strategies in the Paraguayan Chaco, to meet the challenges of expanding agricultural fields and increasing demand of cattle ranching products, which are dominant drivers of tropical forest loss.

Phenology of fine roots and shoots using high frequency temporal resolution images in a temperate larch forest

Purpose Understanding tree phenology reveals the underlying mechanisms through plant functional and productive activities and carbon sinks in forest ecosystems. However, previous research on tree phenology has focused on shoot dynamics rather than tree root dynamics. We aimed to explore seasonal temperature patterns of daily-based root and shoot dynamics by capturing high frequency plant images in a larch forest. Methods We monitored continuous images using an automated digital camera for shoot dynamics and a flatbed scanner for the fine root dynamics in the larch. Using the images, we analyzed the relationship between temperature and plant area index as shoot growth status and total root-area proportion of white and brown roots. Results Larch shoot production had a single mountain-shaped peak with a positive correlation between plant area index and air temperature. Fine root production had two peaks in the bimodal root-growth pattern in early summer and late autumn. Soil temperature was positively correlated with white root proportion and negatively correlated with brown root proportion. Conclusion We found differences between shoots and roots regarding temperature relationships. In particular, the automated flatbed scanner method for the root dynamics allowed the collection of detailed bimodal patterns of root production with shift from whitening to browning color, which had been previously overlooked. Such high frequency temporal resolution analysis can provide an in-depth of mechanisms of fine-root and shoot phenology through different stages of plant development in terms of growth and senescence.

Electromagnetic Induction Is a Fast and Non-Destructive Approach to Estimate the Influence of Subsurface Heterogeneity on Forest Canopy Structure

The spatial forest structure that drives the functioning of these ecosystems and their response to global change is closely linked to edaphic conditions. However, the latter properties are particularly difficult to characterize in forest areas developed on karst, where soil is highly rocky and heterogeneous. In this work, we investigated whether geophysics, and more specifically electromagnetic induction (EMI), can provide a better understanding of forest structure. We use EMI (EM31, Geonics Limited, Ontario, Canada) to study the spatial variability of ground properties in two different Mediterranean forests. A naturally post-fire regenerated forest composed of Aleppo pines and Holm oaks and a monospecific plantation of Altlas cedar. To better interpret EMI results, we used electrical resistivity tomography (ERT), soil depth surveys, and field observations. Vegetation was also characterized using hemispherical photographs that allowed to calculate plant area index (PAI). Our results show that the variability of ground properties contribute to explaining the variability in the vegetation cover development (plant area index). Vegetation density is higher in areas where the soil is deeper. We showed a significant correlation between edaphic conditions and tree development in the naturally regenerated forest, but this relationship is clearly weaker in the cedar plantation. We hypothesized that regular planting after subsoiling, as well as sylvicultural practices (thinning and pruning) influenced the expected relationship between vegetation structure and soil conditions measured by EMI. This work opens up new research avenues to better understand the interplay between soil and subsoil variability and forest response to climate change.

Airborne and Spaceborne Lidar Reveal Trends and Patterns of Functional Diversity in a Semi-Arid Ecosystem

Assessing functional diversity and its abiotic controls at continuous spatial scales are crucial to understanding changes in ecosystem processes and services. Semi-arid ecosystems cover large portions of the global terrestrial surface and provide carbon cycling, habitat, and biodiversity, among other important ecosystem processes and services. Yet, the spatial trends and patterns of functional diversity in semi-arid ecosystems and their abiotic controls are unclear. The objectives of this study are two-fold. We evaluated the spatial pattern of functional diversity as estimated from small footprint airborne lidar (ALS) with respect to abiotic controls and fire in a semi-arid ecosystem. Secondly, we used our results to understand the capabilities of large footprint spaceborne lidar (GEDI) for future applications to semi-arid ecosystems. Overall, our findings revealed that functional diversity in this ecosystem is mainly governed by elevation, soil, and water availability. In burned areas, the ALS data show a trend of functional recovery with time since fire. With 16 months of data (April 2019-August 2020), GEDI predicted functional traits showed a moderate correlation (r = 41–61%) with the ALS predicted traits except for the plant area index (PAI) (r = 11%) of low height vegetation (<5 m). We found that the number of GEDI footprints relative to the size of the fire-disturbed areas (=< 2 km2) limited the ability to estimate the full effects of fire disturbance. However, the consistency of diversity trends between ALS and GEDI across our study area demonstrates GEDI’s potential of capturing functional diversity in similar semi-arid ecosystems. The capability of spaceborne lidar to map trends and patterns of functional diversity in this semi-arid ecosystem demonstrates its exciting potential to identify critical biophysical and ecological shifts. Furthermore, opportunities to fuse GEDI with complementary spaceborne data such as ICESat-2 or the upcoming NASA-ISRO Synthetic Aperture Radar (NISAR), and fine scale airborne data will allow us to fill gaps across space and time. For the first time, we have the potential to monitor carbon cycle dynamics, habitats and biodiversity across the globe in semi-arid ecosystems at fine vertical scales.

Plant and Wood Area Index of Solitary Trees for Urban Contexts in Nordic Cities

Background: We present the plant area index (PAI) measurements taken for 63 deciduous broadleaved tree species and 1 deciduous conifer tree species suitable for urban areas in Nordic cities. The aim was to evaluate PAI and wood area index (WAI) of solitary-grown broadleaved tree species and cultivars of the same age in order to present a data resource of individual tree characteristics viewed in summer (PAI) and in winter (WAI). Methods: All trees were planted as individuals in 2001 at the Hørsholm Arboretum in Denmark. The field method included a Digital Plant Canopy Imager where each scan and contrast values were set to consistent values. Results: The results illustrate that solitary trees differ widely in their WAI and PAI and reflect the integrated effects of leaf material and the woody component of tree crowns. The indications also show highly significant (P < 0.001) differences between species and genotypes. The WAI had an overall mean of 0.91 (± 0.03), ranging from Tilia platyphyllos ‘Orebro’ with a WAI of 0.32 (± 0.04) to Carpinus betulus ‘Fastigiata’ with a WAI of 1.94 (± 0.09). The lowest mean PAI in the dataset was Fraxinus angustifolia ‘Raywood’ with a PAI of 1.93 (± 0.05), whereas Acer campestre ‘Kuglennar’ represents the cultivar with the largest PAI of 8.15 (± 0.14). Conclusions: Understanding how this variation in crown architectural structure changes over the year can be applied to climate responsive design and microclimate modeling where plant and wood area index of solitary-grown trees in urban contexts are of interest.

Retrieval of Arctic Vegetation Biophysical and Biochemical Properties from CHRIS/PROBA Multi-Angle Imagery Using Empirical and Physical Modelling

Mapping and monitoring of Arctic vegetation biochemical and biophysical properties is gaining importance as global climate change is disproportionately affecting this region. Previous studies using remote sensing to model Arctic vegetation biochemical and biophysical properties have generally involved empirical modelling with nadir looking broadband sensors and have typically been conducted at the field scale in one study area. Satellite hyperspectral remote sensing has not been previously investigated for retrieving leaf and canopy biochemical and biophysical properties of Arctic vegetation across multiple sites using either empirical or physically-based modelling approaches. Furthermore, multi-angle hyperspectral sensors (CHRIS/PROBA), which can provide insight into vegetation reflectance anisotropy and potentially improve vegetation parameter estimation, have also not been investigated for this purpose. In this study, three modelling approaches previously investigated with field spectroscopy data (Kennedy et al., 2020) were used with CHRIS Mode-1 imagery to predict leaf chlorophyll content, plant area index and canopy chlorophyll content across a bioclimatic gradient in the Western Canadian Arctic. Modelling approaches included: parametric linear regression based on vegetation indices (VI), non-parametric machine learning Gaussian processes regression (GPR) and inversion of the PROSAIL radiative transfer model using a look-up table approach (LUT). CHRIS imagery was acquired with −55°, −36°, 0°, +36°, +55° view zenith angles (VZA) between 2011 and 2014 over three field sites extending from the Richardson Mountains in central Yukon, Canada to the north end of Banks Island, Northwest Territories, Canada. Field measurements were acquired within several weeks of satellite acquisitions. GPR had the best model fit (mean cross-validated (cv) coefficient of determination, r2cv = 0.61 across all vegetation variables, sites and VZAs vs. 0.59 for the simple ratio, SR) and predictive performance (normalized root mean square error, NRMSEcv = 0.13 vs. 0.14 for SR). The revised optimized soil adjusted VI (ROSAVI) performance was slightly poorer (r2cv = 0.51; NRMSEcv = 0.15). The physically-based PROSAIL model performed poorer than all empirical models (r2 = 0.50; NRMSE = 0.18). This ranking of model performance is similar to that found in the previous field spectroscopy study, where empirical model fits and predictive performance were only slightly worse. With respect to view angle performance, NRMSE varied only slightly, indicating no distinct advantage for any one VZA. Overall, strong potential has been demonstrated for empirical modelling of Arctic vegetation chlorophyll and plant area index using hyperspectral data combined with band selection/optimization procedures in the Arctic. Recently launched and future hyperspectral satellites, including next generation airborne sensors, will likely provide improvements to the model performance reported here.

Optimizing smartphone-based canopy hemispherical photography

1) Hemispherical photography (HP) is a long-standing tool for forest canopy characterization. Currently, there are low-cost fisheye lenses to convert smartphones into high-portable HP equipment; however, they cannot be used whenever since HP is sensitive to illumination conditions. To obtain sound results outside diffuse light conditions, a deep-learning-based system needs to be developed. A ready-to-use alternative is the multiscale color-based binarization algorithm, but it can provide moderate-quality results only for open forests. To overcome this limitation, I propose coupling it with the model-based local thresholding algorithm. I call this coupling the MBCB approach. 2) Methods presented here are part of the R package CAnopy IMage ANalysis (caiman), which I am developing. The accuracy assessment of the new MBCB approach was done with data from a pine plantation and a broadleaf native forest. 3) The coefficient of determination (R^2) was greater than 0.7, and the root mean square error (RMSE) lower than 20 %, both for plant area index calculation. 4) Results suggest that the new MBCB approach allows the calculation of unbiased canopy metrics from smartphone-based HP acquired in sunlight conditions, even for closed canopies. This facilitates large-scale and opportunistic sampling with hemispherical photography.

UAS mounted LiDAR for Estimating LAI Type Metrics for Winter Wheat

&lt;p&gt;&amp;#160;&lt;/p&gt;&lt;p&gt;Metrics such as Leaf Area Index (LAI) are key factors in agricultural monitoring to understand the health and predictive yield of crops. Knowing the spatial distribution and variability in more detail increases the precision of fertilizer and irrigation practices. Unmanned Aircraft Systems (UAS) provide a means to carry sensors at a low altitude below the clouds providing a much higher spatial and temporal resolution than previously seen with satellite remote sensing while also providing more spatially complete data as compared to ground methods. Being that LiDAR is an active sensor and does not depend on solar reflectance and its corresponding zenith angle like commonly used passive optical sensors, it can further improve upon these UAS characteristics. It can also sense further into the canopy as the laser signals can pass through small gaps and are not affected by the shadowing of plant features created by the canopy itself. Evaluating the penetration of these signals and investigating the gap fraction (GF) that relates to canopy density, we are able to retrieve LAI. However, as LiDAR is sensing all above-ground plant elements it may present the ability to estimate Plant Area Index (PAI) rather than LAI when monitoring an entire growing season for a cereal crop like winter wheat that begins browning during senescence. This study investigates the feasibility of using LiDAR to estimate LAI or similar crop canopy density metrics. As LiDAR sensors for UAS are just becoming more accessible, studies related to this topic are scarcely seen.&lt;/p&gt;&lt;p&gt;In this study, a winter wheat field in Selhausen, Germany (~10 ha in size) was monitored throughout the growing season using the following methods: [1] air campaigns with a DJI Matrice 600 UAS with a YellowScan Surveyor LiDAR system, [2] a DJI Matrice 600 UAS with a Micasense RedEdge-M (five band) multispectral sensor, and [3] ground measurements using a SS1 SunScan ceptometer. The resulting LAI type metrics of the UAS LiDAR methods used were compared to methods commonly used with multispectral (MS) and ground instruments to assess the proposed method&amp;#8217;s potential. Additionally, because both products are spatially complete unlike the ground measurements, the LiDAR and multispectral methods were compared for similarities in spatial patterns.&lt;/p&gt;&lt;p&gt;The results showed promise in using UAS LiDAR to estimate metrics that relate to LAI. Pearson correlation coefficient between the LiDAR and multispectral methods were moderate to high (R= 0.39 &amp;#8211; 0.66) over the growing season. The comparison of UAS LiDAR towards the ground reference was within a 3% difference at times before senescence. Later in the growing season, the discrepancy increased between LiDAR and MS sensor retrievals mainly because of plant browning related to changes in plant chlorophyll content. This study covers the benefits of using UAS mounted LiDAR for LAI related measurements and its potential for improving crop health monitoring for precision farming.&lt;/p&gt;

Estimating Canopy Density Parameters Time-Series for Winter Wheat Using UAS Mounted LiDAR

Monitoring of canopy density with related metrics such as leaf area index (LAI) makes a significant contribution to understanding and predicting processes in the soil–plant–atmosphere system and to indicating crop health and potential yield for farm management. Remote sensing methods using optical sensors that rely on spectral reflectance to calculate LAI have become more mainstream due to easy entry and availability. Methods with vegetation indices (VI) based on multispectral reflectance data essentially measure the green area index (GAI) or response to chlorophyll content of the canopy surface and not the entire aboveground biomass that may be present from non-green elements that are key to fully assessing the carbon budget. Methods with light detection and ranging (LiDAR) have started to emerge using gap fraction (GF) to estimate the plant area index (PAI) based on canopy density. These LiDAR methods have the main advantage of being sensitive to both green and non-green plant elements. They have primarily been applied to forest cover with manned airborne LiDAR systems (ALS) and have yet to be used extensively with crops such as winter wheat using LiDAR on unmanned aircraft systems (UAS). This study contributes to a better understanding of the potential of LiDAR as a tool to estimate canopy structure in precision farming. The LiDAR method proved to have a high to moderate correlation in spatial variation to the multispectral method. The LiDAR-derived PAI values closely resemble the SunScan Ceptometer GAI ground measurements taken early in the growing season before major stages of senescence. Later in the growing season, when the canopy density was at its highest, a possible overestimation may have occurred. This was most likely due to the chosen flight parameters not providing the best depictions of canopy density with consideration of the LiDAR’s perspective, as the ground-based destructive measurements provided lower values of PAI. Additionally, a distinction between total LiDAR-derived PAI, multispectral-derived GAI, and brown area index (BAI) is made to show how the active and passive optical sensor methods used in this study can complement each other throughout the growing season.