Optimizing fruit quality and yield in agriculture requires accurately monitoring leaf nitrogen (N) status spatially and temporally throughout the growing season. Standard remote sensing approaches for assessing leaf N rely on proxies like vegetation indices or leaf chlorophyll a + b (Cab) content. However, limitations exist due to the Cab-N relationship’s saturation and early nutrient deficiency insensitivity. The study utilized Sentinel-2 satellite imagery to estimate a set of plant biochemical traits in large almond orchards in a two-year study. These traits, including leaf dry matter, leaf water content, and leaf Cab retrieved from the radiative transfer model, were used to explain the observed variability of leaf N. Airborne hyperspectral imagery-derived leaf N using Cab and solar-induced fluorescence served as a benchmark for validation. Results demonstrate that plant traits quantified from Sentinel-2 were strongly associated with leaf N variability across the orchard, with a strong contribution from the estimated leaf Cab content and leaf dry matter biochemical constituent, outperforming the consistency of vegetation indices. The Sentinel-2 model explaining leaf N variability yielded r2 = 0.82 and nRMSE = 13

Among all essential nutrients, nitrogen (N) is required by plants in large quantities throughout the entire developmental process. This is due to its importance for plant growth and development and as a primary source of energy for photosynthesis. Previous research has demonstrated that solar-induced chlorophyll fluorescence (SIF) coupled with chlorophyll a+b content (C ab ) improved the estimation of leaf N, outperforming standard vegetation indices. The present study investigates the contribution of leaf C x , a measure of the de-epoxidation state of the xanthophyll cycle, for explaining leaf N variability, concluding that it ranks third after C ab and SIF consistently over two growing seasons. Among the rest of the biochemical constituents estimated by model inversion, C x contributed more than anthocyanins (Anth), the total carotenoid content (C car ), and crown-level structural traits.

Accurate, spatially extensive, and frequent assessments of plant nitrogen (N) enabled by remote sensing allow growers to optimize fertilizer applications and reduce environmental impacts. Standard remote sensing methods for N assessment typically involve the use of chlorophyll-sensitive vegetation indices calculated from multispectral or hyperspectral reflectance data. However, the chlorophyll a + b derived from spectral indices is indirectly related to leaf N and saturates at high leaf N levels, dramatically reducing the sensitivity with leaf N under these conditions. Furthermore, these relationships are heavily influenced by canopy structure, variability in leaf area density, proportion of sunlit-shaded tree-crown components, soil background, and understory. Recent studies in uniform crops have demonstrated that estimation of plant N can be improved by considering leaf biochemical constituents derived from radiative transfer model (RTM) and solar-induced fluorescence (SIF). However, it is unclear whether these methods are transferable to tree crops due to their intrinsic physiological differences, structural complexity, and within-tree crown heterogeneity. We investigated how various hyperspectrally derived proxies for leaf N, including RTM-based traits and SIF, could be combined to assess N status on a 1200-ha almond orchard across two growing seasons. RTM-based chlorophyll a + b content (Cab) and SIF were found to be the most important and consistent predictors for leaf N compared to other leaf biochemical and biophysical traits. Cab alone was a modest predictor of leaf N variability (r2 = 0.49, RMSE = 0.16%, p-value <0.001), but when the non-collinear SIF and Cab traits were coupled together, predictions improved dramatically (r2 = 0.95, RMSE = 0.05%, p-value <0.001). Leaf area index (LAI) was poorly associated with leaf N, suggesting that leaf physiological traits may be more important than structural traits in quantifying leaf N in well-managed orchards characterized by high N levels. Consistent results across the 2 years suggests the importance of airborne SIF coupled with Cab for precision agriculture and leaf N status assessment in almond orchards.

Traditional methods to estimate leaf nitrogen (N) from satellite imagery rely on structural and chlorophyll $a+b\,(\mathrm{C}_{\text{ab}})$ vegetation indices. Recent progress with airborne hyperspectral imagery identified Cab and SIF as critical indicators for evaluating leaf N variability, yielding superior performance than standard vegetation indices. In tree orchards, accurate physiological assessments require high-spatial-resolution hyperspectral imagery to minimize canopy architecture and soil background effects. Understanding the potential of coarse-spatial-resolution spaceborne hyperspectral imagery for leaf N estimation is critical. In this study, DESIS hyperspectral imagery collected on board the International Space Station was used to assess the quantification of leaf N, evaluating the relative contributions of physiological plant traits and SIF. High-resolution airborne hyperspectral imagery and ground N data were used for validation. Results show that Cab and SIF were the most critical parameters explaining leaf N both from DESIS and from airborne hyperspectral imagery, yielding strong correlations against ground truth N data ( $r^{2}=0.90, p < 0.0001$ ) and with airborne-predicted $\mathrm{N}\,(r^{2}=0.75, p < 0.0001)$ .

Standard remote sensing methods for nitrogen (N) assessment in precision agriculture rely on empirical relationships built with chlorophyll a+b (Cab) sensitive vegetation indices. Nevertheless, methods of N estimation based on the Cab vs. N relationships are strongly affected by the saturation of these indices at high N levels, and by canopy structure, shadows and soil background variability. These effects are even more pronounced in heterogeneous orchards where the tree crown structural variability is a major factor that limits the transferability of the algorithms within- and across-tree crop species. Solar-induced fluorescence (SIF) has been proposed in precision agriculture as a plant functional trait related to N due to its link with photosynthesis. However, retrieving SIF from orchards is challenging due to the mixture of sunlit and shaded crown components. The present study explored the retrieval of airborne SIF in almond orchards from hyperspectral imagery, assessing its contribution to the estimation of N. Results show that the assessment of N improved when SIF was coupled to the model estimated Cab (e.g., Cab+SIF; r2=0.95) as compared with using Cab alone (r 2 =0.87).

Together with ensuring a stable yield, improving grape composition and aroma is the main goal of wine grape production management as it determines consumer acceptance and ultimately revenue. Understanding the triggers of the synthesis of aromatic components and finding methods to map their variability in the field can aid management practices during the season and planning selective harvest in views of maximizing benefit. Vegetation indices have been shown to track grape colour, sugar and acidity content but it has been demonstrated that aromatic components are the main drivers of the final palate of wine and are not correlated to sugar concentration. Leaf pigments such as chlorophyll, carotenoids and anthocyanins are involved in the metabolic pathways of aroma compounds in grapes. The physiological connections between grape aromatic components and primary and secondary photosynthetic pigments suggest that they could be used to detect processes related to aroma composition. This study investigates the links between grape quality parameters such as aromatic components and image-quantified spectral indices and photosynthetic plant traits derived by physical model inversion methods. Two sets of high-spatial resolution hyperspectral and thermal imagery were collected with an unmanned platform at veraison and harvest. The variability found in the field was partly but not fully explained by the thermal-based crop water stress index as an indicator of water stress (r(2) = 0.51-0.58, p-value<0.01). Fluspect-CX leaf model was coupled to 4SAIL canopy model and inverted to map the main photosynthetic pigment groups and the fraction of pigments acting in photoprotection. Results obtained through radiative transfer model inversion outperformed traditional vegetation indices related to pigment content and degradation. We found statistically significant relationships between image-retrieved pigments and terpenoids responsible for wine aroma (p-value<0.005).

Background Earth Observation ‘EO’ remote sensing technology development enables original insights into vegetation function and health at ever finer temporal, spectral and spatial resolution. Research sites equipped with monitoring infrastructure such as flux towers operate at a key bridging scale between satellite platform measurements and on-the-ground leaf-level processes. Results This paper presents the technical details of the design and operation of a proximal observation system ‘THEMS’ that generates unattended long-term high quality thermal and hyperspectral images of a forest canopy on a short (sub-daily) timescale. The primary purpose of the system is to measure canopy temperature, spectral reflectance and radiance coincident with a highly instrumented flux tower site for benchmarking purposes. Basic system capability is demonstrated through low level data product descriptions of the high-resolution multi-angular imagery and ancillary data streams. The system has been successfully operational for more than 2 years with little to no intervention. Conclusions These data can then be used to derive remotely sensed proxies of canopy and ecosystem function to study temporal forest dynamics over a wide range of wavelengths, spatial scales (individual trees to canopy), and temporal scales (minutes to multiple years). The multi-purpose system is intended to provide unprecedented spatio-temporal ecophysiological insight and to underpin upscaling of remotely sensed dynamic ecosystem water, CO 2 , and energy exchange processes.

Significant advances toward the remote sensing of photosynthetic activity have been achieved in the last decades, including sensor design and radiative transfer model (RTM) development. Nevertheless, finding methods to accurately quantify carbon assimilation across species and spatial scales remains a challenge. Most methods are either empirical and not transferable across scales or can only be applied if highly complex input data are available. Under stress, the photosynthetic rate is limited by the maximum carboxylation rate (V-cmax), which is determined by the leaf biochemistry and the environmental conditions. V-cmax has been connected to plant photoprotective mechanisms, photosynthetic activity and chlorophyll fluorescence emission. Recent RTM developments such as the Soil-Canopy Observation of Photosynthesis and Energy fluxes (SCOPE) model allow the simulation of the sun-induced chlorophyll fluorescence (SIF) and V-cmax effects on the canopy spectrum. This development provides an approach to retrieve V-cmax through RTM model inversion and track assimilation rate. In this study we explore SIF, narrow-band indices and RTM inversion to track changes in photosynthetic efficiency as a function of vegetation stress. We use hyperspectral imagery acquired over an almond orchard under different management strategies which affected the assimilation rates measured in the field. V-cmax used as an indicator of assimilation was retrieved through SCOPE model inversion from pure-tree crown hyperspectral data. The relationships between field-measured assimilation rates and V-cmax retrieved from model inversion were higher (r(2) = 0.7-0.8) than when SIF was used alone (r(2) = 0.5-0.6) or when traditional vegetation indices were used (r(2) = 0.3-0.5). The method was proved successful when applied to two independent datasets acquired at two different dates throughout the season, ensuring its robustness and transferability. When applied to both dates simultaneously, the results showed a unique significant trend between the assimilation measured in the field and V-cmax derived using SCOPE (r(2) = 0.56, p < 0.001). This work demonstrates that tracking assimilation in almond trees is feasible using hyperspectral imagery linked to radiative transfer-photosynthesis models.

We provide a comprehensive review of the empirical and modelling approaches used to quantify the radiation–vegetation interactions related to vegetation temperature, leaf optical properties linked to pigment absorption and chlorophyll fluorescence emission, and of their capability to monitor vegetation health. Part 1 provides an overview of the main physiological indicators (PIs) applied in remote sensing to detect alterations in plant functioning linked to vegetation diseases and decline processes. Part 2 reviews the recent advances in the development of quantitative methods to assess PI through hyperspectral and thermal images. In recent years, the availability of high-resolution hyperspectral and thermal images has increased due to the extraordinary progress made in sensor technology, including the miniaturization of advanced cameras designed for unmanned aerial vehicle (UAV) systems and lightweight aircrafts. This technological revolution has contributed to the wider use of hyperspectral imaging sensors by the scientific community and industry; it has led to better modelling and understanding of the sensitivity of different ranges of the electromagnetic spectrum to detect biophysical alterations used as early warning indicators of vegetation health. The review deals with the capability of PIs such as vegetation temperature, chlorophyll fluorescence, photosynthetic energy downregulation and photosynthetic pigments detected through remote sensing to monitor the early responses of plants to different stressors. Various methods for the detection of PI alterations have recently been proposed and validated to monitor vegetation health. The greatest challenges for the remote sensing community today are (i) the availability of high spatial, spectral and temporal resolution image data; (ii) the empirical validation of radiation–vegetation interactions; (iii) the upscaling of physiological alterations from the leaf to the canopy, mainly in complex heterogeneous vegetation landscapes; and (iv) the temporal dynamics of the PIs and the interaction between physiological changes.

Plants actively regulate excess absorbed energy to protect photosynthetic machinery through heat dissipation in a process known as non-photochemical quenching (NPQ), a process useful for quantifying plant health and productivity. NPQ can be indirectly measured in the visible wavelengths between 500 nm and 560 nm, most commonly through the Photochemical Reflectance Index (PRI). However, there remains a lack of consensus regarding the optimal functional form and band selection to calculate PRI for the purpose of measuring NPQ mechanisms. Here, we quantitatively evaluate the effectiveness of leaf-level parametric and non-parametric spectral formulations, band locations, and number of bands to track the xanthophyll pigment cycle in a tall mature Eucalypt forest. Subsequently, our recommended approach is the new 'tri-PRI' index robust to constitutive pigment pool sizes across the canopy profile. tri-PRI is a Triangular Vegetation Index (TVI) (tri-PRI = 0.5[(520 - 490)(R-545nm - R-490nm) - (545 - 490)(R-520nm - R-490nm)]) using three reflectance bands around 490 nm, 520 nm and 545 nm, and has a physiological photosynthetic basis. We found that tri-PRI significantly outperformed PRI and other two band combinations for quantifying the xanthophyll EPoxidation State 'EPS' (tri-PRI R-2 = 0.75 versus PRI R-2 = 0.23), as well as the Phi NPQ and Phi PSII active chlorophyll fluorescence quenching yields. The new band placement enhanced the dynamic EPS absorption peak, while the third band provided an additional normalisation to minimise the confounding effects of pigments with overlapping spectral features. tri-PRI also performed comparably to parametric and non-parametric hyperspectral techniques and formulations using continuous spectral regions, highlighting the utility of targeted multispectral indices over hyperspectral approaches. This leaf-level study represents a foundational step toward indirectly measuring dynamic photosynthetic activity across the canopy profile in a tall mature Eucalypt forest to inform upscaling efforts from above-canopy remote sensing platforms. The application of tri-PRI and other top-performing multi-band TVI formulations for predicting EPS presented here should be explored across different canopy types, temporal-, and spatial scales.

The so-called clumping factor (Omega) quantifies deviation from a random 3D distribution of material in a vegetation canopy and therefore characterises the spatial distribution of gaps within a canopy. Omega is essential to convert effective Plant or Leaf Area Index into actual LAI or PAL which has previously been shown to have a significant impact on biophysical parameter retrieval using optical remote sensing techniques in forests, woodlands, and savannas. Here, a simulation framework was applied to assess the performance of existing in situ clumping retrieval methods in a 3D virtual forest canopy, which has a high degree of architectural realism. The virtual canopy was reconstructed using empirical data from a Box Ironbark Eucalypt forest in Eastern Australia. Hemispherical photography (HP) was assessed due to its ubiquity for indirect LAI and structure retrieval. Angular clumping retrieval method performance was evaluated using a range of structural configurations based on varying stem distribution and LAI. The CLX clumping retrieval method (Leblanc et al., 2005) with a segment size of 15 was the best performing clumping method, matching the reference values to within 0.05 Omega on average near zenith. Clumping error increased linearly with zenith angle to > 0.3 Omega (equivalent to a 30% PAI error) at 75 for all structural configurations. At larger zenith angles, PAS errors were found to be around 25-30% on average when derived from the 55-60 degrees zenith angle. Therefore, careful consideration of zenith angle range utilised from HP is recommended. We suggest that plot or site clumping factors should be accompanied by the zenith angle used to derive them from gap size and gap size distribution methods. Furthermore, larger errors and biases were found for HPs captured within 1 m of unrepresentative large tree stems, so these situations should be avoided in practice if possible.

This paper presents a methodology for the attribution and characterisation of Sclerophyll forested landscapes over large areas. First we define a set of woody vegetation data primitives (e.g. canopy cover, leaf area index (LAI), bole density, canopy height), which are then scaled-up using multiple remote sensing data sources to characterise and extract landscape woody vegetation features. The advantage of this approach is that vegetation landscape features can be described from composites of these data primitives. The proposed data primitives act as building blocks for the re-creation of past woody characterisation schemes as well as allowing for re-compilation to support present and future policy and management and decision making needs. Three main research sites were attributed; representative of different sclerophyll woody vegetated systems (Box Iron-bark forest; Mountain Ash forest; Mixed Species foothills forest). High resolution hyperspectral and full waveform LiDAR data was acquired over the three research sites. At the same time, land management agencies (Victorian Department of Environment, Land Water and Planning) and researchers (RMIT, CRC for Spatial Information and CSIRO) conducted fieldwork to collect structural and functional measurements of vegetation, using traditional forest mensuration transects and plots, terrestrial lidar scanning and high temporal resolution in-situ autonomous laser (VegNet) scanners. Results are presented of: 1) inter-comparisons of LAI estimations made using ground based hemispherical photography, LAI 2200 PCA, CI-110 and terrestrial and airborne laser scanners; 2) canopy height and vertical canopy complexity derived from airborne LiDAR validated using ground observations; and, 3) time-series characterisation of land cover features. 1. Accuracy targets for remotely sensed LAI products to match within ground based estimates are ± 0.5 LAI or a 20% maximum (CEOS/GCOS) with new aspirational targets of 5%). In this research we conducted a total of 67 ground-based method-to-method pairwise comparisons across 11 plots in five sites, incorporating the previously mentioned LAI methods. Out of the 67 comparisons, 29 had an RMSE ≥ 0.5 LAIe. This has important implications for the validation of remotely sensed products since ground based techniques themselves exhibit LAI variations greater than internationally recommended guidelines for satellite product accuracies. 2. Two methods of canopy height derivation are proposed and tested over a large area (4 Million Ha). 99th percentile maximum height achieved a RMSE of 6.6%, whilst 95th percentile dominant height a RMSE = 10.3%. Vertical canopy complexity (i.e. the number of forest layers of strata) was calculated as the local maxima of vegetation density within the LiDAR canopy profile and determined using a cubic spline smoothing of Pgap. This was then validated against in-situ and LiDAR observations of canopy strata with an RMSE 0.39 canopy layers. 3. Preliminary results are presented of landcover characterisation using LandTrendr analysis of Landsat LEDAPS data. kNN is then used to link these features to a dense network of 800 field plots sites.

Estimating the proportion of woody-to-total plant material 'alpha' is an essential step to convert Plant Area Index 'PAP' estimates into Leaf Area Index 'LAI'. alpha has also been shown to have a significant impact on the passive optical remote sensing signal for retrieval of biophysical parameters in forests, woodlands, and savannas. However, benchmarked indirect a retrieval methods are lacking and thus it is common for this pivotal correction to be ignored. In this paper we validate an alpha retrieval method using a 3D radiative transfer simulation framework, enabling the retrieval method to be benchmarked against a known and precise model truth. The 3D framework consists of a representative and highly detailed 3D explicit Eucalypt forest reconstructed from field measurements. The 3D structure is coupled with a 3D scattering model to enable simulation of remote sensing instruments. The retrieval method utilises classified hemispherical photography 'HP', but is applicable to all ground-based optical instruments that can separate leaf and woody elements. The method is applicable to evergreen forests and thus independent of the estimation of PAI or LAI. The unknown degree of mutual shading or occlusion of leaf and woody elements was traditionally a key impediment to the, operational use of this method and was therefore closely examined. The indirect a method utilising classified HP imagery agreed on average to within 0.01 a of the reference (alpha(ref) = 0.37). In addition, the method demonstrated robustness to a range of LAI, stem density, and stem distribution values, matching to within +/- 0.05 alpha of the reference. Angular dependence on indirect alpha retrieval was also found; where the entire HP image (180 degrees FOV) was needed to produce the most accurate estimate. Conversely, the classified narrow view zenith angle range around 55-60 degrees zenith also provided an a estimate matching the reference. At this narrow zenith angle the method is insensitive to leaf angle distribution. As such, careful consideration of zenith angle range utilised from the instrument is recommended. The results demonstrate the method's applicability for accurate indirect estimation of alpha in single-storey forest types. The simple and efficient method can be used to convert estimates of PAI into LAI from a variety of optical ground-based instruments. Quantitative a estimates can and should be used to aid interpretation of the remote sensing signal from satellite imagery, which has been shown to be sensitive to the proportion and spatial distribution of woody canopy materials. (C) 2016 Elsevier B.V. All rights reserved.

Hyperspectral remote sensing serves as an effective tool for estimating foliar nitrogen using a variety of techniques. Vegetation indices (VIs) are a simple means of retrieving foliar nitrogen. Despite their popularity, few studies have been conducted to examine the utility of VIs for mapping canopy foliar nitrogen in a mixed forest context. In this study, we assessed the performance of 32 vegetation indices derived from HySpex airborne hyperspectral images for estimating canopy mass-based foliar nitrogen concentration (%N) in the Bavarian Forest National Park. The partial least squares regression (PLSR) was performed for comparison. These vegetation indices were classified into three categories that are mostly correlated to nitrogen, chlorophyll, and structural properties such as leaf area index (LAI). %N was destructively measured in 26 broadleaf, needle leaf, and mixed stand plots to represent the different species and canopy structure. The canopy foliar %N is defined as the plot-level mean foliar %N of all species weighted by species canopy foliar mass fraction. Our results showed that the variance of canopy foliar %N is mainly explained by functional type and species composition. The normalized difference nitrogen index (NDNI) produced the most accurate estimation of %N (R2CV = 0.79, RMSECV = 0.26). A comparable estimation of %N was obtained by the chlorophyll index Boochs2 (R2CV = 0.76, RMSECV = 0.27). In addition, the mean NIR reflectance (800–850 nm), representing canopy structural properties, also achieved a good accuracy in %N estimation (R2CV = 0.73, RMSECV = 0.30). The PLSR model provided a less accurate estimation of %N (R2CV = 0.69, RMSECV = 0.32). We argue that the good performance of all three categories of vegetation indices in %N estimation can be attributed to the synergy among plant traits (i.e., canopy structure, leaf chemical and optical properties) while these traits may converge across plant species for evolutionary reasons. Our findings demonstrated the feasibility of using hyperspectral vegetation indices to estimate %N in a mixed temperate forest which may relate to the effect of the physical basis of nitrogen absorption features on canopy reflectance, or the biological links between nitrogen, chlorophyll, and canopy structure.

Summary The vertical arrangement of forest canopies is a key descriptor of canopy structure, a driver of ecosystem function and indicative of forest successional stage. Yet techniques to attribute for canopy vertical structure across large and potentially heterogeneously forested areas remain elusive. This study introduces a new technique to estimate the Number of Strata ( NoS ) that comprise a canopy profile, using discrete‐return Airborne Laser Scanning (ALS) data. Vertically resolved gap probability ( P gap ) aggregated over a plot is generalized with a nonparametric cubic spline regression ( P s ). Subsequently a count of the positive zero‐crossings of second derivative of 1 − P s is used to estimate NoS . Comparison with inventory derived estimates at 24 plots across three diverse study areas shows a good agreement between the two techniques (RMSE = 0·41 strata). Furthermore, this is achieved without altering model parameters, indicating the transferability of the technique across diverse forest types. NoS values ranged from 0 to 4 at a further 239 plots, emphasizing the need for a method to quantify canopy vertical structure across forested landscapes. Comparison of NoS with other commonly derived ALS descriptors of canopy structure (canopy height, canopy cover and return height coefficient of determination) returned only a moderate correlation ( r 2 < 0·4). It is proposed the presented method provides a primary descriptor of canopy structure to complement canopy height and cover, as well as a candidate Ecological Biodiversity Variable for characterizing habitat structure.

Operational assessment of forest structure is an on-going challenge for land managers, particularly over large, remote or inaccessible areas. Here, we present an easily adopted method for generating a continuous map of canopy height at a 30 m resolution, demonstrated over 2.9 million hectares of highly heterogeneous forest (canopy height 0–70 m) in Victoria, Australia. A two-stage approach was utilized where Airborne Laser Scanning (ALS) derived canopy height, captured over ~18% of the study area, was used to train a regression tree ensemble method; random forest. Predictor variables, which have a global coverage and are freely available, included Landsat Thematic Mapper (Tasselled Cap transformed), Moderate Resolution Imaging Spectroradiometer Normalized Difference Vegetation Index time series, Shuttle Radar Topography Mission elevation data and other ancillary datasets. Reflectance variables were further processed to extract additional spatial and temporal contextual and textural variables. Modeled canopy height was validated following two approaches; (i) random sample cross validation; and (ii) with 108 inventory plots from outside the ALS capture extent. Both the cross validation and comparison with inventory data indicate canopy height can be estimated with a Root Mean Square Error (RMSE) of ≤ 31% (~5.6 m) at the 95th percentile confidence interval. Subtraction of the systematic component of model error, estimated from training data error residuals, rescaled canopy height values to more accurately represent the response variable distribution tails e.g., tall and short forest. Two further experiments were carried out to test the applicability and scalability of the presented method. Results suggest that (a) no improvement in canopy height estimation is achieved when models were constructed and validated for smaller geographic areas, suggesting there is no upper limit to model scalability; and (b) training data can be captured over a small percentage of the study area (~6%) if response and predictor variable variance is captured within the training cohort, however RMSE is higher than when compared to a stratified random sample.