Green Area Index (GAI) serves as an important link in the analysis of the relationship between solar energy radiation and crop grain yield. Voxel-based approach from terrestrial laser scanning (TLS) data has been widely used to measure the canopy structure due to its ability to characterize plant material within a 3D grid. However, few studies have explored how TLS sampling design and voxel size affect the accuracy of crop GAI estimation. In this study, we assessed these effects on wheat GAI estimation based on measured and simulated TLS data. We simulated different TLS sampling designs, including varying numbers of scanning sites (1, 4, 8, 12, 16) and scanning strategies (Strategy 1: peripheral distribution; Strategy 2: uniform distribution). Furthermore, a LiDAR algorithm for GAI estimation based on voxel size optimization, which we call the k-neighborhood voxel approach (KNV), was developed. The results demonstrated that TLS sampling design 2-12 (12-scan sites uniform distribution) was the most effect approach for comprehensively characterizing canopy structure across the wheat growing seasons. Uniform distribution (Strategy 2) resulted in a greater number of laser returns from lower part of the canopy, and enhanced the homogeneity of the distribution of laser points. In addition, the optimal voxel method could greatly mitigate the impacts of changes in TLS sampling design and improve the accuracy of TLS technology for estimating wheat GAI (RMSE = 0.57, RRMSE = 18.26%). This study provides valuable guidance for effective acquisition and accurate estimation of crop attributes in crop breeding and plant phenotyping.

In the context of global climate change, understanding forest dynamics and formulating sustainable forest development strategies require accurate quantification of forest structural changes over time. However, the majority of studies have focused on exploring forest structure at one specific time point, lacking repeated observations of forest structure. This limitation hinders the comprehensive understanding of structural dynamics in forests, particularly its link with changes in forest ecosystem functionality. To address this issue, this study proposes a method for quantifying forest structural dynamics using bi-temporal terrestrial laser scanning (TLS) data. We utilized bi-temporal TLS data in combination with a quantitative structure model (QSM) algorithm, to reconstruct 3D models of individual trees. Subsequently, tree parameters including diameter at breast height (DBH), tree height, crown projection area (CPA), crown volume (CV) and biomass were extracted. By quantifying the changes in these parameters and analyzing their relationships with the 3D spatial structure of trees, a comprehensive analysis was conducted to quantitatively assess forest dynamics. The results indicate that bi-temporal TLS data possesses unprecedented advantages and tremendous potential in quantifying forest dynamics.

Terrestrial laser scanning (TLS) has proven to be an invaluable tool in various forest ecology applications and forestry research. A crucial step in most TLS forest point cloud processing pipelines is instance segmentation; separating individual trees from the forest. However, automation in this area proves difficult, largely due to the heterogeneity of tree features and composition as well as overlapping, dense crown areas and understory. A lack of benchmarks and standard metrics complicates intercomparison of methods and hinders development in the field. This work proposes a set of metrics and methodology for benchmarking methods, and applies this to four open source TLS instance segmentation methods on a fully segmented 1.2 hectare benchmark dataset of a deciduous forest.

Using satellite remote sensing to map degradation in dense tropical forests is challenging. Accurate ground data are desperately needed to calibrate and validate detection algorithms. To improve our measurements of small-scale disturbance events, such as those caused by selective logging, we established a degradation experiment in eight 1-ha plots located in the tropical forests of Peru and Gabon. Biomass data was collected before and after the extraction of a small number of trees (resulting in the loss of between 4 and 31% initial biomass), and scenes of high-resolution optical satellite imagery from five satellites from 0.3 - 30 m pixel resolution compared before and after. Preliminary results suggest visual analysis of data up to and including 10m resolution (from Worldview-3 to Sentinel-2) can detect the disturbance caused by logging, but it is invisible to 30 m resolution Landsat. However, further work is needed to automate the detection process using high resolution data.

The TomoSense experiment comprises campaign and research activities in support of future Synthetic Aperture Radar (SAR) mission concepts at P-, L-, and C-band by the European Space Agency (ESA). The research is intended to provide a quantitative basis for the evaluation of single-pass interferometry over temperate forests at L- and C-band and investigate potential synergies between C-band convoy mission concepts and future P- and L-band missions. SAR acquisitions include P- L-, and C-band data acquired at the Eifel National Park in Germany by flying approximately 25 trajectories to provide tomographic imaging capabilities. Land C-band data were acquired by simultaneously flying two aircraft to gather bistatic data with varying interferometric baselines. Field activities include forest census (dbh, tree height and species) at 80 plots and Terrestrial Laser Scanning (TLS). The dataset is complemented by small-footprint Airborne Lidar Scanning (ALS) and derived products. Preliminary results are here shown relative to polarimetric tomography at P-band and L-band imaging.

Obtaining accurate and detailed structural forest information has been revolutionized with the emergence of laser scanning. The sampling limitations and potential of the different laser scanning platforms (e.g. TLS, UAV -LS) have, however, not been fully explored for dense tropical forests. We fused laser scanning data from the terrestrial (TLS) and drone (UA V -LS) platform for two dense tropical forest plots and calculated their vertical point density profiles to gain insight in their sampling abilities. Our results reveal the limitations of TLS to fully sample the top of the canopy of a dense tropical rainforest. We also demonstrate how multiple returns but also cheaper single returns UAV -LS systems can be applied to sample the forest structure.

Terrestrial laser scanning (TLS) data provide 3-D measurements of vegetation structure and have the potential to support the calibration and validation of satellite and airborne sensors. The increasing range of different commercial and scientific TLS instruments holds challenges for data and instrument interoperability. Using data from various TLS sources will be critical to upscale study areas or compare data. In this paper, we provide a general framework to compare the interoperability of TLS instruments. We compare three TLS instruments that are the same make and model, the RIEGL VZ-400. We compare the range accuracy and evaluate the manufacturer's radiometric calibration for the uncalibrated return intensities. Our results show that the range accuracy between instruments is comparable and within the manufacturer's specifications. This means that the spatial XYZ data of different instruments can be combined into a single data set. Our findings demonstrate that radiometric calibration is instrument specific and needs to be carried out for each instrument individually before including reflectance information in TLS analysis. We show that the residuals between the calibrated reflectance panels and the apparent reflectance measured by the instrument are greatest for highest reflectance panels (residuals ranging from 0.058 to 0.312).

Combining virtual forests with radiative transfer is a powerful tool to calibrate and validate ground-based, airborne and spaceborne sensors. In this set-up, we can control and calculate all aspects of the forest structure and the simulated signal, which would not be possible using measured data only. Terrestrial laser scanning (TLS) enables us to measure forest structure directly with high detail. We present a processing chain that uses TLS data as input data to assess the end-to-end traceability of various in situ LAI and fAPAR products via radiative transfer modelling. Tree reconstruction from TLS data is used to represent the explicit 3D forest structure in radiative transfer models.

Raumonen et al.[1] have developed a new method for reconstructing topologically consistent tree architecture from TLS point clouds. This method generates a cylinder model of tree structure using a stepwise approach. Disney et al.[2] validated this method with a detailed 3D tree model where structure is known a priori, establishing a reconstruction relative error of less than 2%. Here we apply the same method to data acquired from Eucalyptus racemosa woodland, Banksia ameula low open woodland and Eucalyptus spp. open forest using a RIEGL VZ-400 instrument. Individual 3D tree models reconstructed from TLS point clouds are used to drive Monte Carlo ray tracing simulations of TLS with the same characteristics as those collected in the field. 3D reconstruction was carried out on the simulated point clouds so that errors and uncertainty arising from instrument sampling and reconstruction could be assessed directly. We find that total volume could be recreated to within a 10.8% underestimate. The greatest constraint to this approach is the accuracy to which individual scans can be globally registered. Inducing a 1cm registration error lead to a 8.8% total volumetric overestimation across the data set.

Calibration and validation of LAI products require accurate ground-based measurements. Many indirect ground-based sensors such as digital hemispherical photography (DHP), ceptometers, and terrestrial laser scanners (TLS) are used interchangeably to estimate reference values. However these sensors have biases in regards to the true LAI value, which can never be known in the field. Results from three representative woody ecosystems in Eastern Australia are presented from real field measurements. Significant differences were found between methods at the individual measurement and plot scale. Furthermore, one of the sites in South East Australia was measured and modeled in a 3D deterministic model. In this digital environment where the truth is known, sensors can be simulated to determine their bias.

Empirical relations are frequently used to derive leaf area index (LAI). Such relations often make assumptions that make it hard to link the derived LAI to realistic trees and forest canopies. In previous work we developed a set of analytical expressions to describe LiDAR waveforms with only a limited number of assumptions based on radiative transfer. These expressions were a function of crown macro-structure and LAI. The expressions were successfully tested when applied on crown archetypes, but showed significant error when applied to more realistic crowns. In this study, we analyse the effect of clumping on inferring LAI from realistic trees. Despite the potential of the expressions to detect subtle changes in LAI, absolute inferred LAI values can be significantly off. However, the strong correlation between true and inferred LAI (R2>; 0.97) for the two test cases in this study, allows for calibration of the inferred LAI values.

This paper presents a method to extract information on the impact of wildfire on vegetation canopies. A simple linear model is proposed with a ‘generic’ spectral model of the impacts of wildfire on vegetation canopies. This allows a term related to the projected proportion of a pixel affected by fire (fcc) to be estimated from measurements of pre- and post-fire spectral reflectance. The properties of fcc are investigated using a hybrid radiative transfer model to simulate the impacts of wildfire in a multi-layered canopy. Spectral sampling from MODIS is assumed (7 bands). The fcc is confirmed to relate to the fractional area of a pixel affected by fire, although in multi-layer canopies it is modulated by a term dependent on the contrast between the pre-fire reflectance of areas affected by fire and those unaffected.

This study explores the capability of satellite remote sensing to detect relatively rapid changes of vegetation cover in northern Fennoscandian regions in response to disturbance more generally, and insect defoliation damage in particular. The data used is a long term time series of leaf area index (LAI) at 8Km resolution derived from the Advanced Very High Resolution Radiometer (AVHRR) between 1982 and 2006, developed to be structurally consistent with the Moderate Resolution Imaging Spectrometer (MODIS) record. The study explores the potential of frequentist traditional statistics to detect disturbances at this coarse spatial resolution over the 25 years time series, and outlines the possibilities that Bayesian methods offer to improve the detection by including prior information on the profile of such disturbance events.

Spaceborne differential absorption lidar has been proposed for accurate measurements of atmospheric CO2(and surface properties). Lidar instruments typically observe the highest possible surface reflectance due to observing in the retroreflection direction (the so-called ldquohotspotrdquo) where viewed shadow is minimized. The range of observed reflectance will determine instrument dimensions and signal-to-noise ratio, but it is difficult to predict this range globallyapriori. Two complementary methods are presented for estimating lidar reflectivity over a range of vegetated surface types. The first method simulates the expected response of a lidar instrument from multiangle multispectral reflectance data. The second method uses detailed 3-D vegetation structural models and Monte Carlo ray tracing to simulate the lidar signal. The simulations are used to validate the first method and assess the impact of possible instrument configurations. Both methods agree well and are robust to error in observations, with predicted lidar reflectivity (at 1570 and 2050 nm here) typically between 10% and 33% higher relative to off-nadir reflectance and ranging from 0.02 to ~ 0.7. We use the 3-D simulations to show that the impact of shifted on-off lidar pulses is not likely to be significant for accuracy of retrieved CO2, and we demonstrate that the 3-D simulation method is a flexible and powerful way of prototyping future spaceborne lidar missions.

Spaceborne lidar missions are being studied to estimate atmospheric concentrations of CO2, water vapour and O3, as well as for measuring surface biophysical properties. Lidar instruments typically observe the highest possible surface reflectance due to observing in the retroreflection peak (the so-called 'hotspot'), where shadowing on the surface is minimised. The likely range of observed reflectance will determine the required dynamic range and desired signal-to-noise ratio (SNR) of such an instrument, but it is difficult to predict this range a priori. A method is presented for estimating lidar surface reflectance over a range of vegetated surface types using multi-angle, multi-spectral reflectance data. The approach is validated using radiative transfer simulations of highly detailed 3D vegetation canopy models. The method is particularly useful for testing proposed lidar instrument configurations.

It is generally agreed that the optimal footprint size for a spaceborne lidar is 30 m. Over topography such a large footprint can blur the canopy and ground signal together preventing information extraction. Multi-spectral lidar waveforms have been simulated with Monte-Carlo ray tracing over explicit geometric forest models. A method for using multi-spectral waveform lidar to distinguish ground from canopy returns has been tested over a range of ground slopes. The results are promising, with an initial error of +/-5 m for a signal level of only 5,000 photons with noise; an easily achievable figure. The inversion algorithms completely dominate inversion errors for all cases above 10,000 signal photons.

This paper explores the potential to improve spatial estimates of key carbon fluxes by combining Earth Observation data with a simple ecosystem model. Spatial estimates of Leaf Area Index from MODIS at the kilometre scale over a coniferous forest site in Oregon are assimilated into an ecosystem model with an Ensemble Kalman filter. Results show that assimilating EO data improves the magnitude of estimates of Net Ecosystem Productivity relative to running the model alone, however the uncertainty is not significantly constrained. Spatially there is an underestimate in modelled carbon fluxes. This is attributed to error in the EO data which induces an underestimate in model stock estimates, as well as inadequacies in the model parameterisation.

A time-series of MODIS Terra daily 500m resolution surface reflectance data is investigated to examine the effect of current land management activity taking place on an upland peat site in North Wales. A catchment-scale controlled experiment has been set up to monitor the effect blocking of the artificial drainage channels (`grips') has on the ecosystem dynamics. The reflectance data and the Normalised Difference Vegetation Index (NDVI) and Normalised Difference Water Index (NDWI) were compared both before and after the period of grip blocking and between the control and treated catchmensts. No change was observed as a result of the grip blocking. Reasons why this might be the case are discussed.

Assimilation of Earth Observation (EO) data into Dynamic Vegetation Models (DVMs) can either be via derived products (e.g., LAI or fAPAR) or through radiances. Successful assimilation generally requires that the distribution of errors in an observed variable is well known. Radiance measurements conform to this requirement more strongly than derived products which have undergone more complex processing and whose relation to the true value of the estimated variable is poorly understood. To enable radiances to be assimilated a DVM must be able to predict canopy leaving radiation. To do this it is necessary to couple it with a Canopy Reflectance Model (CRM). As DVMs require some concept of intercepted radiation to drive photosynthesis and ultimately plant growth, there is a common framework between DVMs and CRMs which may be exploited for this purpose. This work describes the mechanisms by which a simple radiative transfer CRM can be coupled with a DVM and discusses the disparities in the assumptions made by each concerning photon-vegetation interactions. Results of forward modelled canopy reflectances are presented and compared with EO estimates of reflectance.

A dynamic 3D structural model is used to simulate the structural growth stages of a Scots pine canopy from age 5 to 50 years. The 3D structural output of the model agrees with observed measures of Scots pine canopy structure. Needles are added to the structural model according to measured density and phyllotaxy (distribution). The 3D structural models are used to drive both optical and microwave models of canopy radiometric response. Simulated canopy radiometric response is compared with airborne hyperspectral reflectance data (HyMAP) and airborne synthetic aperture radar (ASAR) backscatter data, recorded during the SAR and Hyperspectral Airborne Campaign (SHAC) conducted over the UK during 2000. Simulations are shown to agree well in general with observations. This method is shown to be suitable for exploring the impact of canopy structure on the measured remotely sensed signal.

A dynamic 3D structural model of wheat is used to simulate growth stages of a wheat canopy. The 3D information regarding canopy structure output by this model is used to drive two models of canopy radiometric response, one in the optical and one in the microwave domain. The radiometric response of the canopy in the optical and microwave domains is simulated and the sensitivity of the canopy response to variation in the canopy structural and radiometric parameters is examined. The modelled canopy response is compared to field-measured hyperspectral reflectance in the optical, and airborne synthetic aperture radar (E-SAR) backscatter measurements in the microwave. The modelled signal is shown to agree in general with observations. It is demonstrated that the major growth development stages of the wheat canopy which impact the remote sensing signal in the optical and microwave domains can be modelled in this way. This method is shown to be suitable for exploring the impact of canopy structure on remotely sensed measurements.

Large swath width sensors with short revisit periods (circa. 1 day) between successive data acquisitions of a given point on the Earth's surface provide an excellent opportunity to study the phonological developments of vegetation cover. Although very detailed phenological measurements are probably beyond the scope of such sensors, broad changes in vegetation should be readily detected and this information may be used to constrain and check the performance of dynamic vegetation models. This paper compares time series data from a variety of Earth Observation platforms (principally AVHRR and SPOT-4 VGT) and examines their ability to describe the phenology of the United Kingdom over a period several years. Phenological parameters are derived from vegetation indices and the ability of these parameters to drive or be assimilated into vegetation models is discussed with reference to the Sheffield Dynamic Global Vegetation Model.

3D structural models of a (dynamic) wheat canopy and a Scots pine canopy are used to drive models of canopy scattering behaviour in the optical and microwave domains. These models have been shown to be capable of simulating canopy reflectance and backscatter using scattering models in the optical and microwave domains. By varying the structural and radiometric parameters governing the simulated canopy response, look-up tables (LUTs) of reflectance (optical) and backscatter (microwave) are constructed. These LUTs are used to 'invert' the scattering models against reflectance and backscatter observations made from airborne reflectance and SAR backscatter data. The practicability of this approach is demonstrated. It is shown that combined optical and microwave retrievals are possible without the need for simplified mathematical models of scattering behaviour and time-consuming numerical inversion techniques. Once LUTs have been generated, inversion against observations is very rapid and flexible.