Errors in the Estimation of Leaf Area Density From Aerial LiDAR Data: Influence of Statistical Sampling and Heterogeneity

Halubok, Maryia; Kochanski, Adam K.; Stoll, Rob; Bailey, Brian N.

doi:10.1109/tgrs.2021.3123585

Cited by 5 publications

(3 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, data of 3D LAD distribution have not been readily available. However, airborne high‐density point clouds are now becoming publicly available (e.g., Shizuoka point cloud DB, 2022; USGS 3D elevation program, 2022), and methods for estimating the 3D distribution of LAD or plant area density using airborne point clouds have been developed (Arnqvist et al., 2020; de Almeida et al., 2019; Halubok et al., 2022; Oshio et al., 2015). In addition, small and inexpensive laser scanners have been developed, and point clouds are readily available by measurements from the ground or an unmanned aerial vehicle (Hu et al., 2021; Lu et al., 2020).…”

Section: Discussionmentioning

confidence: 99%

Simulating the 3D Distribution of Absorbed Shortwave Radiation in a Tree Crown: Comparison of Simplified and Monte Carlo Models

Oshio,

Asawa

2023

JGR Atmospheres

View full text Add to dashboard Cite

Urban trees are expected to alleviate severe thermal environments in summer. The absorbed shortwave radiation is a key variable to estimate the microclimatic effects of trees. In numerical simulation tools for the urban and built thermal environment, the difficulty lies in how to calculate the absorbed radiation in terms of accuracy and computational load. Here we investigated the difference between a radiative transfer model calculating scattering by the Monte Carlo method and a simpler model that accounts for scattering by modifying the optical depth. The calculations of absorbed radiation were performed under various conditions of leaf normal distribution and zenith angle of incident radiation, using tree models represented by voxels with different leaf area density values. For the photosynthetically active radiation, the relative difference between models was less than 10% for most cases. For the near‐infrared radiation, the difference was also small for a large canopy consisting of several trees. The difference was 30%–60% and 0%–20% for zenith angles of incident radiation of 0° and 60°, respectively, for small isolated trees with a crown height and width of less than 5 m, which was a significantly different condition from a uniform canopy. This large difference was related to the uniformity of the canopy and was explained by the transmittance from each voxel to the outside of the canopy. The present study enables us to evaluate the applicability of the simple model and correct the difference from the Monte Carlo model.

show abstract

Section: Discussionmentioning

confidence: 99%

Simulating the 3D Distribution of Absorbed Shortwave Radiation in a Tree Crown: Comparison of Simplified and Monte Carlo Models

Oshio,

Asawa

2023

JGR Atmospheres

View full text Add to dashboard Cite

show abstract

“…Different species of vegetation cohabit with distinct LAD distribution and height. The LAD was estimated based on high-resolution aerial LiDAR data collected during the RxCADRE campaign using the approached described in Halubok et al (2021). LAD was computed on a voxel grid of 5 × 5×3 m 3 or 25 × 25×3 m 3 (both cases are used).…”

Section: Environmental Conditionsmentioning

confidence: 99%

“…Observed wind speeds were much weaker than the simulation results in Region 3. This discrepancy could be caused by underestimation of the LAD by LiDAR within the forest, where the vegetation heterogeneity made it difficult to measure accurate values even at 5-m resolution (Halubok et al, 2021, details how the heterogeneity and clumping cause a negative bias in LAD). Finally, there were no trees in Region 4 and the southwest road was aligned with the wind direction.…”

Section: Rxcadre -A Heterogeneous Forestmentioning

confidence: 99%

A rapid method for computing 3-D high-resolution vegetative canopy winds in weakly complex terrain

Renault,

Bailey,

Stoll

et al. 2024

Front. Earth Sci.

Self Cite

View full text Add to dashboard Cite

To determine near-surface winds within and above vegetation canopies for operational environmental applications, a wind model must run at high-resolution (O(1–10 m)), in a few minutes, using limited input information, and requiring minimal computing resources (e.g., personal computers). Current research models simulate large domains at coarse resolution or small domains at fine scale, but canopy simulations can take days. Fast-modeling approaches are used to solve large complex wind fields, but they oversimplify the roughness elements’ distribution impact on momentum exchanges. To overcome these deficits, the fast-running wind model QUIC-URB (Quick Urban and Industrial Complex) was augmented with a high-resolution canopy wind solver. The wind model includes a non-local factor that describes how momentum propagates through the canopy and how sub-canopy jets appear under certain conditions. QUIC-URB was also coupled with the mesoscale WRF (Weather Research and Forecasting) model to downscale wind fields from a few kilometers to a meter. The new QUIC Canopy Model resolves 3-D wind fields over hundreds of millions of cells in less than 30 s per time step on a personal computer. It was compared to two canopy models for real quasi-homogeneous and heterogeneous canopies. An error analysis shows that the model was relatively accurate with a normalized root-mean-square error of about 0.2 m s−1 in the quasi-homogeneous canopy, and a mean absolute error of 0.3 m s−1. The new model is suitable for coupling with pollution dispersion, wildfire spread, and numerical weather prediction models over weakly complex terrain, defined here as a mildly undulating environment with gradual changes in elevation and a heterogeneous distribution of plants.

show abstract

Does intensity-based weighting of multiple-return terrestrial LiDAR data improve leaf area density estimates?

Kent,

Bailey

2024

Remote Sensing of Environment

View full text Add to dashboard Cite

Errors in the Estimation of Leaf Area Density From Aerial LiDAR Data: Influence of Statistical Sampling and Heterogeneity

Cited by 5 publications

References 75 publications

Simulating the 3D Distribution of Absorbed Shortwave Radiation in a Tree Crown: Comparison of Simplified and Monte Carlo Models

Simulating the 3D Distribution of Absorbed Shortwave Radiation in a Tree Crown: Comparison of Simplified and Monte Carlo Models

A rapid method for computing 3-D high-resolution vegetative canopy winds in weakly complex terrain

Does intensity-based weighting of multiple-return terrestrial LiDAR data improve leaf area density estimates?

Contact Info

Product

Resources

About