IEEE Geoscience and Remote Sensing Magazine - March 2021 - 243

a linear correlation with water availability. Xue et al. [258]
further proved that tuning vegetation dynamic models
with lidar-derived attributes can lead to significantly different carbon modeling results. These studies suggest that
interpreting the spatial and temporal dynamics of forest
structure and incorporating them into the current dynamic vegetation models are necessary and important research
directions [256].
Moreover, the vertical dimension provided by lidar data
enables much more comprehensive and accurate descriptions of animal habitats [259]. Forest canopy structure is
a very important factor influencing animal behavior (e.g.,
nesting, hunting, and breeding) that is often neglected in
current niche modeling studies [26], [260]. Feeding niche
models with lidar-derived vegetation structural attributes
has the potential to overcome this obstacle. For example,
Zhao et al. [261] showed that lidar-derived structural attributes could better characterize fisher (a rare mammal)
dens; García-Feced et al. [262] and North et al. [263] found
that using lidar-derived snag trees and the cover of tall trees
could predict California spotted owl nesting habitats; Loarie et al. [264] used lidar-derived vegetation structure to
explain the hunting behavior of male lions; and Blakey et
al. [265] and Eichhorn et al. [266] used lidar-derived structure information to understand how deer, birds, and bats
choose their habitats. Coops et al. [267] proposed a 3D index using lidar data capturing the three main components
of vertical and horizontal vegetation structure (i.e., height,
cover, and complexity) and proved that this index can be
used to predict avian species richness more accurately than
using climate or land cover data alone. However, animals of
different taxonomic groups interact with forest canopy differently. Thus, investigating the influence of 3D forest canopy structure on a wider range of animals can benefit our
understanding of the correlation between canopy structure
and biodiversity.
In agricultural ecosystems, lidar has improved the accuracy of crop growth modeling. Crop growth models
aim to simulate the crop growth cycle with the input of
environmental variables (e.g., temperature, precipitation,
soil moisture, and fertilizer) [268], [269]. Lidar can provide detailed and accurate structural traits at an individual
organ level, which could be a robust information source
for validating and calibrating crop growth models. Additionally, the 3D light regime of agricultural ecosystems
can be more accurately modeled with the aid of 3D radiative transfer models [270], which can refine the environmental input of crop growth models. The combination of
improved crop growth models and accurate lidar observations could ultimately lead to more efficient crop management. However, the incorporation of lidar data into crop
growth models is still in its very early stages. A recent study
by Liu et al. [271] used a function-structure-plant model
to provide in silico experiments for exploring the feasibility of lidar-based phenotype estimation (e.g., a green area
index) with machine learning methods. More studies are
MARCH 2021

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

still urgently needed to investigate the potential of lidar in
the crop growth modeling process.
In grassland ecosystems, airborne and UAV lidar data
have been used to model spatial heterogeneity and biodiversity [211], [272]. For example, Jansen et al. [273] used
lidar-derived structural attributes to model the influence
of grazing on the spatial heterogeneity of a bunchgrass
prairie ecosystem; Moeslund
et al. [274] successfully used
airborne lidar data to asTHE COMBINATION OF
sess the local biodiversity of
IMPROVED CROP GROWTH
grasslands and meadows in
MODELS AND ACCURATE
bushes; and MarcinkowskaOchtyra et al. [275] and FishLIDAR OBSERVATIONS
er et al. [276] modeled the
COULD ULTIMATELY LEAD
distribution of grassland inTO MORE EFFICIENT CROP
vasive species. These results,
MANAGEMENT.
c ombi ne d w it h biom a s s
modeling results from lidar
[199], can ultimately provide
much more accurate information for managing and protecting grassland ecosystems. However, currently, studies
in this direction are still few in number. More studies are
urgently needed to link lidar-derived structural attributes
with functional attributes to better simulate grassland ecosystem processes.
In urban ecosystems, detailed 3D information of built
and vegetation structures can be assimilated into models to understand urban microclimates, investigate urban
biodiversity, and ensure the health of urban ecosystems.
The 3D arrangements of buildings and vegetation can
alter the spatial and temporal variability of solar radiation [277]-[281], and assimilating the 3D information of
buildings and vegetation into climate models can improve
the accuracy of microclimate simulations [282]-[285].
Moreover, urban ecosystems have become an important
habitat for both flora and fauna, and anthropogenic activities are important factors influencing the biodiversity
of urban ecosystems. The integration of lidar-derived
information into biodiversity studies can help better describe the spatial distribution of vegetation and animal
species in an urban environment [286], [287]. Modeled
vegetation and animal distribution can be further used
to analyze and predict human health risks (e.g., exposure
to pollen and mosquito abundance) [288]-[290]. In addition, lidar data can be used to model risks related to
power lines, flooding, and so on, and thus help in devising management plans that ensure the health of urban
ecosystems [196], [291]-[295].
To conclude, lidar-derived 3D information is an important input for ecological models. The integration of lidar
observations into ecological models can improve our understanding and may even challenge our existing theories on
ecological processes. Despite the consensus on the importance of integrating lidar observations into ecological models, this process is still in the early stages across ecosystems.
243

IEEE Geoscience and Remote Sensing Magazine - March 2021

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2021