The use of drones in forestry [630328]

Review article

The use of drones in forestry
T.P. Banu1, G.F. Borlea1, C. Banu1

1Banat’s University of Agricultural Sciences and Veterinary Medicine “King Michael I of Romania” from Timișoara

Abstract
Recently drones have found applicability in a variety of study fields, one of these being forestry, where
an increasing interest is given to this segment of technology, especially due to the high -resolution data that can
be collected flexibly in a short time and at a relatively low price. Also, drones have an important role in filling the
gaps of common data collected using manned aircraft or satellite remote sensing, while having many advantages
both in research and in various practical applications particular ly in forestry as well as in land use in ge neral. This
paper aims to briefly describe the different approaches of applications of unmanned aircraft vehicles (UAVs) in
forestry, such as forest mapping, forest management planning, canopy height model creation or mapping forest
gaps. These approac hes have great potential in the near future applications and their quick implementation in a
variety of situations is desirable for the sustainable management of forests.

Keywords: drones, UAV, remote sensing, forest management

1. Introduction

The first aerial photos were aquired in 1860 using
balloons, which were later replaced by aicrafts in the
first and second world war. Afterwards, t he first
satellites were launched into orbit around the 1960s
for military purposes, and since the 1970s, once with
the development of digital sensors, first civilian
applications appeared.
In recent decades, remote sensing techniques
applied in forestry are given an increase d attention ,
which leads to the ability of extracting important
informatio n for forest planning and sustainable
management such as th e forest structure,
composition , volume or growth. (1Shao , 2012 ).
At the same time with the development of
sensors, computers and computational techniques,
the applicability of remote s ensing in forestry has
evolved from aerial photography data (2Lyons , 1966 )
to satellite imagery data (3Asner et al , 2005 ), which led
to different calcuations of forest indexes (eg. 4Thakur et al , 2014 ) and up to estimations of forest volume and
biomass (5Lu, 2006 ; 6Koch , 2010 ).
Satellite programs like Landsat, which is still one
of the most used remote sensing worldwide program,
commonly used in forestry , (7Alberts , 2012 ) and the
recent Sentinel 2 , which has a higher revisit frequency,
narrower bandwidths and finer resolution (8Hojas –
Gascón et al, 2015 ) are still limited and not suitable in
applications where very high spatial resolution is
needed, such as individual trees or even leafs, or those
requiring a very short period of area revisiting.
In this context, the applicability and effectiveness
of drones has a great potential in fillin g the gaps of
other data types even if currently forest applications
are still in experimental stage (9Shahbazi et al, 2014 ).
Drones are unmanned aerial vehicles (U.A.V.s)
which were originally developed in the early last
century and after the 1950 s, the main application of
drones relied in the military field, through
reconnaissance or surveillance . In the last decade,

drones of different sizes, shapes and capabilities have
grown rapidly and have acquired a growi ng interest
(Fig.1) in civilian applications (10Colomina and Molina ,
2014 ), as well as precision agriculture, forestry,
biodiversity, meteorology, emergency management,
wildlife research, land management, traffic monitoring
and many others (9Shahbazi et al , 2014 ).

The most common classification of drones is made
according to takeoff and landing type (11Tang and
Shao , 2015 ):
• takeoff and landing horizontally , typical for
fixed -wing drones (aircrafts) – Fig.2;
• takeoff and landing vertically , typical for
rotary -wing drones ( helicopter s, quadcopters,
hexacopters etc.) – Fig. 3 ;

The stability of drones and the area covered per
flight are key elements in applications of remote
sensing. The first category of drone has the advantage
of higher coverage areas per flight and the second
category has the advantage of better stability whic h
gives a higher spatial resolution, but with reduced areas that can be covered per flight (11Tang and Shao ,
2015 ).
Another classification of drones is based on their
power supply and it directly affects the maximum
flight time (12Dudek et al, 2013 ). On this matter, UAVs
can be split into two different types: electric (Fig. 2,3)
and internal combustion (Fig. 4). Drones with electrical
power supply are recommended for remote sensing
applications compared to those with internal
combustion that are not so econom ical and have
higher vibrations (11Tang and Shao, 2015 ).

The drones can fly autonomously or by remote
piloting using a remote control. Autonomous flight is
scheduled previously and is well suited to systematic
mapping.
A multitude of details for drone remote sensing
can be found in the literature prepared by 13Anderson
and Gaston (2013) ; 10Colomina and Molina (2014) .

2. Applications of UAVs in forestry
Remote sensing using drones has a range of
benefits such as reduced costs, flexibility in time and
space, high accuracy data and the advantage of no
human risks.
It is important to mention that even though forest
fires ´ moni toring and management was one of the first
field in forestry that showed the importance of drones
in forestry (14Ambrosia et al, 2003 ; 15Ollero, 2006 ),
getting NASA and th e US Forest Service to present a
drone that was able to fly up to 24 hours (11Tang and
Shao, 2015 ), at the moment the use of drones spread
in other more popular forestry fields.
The following will briefly present recent examples
of applications of drones in forestry.

2.1 Mapping forests and biodiversity

An application in order to map forest areas was
made by 16Koh and Wich (2012) , where a drone was
used in mapping tropical forests in Indonesia. The
experiment involved a small UAV type aircraft (under
1 kg) with a flight time range of approximately 25 min
per flight a nd a maximum distance traveled per flight
of about 15 km. The acquired images were assembled
after the flights and a land cover map resulted with a
spatial resolution of 5.1 cm. Also, shots and videos that
caught different human activities, logging, wildlife (Fig.
5) or flora species were made. Authors suggested that
using UAV remote sensing can save time, costs and
labor power for these purposes.

2.2 Precision forestry and sustainable forest
planning management
Parameters such as canopy cover, number of
trees, volume estimation, vitality or composition of
stands are important parameters in forest planning
and sustainable forest management. Quick and
accurate determination of canopy cover can b e
reached using unmanned aircraft (eg. 17Chianucci et al,
2016), leading to fast an d better decisions that
improve optimal quality and productivity of stands
(Fig. 6). A study that estimated biomass volume and
basal area in forests of Pinus Taeda was done by
18Marks et al in 2014 using a drone with radar SAR (L –
band – Synthetic Aperture Radar) sensor, the authors
concluding that it is feasible to use L -band or shorter
wavelength radar to measure these parameters only
in young stands with lower basal area (BA) and lower
above ground biomass (AGB) similar to stands from
their study.

19Mokroš et al (2016) used a commercial low -cost
quadcopter (DJI Phantom 3 Professional) to fly at an
altitude of ~ 20 m eters in order to estimate the volume
of wood chips pile. After comparing the results with
the volume resulted with a GNSS device, the authors
concluded that the use of UAVs does not lead to
significantly different results (10.4% more volume
estimated via drone method) and the time to collect
data is significantly lower (12 -20 times less) with the
advantage of documentation through orthomosaic
(19Mokroš et al 2016 ).
20Hassaan et al (2016) used a commercial
quadcopter (DJI Phantom 2) with a maximum flight
time of ~20 min to count trees in urban areas and
successfully identified trees with a 72% accuracy. Also,
a paper that consisted in detecting the number of
trees using LiDAR sensor s mount ed on a UAV was
successfully performed by 21Wallace et al in 2014 .
Fast-growing forest plantations can have similar
approaches to precision agriculture from the drone
remote sensing perspective (22Wang et al , 2014 ) and
increasing their productivity is a major concern,
especially nowadays when the demand of timber is
highly increasing. Regarding this field, 23Felderhof and
Gillieson (2011) acquired near -infrared (NIR) images
using drone r emote sensing to map the vitality of tree
canopy in a macadamia plantation, where they found
significant correlations between spectral radiation of
trees and the levels of nitrogen in leaves measured in –
situ.
Regarding the monitoring of forest vegetation
reco very in tropical areas, the work of 24Zahawi et al
(2015) indicated that the methodology based on the
use of drones is viable even for large data volumes.
25Lehmann et al (2015) study showed the
potential of UAVS for the detection of pest infestation
(Agrilus biguttatus ) levels in small oak forest stands.
Authors used a commercial quadcopter with an
estimated flight time of ~30 min and ~200g payload
used for a low -priced digital CIR-camera (Canon IXUS
100 – equipped with an infrared sensor) that acquired
images with a very high spatial resolution (~2cm).
Their results were based on five classes had an overall
accuracy over 82.5% with an estimated saving time
and financial cost by more than 50% for small/medium
sized stands compared to traditionally ground based
pest detection workflow (25Lehmann et al 2015 ).
In future studies, fields such as forest dynamics,
forest species proportions in stands, mapping and
assessing forest disturbances will grow considerably
with the benefits of unmanned aircraft technology.

Fig. 5 Wildlife
photographs captured
by 16Koh and Wich , 2012
a. Sumatran orangutan
b. Sumatran elephant
Fig. 6 Cropped RGB
orthomosaic used for canopy
cover determination
(17Chianucci et al , 2016 )

2.3 Mapping canopy gaps
Forest disturbances, especially those caused
by wind and snow, directly affect regeneration,
biodiversity and productivity of the stands and
mapping the forest canopy gaps can provide an
accurate situation of these type s of disturbances. So
far, small gaps cannot be measured accurately using
satellite data (26Frolking et al, 2009 ), but it can now be
achieved using drone remote sensing.
In Germany, 27Getzin et al (2012) made drone
flights in beech -dominated deciduous and deciduous –
conifers mixed forests, obtaining images (Fig. 7) with 7
cm resolution that could accurately identify gaps in the
canopy down to 1 sq m size. The flights took place at an
altitude of 250 m with an aircraft -type drone that
weighs about 6kg and having flight -times of up to 60
minutes. After processing the obtained images, a
significant correlation be tween the measured gaps
parameters and the biodiversity indicators was found.
The paper suggested that the use of drones and their
high resolution imagery can accurately measure the
canopy gaps parameters which could actually be some
valuable biodiversity indicators.
Also, 28Getzin et al (2014) used the above
described equipment to successfully quantify and
compare spatial gap patterns (Fig. 8) in age-class,
selection -cutting and unmanaged forests. The authors
also recommend ed that f lights should be made at low
altitudes and under low cloud cover conditions ,
without direct sunlight , in order to get very good
images (27Getzin et al , 2012 ).

2.4 Measuring forest canopy height and
attributes
Canopy height is a valuable parameter in
forestry and is normally determined with field
measurements. Currently, LiDAR technology became a
lot more accessible and represent s a new solution in
canopy height measurements (29Lefsky et al, 2002 ),
especially now due to the fact that it can be mounted
on UAVs (30Tulldahl et al, 2014 ; 21Wallace et al , 2014 ).
Also, drones with oblique optical sensor combined with new digital photogrammetry techniques can
provide new methods in measuring canopy height
(31Siebert et al, 2014 ).
In Belgium, 32Lisein et al (2013) used a small
drone (about 2kg), aircraft type, with 40 minutes
autonomy/flight to acquire a set of near -infrared
images with a spatial resolution of 7,6 cm which were
used to measure canopy height. The results obtained
with this equipment were similar to measurements
obtained by a more expensive LiDAR crewed
equipment (Fig. 9).

In Spain, a study on a forest area of 158 ha
that u sed an aircraft type drone with an infrared
camera acquired images with 5 cm spatial resolution.
Authors (33Zarco -Tejada et al, 2014 ) obtained accurate
canopy height measurements (Fig. 10) when
comparing with classical field determinations.

In the U.S.A., 34Dandois and Ellis (2013 ) used
a hexacopter to fly at lower altitudes (below 130m)
over an area of 625 h ectares of deciduous forest and
made observations of the canopy phenology at a high
temporal resolution.
17Chianucci et al ( 2016 ) tested in pure beech
stands if UAV digital true color photography can be
used to estimate canopy attributes. A small fixed -wing
Fig. 7 Very high -resolution
image used for gap analysis
(27Getzin et al , 2012 )
Fig. 8 Study plots with
delineated gaps.
(28Getzin et al , 2014 )
Fig. 9 Comparison between photogrammetric
and LiDAR point clouds (32Lisein et al, 2013 )
Fig. 10 Single tree photo -reconstruction obtain from
UAV imagery, used for tree height quantification
(33Zarco -Tejada et al, 2014 )

UAV (~700g) with a maxim um flight time of ~50 min,
equipped with a standard RGB digital camera was used
to fly at an altitude of 170 m acquiring images with
~7.5 cm resolution per pixel. Authors demonstrated
that in this way it is very effective to obtain cheap,
rapid and useful estimates of forest canopy attributes
at medium -large scales (17Chianucci et al, 2016 ).
Therefore, the studies described above
suggest great potential in using drones for
determination of canopy height and attributes with
much lower costs compared to the solutions provided
by Li DAR technology.

3. Concluding remarks
 Drones are unmanned aircrafts of remarkably
reduced dimensions, very low energy consumption
and low cost for their utilization, no human life being
endangered.
 The current use of drones in forestry applications
is still at an experimental stage, but with grea t
potential in the near future.
 The increasing accessibility in terms of cost and
size for LiDAR and infrared sensors along with data
combining methodologies will highly improve the
utilization of UAVs in forestry.
 Future generations of UAVs will continually evolve
and offer increased flight time and improved sensors.
 Future applications will include studies over a
large range of forestry fields, such as forest dynamics,
species detection, forest disturbance evaluation and
other s, all these with a quick implementation
requirement in the variety of situations that occur in
the sustainable management of forests.

References

1. Shao GF (2012 ). Remote sensing:
Encyclopedia of environmetrics , p. 2187 –
2193 ;
2. Lyons EH (1966) . Fixed air -base 70 mm
photography, a new tool for forest sampling .
For Chron 42 : p. 420–431;
3. Asner G, Knapp D, Broadbent P, Keller M,
Silva J (2005) . Selective logging in the
Brazilian Amazon . Science 310(5747) : p.
480–482;
4. Thakur T, Swamy SL, Nain AS (2014) .
Composition, structure and diversity
characterization of dry tropical forest of
Chhattisgarh using satellite data . J For Res
25(4): p. 819–825;
5. Lu DS (2006) . The potential and challenge of
remote sensing -based biomass estimation .
Int J Remote Sens 27(7): p. 1297 –1328 ;
6. Koch B (2010) . Status and future of laser
scanning, synthetic aperture radar and
hyperspectral remote sensing data for forest
biomass assessment . ISPRS J Photogramm
Remote Sens 65(6): p. 581–590;
7. Alberts K (2012) . Landsat data characteristics
and holdings. A presentation of USGS Landsat
Ground System Lead
(http://www.slideserve.com/keahi/landsat –
data -characteristics -and-holdings) ;
8. Hojas -Gascón L, Belward A, Eva H, Ceccherini
G, Hagolle O, Garcia J, Cerutti P (2015). Potential improvement for forest cover and
forest degradation mapping with the
forthcoming Sentinel -2 program . The
International Archives of the
Photogrammetry, Remote Sensing and
Spatial Information Sciences, Volume XL –
7/W3 : p. 417 -423;
9. Shahbazi M, Theau J, Me nard P (2014) .
Recent applications of unmanned aerial
imagery in natural resource management .
GIScience & Remote Sensing, p. 339 -365;
10. Colomina I, Molina P (2014) . Unmanned
aerial systems for photogrammetry and
remote sensing: a review . ISPRS J
Photogramm Remote Sens 92, p. 79–97;
11. Tang L , Shao G (2015) . Drone remote sensing
for forestry research and practices: review
article , J. For. Res. : p. 791–797;
12. Dudek M, Tomczyk P, Wygonik P, Korkosz M,
Bogusz P, Lis B (2013) . Hybrid fuel cell —
battery system as a main power unit for small
unmanned aerial vehicles (UAV). Int J
Electrochem Sci8(6): p. 8442 –8463 ;
13. Anderson K, Gaston KJ (2013). Lightweight
unmanned aerial vehicles will revolutionize
spatial ecology . Front Ecol Environ 11(3): p.
138–146;
14. Ambrosia VG, Wegener SS, Sullivan DV,
Buechel SW, Brass JA, Dunagan SE, Higgins
RG, Hildum EA, Schoenung SM (2002) .

Demonstrating UAV -Acquired Real -Time
Thermal Data Over Fires . Photogrammetric Engineering and Remote Sensing 69(4): p.
391-402;
15. Ollero A, Merino L (2006). Unmanned aerial
vehicles as tools for forest -fire fighting . For.
Ecol. Manage 234(1): p. 263 -274;
16. Koh LP, Wich SA (2012) . Dawn of drone
ecology: low -cost autonomous aerial vehicles
for conservation . Trop Conserv Sci5: p. 121–
132;
17. Chianucci F, Disperati L, Guzzi D, Bianchini D,
Nardino V, Lastri C, Rindinella A, Corona P
(201 6). Estimation of canopy attributes in
beech forests using true colour digital images
from a small fixed -wing UAV . International
Journal of Applied Earth Observat ion and
Geoinformation 47: p. 60–68;
18. Marks W, Iiames JS , Lunetta RS, Khorram S,
Mace TH (2014) . Basal Area and Biomass
Estimates of Loblolly Pine Stands Using L –
band UAVSAR . Photogrammetric engineering
& Remote Sensing : p. 33 -42;
19. Mokroš M, Tabačák M, Lieskovský M,
Fabrika M (2016). Unmanned aerial vehicle
use for wood chips pile volume estimation.
The International Archives of the
Photogrammetry, Remote Sensing and
Spatial Information Sciences, Volume XLI -B1,
p. 953 -956;
20. Hassaan O, Nasir AK, Roth H, Khan MF
(2016). Precision Forestry: Trees Counting in
Urban Areas Using Visible Imagery based on
an Unmanned Aerial Vehicle . IFAC –
PapersOnLine vol 49, p. 16 -21;
21. Wallace L, Lucieer A,Watson CS, (2014) .
Evaluating Tree Detection and Segmentation
Routines on Very High Resolution UAV LiDAR
Data . IEEE TRANSACTIONS ON GEOSCIENCE
AND REMOTE SENSING, VOL. 52, NO. 12,
DECEMBER 2014 : p. 7619 – 7628 ;
22. Wang YY, Bai GX, Shao GF, Cao YK (2014) .
An analysis of potential investme nt returns
and their determinants of poplar plantations
in state -owned forest enterprises of China .
New Forest 45(2): p. 251 –264;
23. Felderhof L, Gillieson D (2011) . Near -infrared
imagery from unmanned aerial systems and
satellites can be used to specify fert ilizer
application rates in tree crops . Can J Remote
Sens 37(4): p. 376 –386;
24. Zahawi RA, Dandois JP, Holl K, Nadwodny D,
Reid JL, Ellis EC (2015) . Using lightweight
unmanned aerial vehicles to monitor tropical
forest recovery . Biological Conservation 186,
p. 287–295;
25. Lehmann J RK, Nieberd ing F, Prinz T, Knoth C
(2015) . Analysis of Unmanned Aerial System -Based CIR Images in Forestry – A New
Perspective to Monitor Pest Infestation
Levels. Forests 2015, 6,
doi:10.3390/f6030594, ISSN 1999 -4907, p.
594-612;
26. Frolk ing S, Palace MW, Clark DB, Chambers
JQ, Hugart HH, Hurtt GC (2009) . Forest
disturbance and recovery: a general review in
the context of spaceborne remote sensing of
impacts on abovegrou nd biomass and canopy
structure. J Geophys Res, 114: p. 1 -27;
27. Getzin S, Wiegand K, Scho ning I (2012) .
Assessing biodiversity in forests using very
high -resolution images and unmanned aerial
vehicles . Methods Ecol Evol 3(2): p. 397–404;
28. Getzin S, Nuske RS, Wiegand K (2014) . Using
Unmanned Aerial Vehicles (UAV) to Quantify
Spatial Gap Patterns in Forests . Remote Sens.
Journal 2014, 6, p. 6988 -7004;
29. Lefsky MA, Cohen WB, Parker GG, Harding
DJ (2002) . Lidar remote sensing for ecosystem
studies . Bioscience 52(1): p. 19–30;
30. Tulldahl HM, Larsson H. (2014 ). Lidar on
small UAV for 3D mapping. In: Proceedings of
SPIE sponsored conference on electro -optical
remote sensing, photonic technologies, and
applications VIII; and military applications in
hyperspectral imaging and high spatial
resolution sensing II . Amsterdam,
Netherlands, p. 22–23;
31. Siebert S, Teizer J (2014). Mobile 3D mapping
for surveying earthwork projects using an
unmanned aerial vehicle (UAV) system.
Autom Constr 41: p. 1 –14;
32. Lisein J, Pierrot -Deseilligny M, Bonnet S,
Lejeune P (2013) . A photogrammetric
workflow for the cre ation of a forest canopy
height model from small unmanned aerial
system imagery . Forests 4(4): p. 922–944;
33. Zarco -Tejada PJ, Diaz -Varela R, Angileri V,
Loudjani P (2014) . Tree height quantification
using very high resolution imagery acquired
from an unmanne d aerial vehicle (UAV) and
automatic 3D photo -reconstruction methods .
Eur J Agron 55: p. 89–99;
34. Dandois JP, Ellis EC (2013) . High spatial
resolution three -dimensional mapping of
vegetation spectral dynamics using computer
vision . Remote Sens Environ 136: p. 259–276;
35. Website: www. indiamart .com ;
36. Website: www.questuav.com;
37. Website: www.smnrc.org;
38. Website: www.digitaltrends.com .