Evaluating unmanned aerial vehicle images for estimating forest canopy fuels in a ponderosa pine stand

Patrick Shin, Temuulen Sankey, Margaret M Moore, Andrea E Thode

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Forests in the Southwestern United States are becoming increasingly susceptible to large wildfires. As a result, forest managers are conducting forest fuel reduction treatments for which spatial fuels and structure information are necessary. However, this information currently has coarse spatial resolution and variable accuracy. This study tested the feasibility of using unmanned aerial vehicle (UAV) imagery to estimate forest canopy fuels and structure in a southwestern ponderosa pine stand. UAV-based multispectral images and Structure-from-Motion point clouds were used to estimate canopy cover, canopy height, tree density, canopy base height, and canopy bulk density. Estimates were validated with field data from 57 plots and aerial photography from the U.S. Department of Agriculture National Agriculture Imaging Program. Results indicate that UAV imagery can be used to accurately estimate forest canopy cover (correlation coefficient (R2) = 0.82, root mean square error (RMSE) = 8.9%). Tree density estimates correctly detected 74% of field-mapped trees with a 16% commission error rate. Individual tree height estimates were strongly correlated with field measurements (R2 = 0.71, RMSE = 1.83 m), whereas canopy base height estimates had a weaker correlation (R2 = 0.34, RMSE = 2.52 m). Estimates of canopy bulk density were not correlated to field measurements. UAV-derived inputs resulted in drastically different estimates of potential crown fire behavior when compared with coarse resolution LANDFIRE data. Methods from this study provide additional data to supplement, or potentially substitute, traditional estimates of canopy fuel.

Original languageEnglish (US)
Article number1266
JournalRemote Sensing
Volume10
Issue number8
DOIs
StatePublished - Aug 1 2018

Fingerprint

forest canopy
canopy
bulk density
imagery
agriculture
fire behavior
multispectral image
aerial photography
vehicle
wildfire
spatial resolution

Keywords

  • Base height
  • Bulk density
  • Cover
  • Drone
  • Fire behavior
  • Lidar
  • SfM
  • Structure-from-motion
  • Unmanned aerial vehicle (UAV)
  • Wildfire

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)

Cite this

Evaluating unmanned aerial vehicle images for estimating forest canopy fuels in a ponderosa pine stand. / Shin, Patrick; Sankey, Temuulen; Moore, Margaret M; Thode, Andrea E.

In: Remote Sensing, Vol. 10, No. 8, 1266, 01.08.2018.

Research output: Contribution to journalArticle

@article{4da4ca85303d4f1985ea9c458242f2bc,
title = "Evaluating unmanned aerial vehicle images for estimating forest canopy fuels in a ponderosa pine stand",
abstract = "Forests in the Southwestern United States are becoming increasingly susceptible to large wildfires. As a result, forest managers are conducting forest fuel reduction treatments for which spatial fuels and structure information are necessary. However, this information currently has coarse spatial resolution and variable accuracy. This study tested the feasibility of using unmanned aerial vehicle (UAV) imagery to estimate forest canopy fuels and structure in a southwestern ponderosa pine stand. UAV-based multispectral images and Structure-from-Motion point clouds were used to estimate canopy cover, canopy height, tree density, canopy base height, and canopy bulk density. Estimates were validated with field data from 57 plots and aerial photography from the U.S. Department of Agriculture National Agriculture Imaging Program. Results indicate that UAV imagery can be used to accurately estimate forest canopy cover (correlation coefficient (R2) = 0.82, root mean square error (RMSE) = 8.9{\%}). Tree density estimates correctly detected 74{\%} of field-mapped trees with a 16{\%} commission error rate. Individual tree height estimates were strongly correlated with field measurements (R2 = 0.71, RMSE = 1.83 m), whereas canopy base height estimates had a weaker correlation (R2 = 0.34, RMSE = 2.52 m). Estimates of canopy bulk density were not correlated to field measurements. UAV-derived inputs resulted in drastically different estimates of potential crown fire behavior when compared with coarse resolution LANDFIRE data. Methods from this study provide additional data to supplement, or potentially substitute, traditional estimates of canopy fuel.",
keywords = "Base height, Bulk density, Cover, Drone, Fire behavior, Lidar, SfM, Structure-from-motion, Unmanned aerial vehicle (UAV), Wildfire",
author = "Patrick Shin and Temuulen Sankey and Moore, {Margaret M} and Thode, {Andrea E}",
year = "2018",
month = "8",
day = "1",
doi = "10.3390/rs10081266",
language = "English (US)",
volume = "10",
journal = "Remote Sensing",
issn = "2072-4292",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "8",

}

TY - JOUR

T1 - Evaluating unmanned aerial vehicle images for estimating forest canopy fuels in a ponderosa pine stand

AU - Shin, Patrick

AU - Sankey, Temuulen

AU - Moore, Margaret M

AU - Thode, Andrea E

PY - 2018/8/1

Y1 - 2018/8/1

N2 - Forests in the Southwestern United States are becoming increasingly susceptible to large wildfires. As a result, forest managers are conducting forest fuel reduction treatments for which spatial fuels and structure information are necessary. However, this information currently has coarse spatial resolution and variable accuracy. This study tested the feasibility of using unmanned aerial vehicle (UAV) imagery to estimate forest canopy fuels and structure in a southwestern ponderosa pine stand. UAV-based multispectral images and Structure-from-Motion point clouds were used to estimate canopy cover, canopy height, tree density, canopy base height, and canopy bulk density. Estimates were validated with field data from 57 plots and aerial photography from the U.S. Department of Agriculture National Agriculture Imaging Program. Results indicate that UAV imagery can be used to accurately estimate forest canopy cover (correlation coefficient (R2) = 0.82, root mean square error (RMSE) = 8.9%). Tree density estimates correctly detected 74% of field-mapped trees with a 16% commission error rate. Individual tree height estimates were strongly correlated with field measurements (R2 = 0.71, RMSE = 1.83 m), whereas canopy base height estimates had a weaker correlation (R2 = 0.34, RMSE = 2.52 m). Estimates of canopy bulk density were not correlated to field measurements. UAV-derived inputs resulted in drastically different estimates of potential crown fire behavior when compared with coarse resolution LANDFIRE data. Methods from this study provide additional data to supplement, or potentially substitute, traditional estimates of canopy fuel.

AB - Forests in the Southwestern United States are becoming increasingly susceptible to large wildfires. As a result, forest managers are conducting forest fuel reduction treatments for which spatial fuels and structure information are necessary. However, this information currently has coarse spatial resolution and variable accuracy. This study tested the feasibility of using unmanned aerial vehicle (UAV) imagery to estimate forest canopy fuels and structure in a southwestern ponderosa pine stand. UAV-based multispectral images and Structure-from-Motion point clouds were used to estimate canopy cover, canopy height, tree density, canopy base height, and canopy bulk density. Estimates were validated with field data from 57 plots and aerial photography from the U.S. Department of Agriculture National Agriculture Imaging Program. Results indicate that UAV imagery can be used to accurately estimate forest canopy cover (correlation coefficient (R2) = 0.82, root mean square error (RMSE) = 8.9%). Tree density estimates correctly detected 74% of field-mapped trees with a 16% commission error rate. Individual tree height estimates were strongly correlated with field measurements (R2 = 0.71, RMSE = 1.83 m), whereas canopy base height estimates had a weaker correlation (R2 = 0.34, RMSE = 2.52 m). Estimates of canopy bulk density were not correlated to field measurements. UAV-derived inputs resulted in drastically different estimates of potential crown fire behavior when compared with coarse resolution LANDFIRE data. Methods from this study provide additional data to supplement, or potentially substitute, traditional estimates of canopy fuel.

KW - Base height

KW - Bulk density

KW - Cover

KW - Drone

KW - Fire behavior

KW - Lidar

KW - SfM

KW - Structure-from-motion

KW - Unmanned aerial vehicle (UAV)

KW - Wildfire

UR - http://www.scopus.com/inward/record.url?scp=85051623216&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051623216&partnerID=8YFLogxK

U2 - 10.3390/rs10081266

DO - 10.3390/rs10081266

M3 - Article

AN - SCOPUS:85051623216

VL - 10

JO - Remote Sensing

JF - Remote Sensing

SN - 2072-4292

IS - 8

M1 - 1266

ER -