Mapping forest landscape patterns

333 Pages·2017·14.717 MB·English

by  Perera, Ajith H., Remmel, Tarmo K

Tarmo K. Remmel 
Ajith H. Perera Editors
Mapping
Forest
Landscape
Patterns
Mapping Forest Landscape Patterns
Tarmo K. Remmel • Ajith H. Perera
Editors
Mapping Forest Landscape
Patterns
Editors
Tarmo K. Remmel Ajith H. Perera
Department of Geography Ontario Forest Research Institute
York University Ministry of Natural Resources and Forestry
Toronto, ON, Canada Sault Ste. Marie, ON, Canada
ISBN 978-1-4939-7329-3 ISBN 978-1-4939-7331-6 (eBook)
DOI 10.1007/978-1-4939-7331-6
Library of Congress Control Number: 2017951195
© Springer Science+Business Media LLC 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature
The registered company is Springer Science+Business Media LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface
Forest landscape maps provide the foundational information for managing, con-
serving, and utilizing forests—one of the most important global resources. Use of
the spatially and temporally explicit knowledge generated by these maps, once lim-
ited to professionals engaged in research, management, and education, is now pos-
sible for nonprofessionals, including the general public. Recent interest in global
issues such as conserving biodiversity, the impacts of anthropogenic activities, and
the effects of climate change have further increased the demand for mapped infor-
mation about forest landscapes. This demand is being met by a rapidly expanding
capability to supply that information readily and inexpensively. Given this positive
and dynamic scenario, and the growing enthusiasm of academics and students, it is
perhaps time to pause and ponder the status quo for using mapped information.
Thus, in this book, we examine the mapping of forest landscapes with the goals of
providing a broad overview of the current status of mapping and highlighting some
advances in the techniques available for obtaining and visualizing this information.
It is not intended to be an exhaustive review of the state of our current knowledge of
forest landscape mapping, but rather a recap of some key aspects related to the util-
ity and use of these maps.
The book is composed of eight chapters. We begin with a broad introduction to
mapping forest landscapes, which will serve as a primer for readers who are either
not well versed in this topic or who need a refresher course in the basics. Chapter 1
reminds the reader that maps are abstractions of reality, and that the degree of the
abstraction is influenced by the map’s scale and the developers’ decisions regarding
the geometric projection and representation; forest landscapes represent a high level
of ecological organization that includes many hierarchically structured elements at
multiple scales, and the elements we map are often fuzzy and heterogeneous with
respect to their attributes and geometry. Data for maps are gathered from numerous
sources at multiple scales; the collection can span many sources, so assessments of
the inherent errors and of map accuracy are essential for defining the validity and
credibility of maps.
Chapter 2 describes the concept of mapping forest landscapes as fuzzy elements,
since both classification themes and the boundaries of mapped entities are often
v
vi Preface
unknown or imprecise. This chapter describes the approaches that can be used to
obtain probabilistic and membership values for mapped ecosystems, how the mem-
bership functions that provide these values can take many forms that depend on the
spatial resolution, and the consequences for the uncertainty of mapped themes and
spatial details.
Chapter 3 addresses the specific case of mapping wildfires in forest landscapes
as discrete but complex objects. It describes how wildfire footprints, despite com-
monly being mapped as simple polygons with a uniform interior and a definitive
boundary, are highly complex with respect to their internal composition and spatial
attributes. This complexity results from the heterogeneous and stochastic nature of
the processes that underlie fire behavior. Accurate portrayal of these processes is
scale-dependent.
Chapter 4 describes the utility of three-dimensional mapping of forest landscapes
using airborne light detection and ranging (LiDAR). This chapter compares the uses
of discrete and full-waveform LiDAR data collection and discusses the types of
summary statistics about forests that each can provide, including various character-
izations of tree crowns and the development of canopy and terrain models. Having
access to 3D forest landscape data has implications for mapping biomass and stem
density that go beyond traditional two-dimensional stand-level genus or species
classifications.
Chapter 5 explores the mapping of outputs from spatial models to understand the
interactions among processes from the resulting spatial patterns. Such mapping
often extends into the construction of mathematical surfaces, for which the mapped
quantities represent the outputs of a model (e.g., CART, Random Forests) and are
themselves abstractions of a landscape. In such cases, the spatial patterns can be
used as a surrogate for studying complex ecological processes, and ensemble meth-
ods can be used to integrate data within analyses of growing archives of spatial data.
Chapter 6 continues the discussion of landscape abstraction through the use of
landscape metrics in two, three, and four dimensions. Although numerous metrics
exist, most capture some aspect of the landscape’s composition or configuration by
summarizing the size, density, shape, core, edge, or connectivity characteristics of
specific land cover types. Transition zones (e.g., ecotones, edges) can be visualized
as concentric bands, which are conceptually simpler than fuzzy membership func-
tions in terms of how they describe changes with respect to the distance across an
interface, since fuzzy membership functions can take on highly complex forms.
Chapter 7 explores the automation of data processing to standardize the produc-
tion of forest maps and to ensure both consistency and more rapid map develop-
ment. Ensuring the effectiveness of such workflows will require consistent
terminology in the context of an increasingly automated environment. In this chap-
ter, we reiterate the distinction between land use and land cover, particularly since
the important information obtained from land use classification remains difficult to
extract from remote-sensing imagery. The chapter concludes with a discussion of
tools that can integrate this imagery with LiDAR data in a logistic regression model
to allow interpretation of fractional tree cover, which connects neatly with the con-
Preface vii
cept and fuzziness or mixed pixels that contain various combinations of land cover
types.
The book concludes with a brief synopsis of the need to improve the application
of forest landscape maps, particularly in terms of the efficiency and effectiveness of
using the mapped information. In this epilogue, we remind readers that all applica-
tions of maps are scale-related; the best information is not necessarily the most
detailed or presented at the highest resolution, but rather uses the optimal (most
appropriate) scale for a specific application. We also reiterate that maps are abstrac-
tions and simplifications of complex natural systems, and that these abstractions
depend on many assumptions and are undermined by the many sources of error that
are associated with any map.
Our target audience for this book is readers who are involved with generating and
using forest landscape maps. We anticipate this volume to benefit readers from the
communities of developers and users of geospatial data about forest landscapes.
Throughout the book, we emphasize how creators and users of maps must actively
and continuously communicate their needs and interact to achieve those needs, with
the ultimate goal being to improve the efficiency and effectiveness of creating and
using forest landscape maps.
On behalf of all authors who contributed to this book, we thank the colleagues
who critically reviewed drafts of the chapter manuscripts and suggested improve-
ments: Raivo Aunap, Den Boychuk, Nicholas Coops, Christoph Fischer, Curtis
Gautschi, Wendy Goetz, Emilie Henderson, John Lindsay, Jed Long, Kevin
McGarigal, Don McKenzie, Scott Mitchell, Achilleas Psomas, Marc Simard,
Christoph Straub, and Hannah Wilson. Finally, we gratefully acknowledge Geoff
Hart for improving the clarity of the messages in several chapters, Marc Ouellette
for refining many images and illustrations, and Janet Slobodien for guiding the pub-
lication process.
Tarmo K. Remmel
Toronto, ON, Canada
Ajith H. Perera
Sault Ste. Marie, ON, Canada
Contents
Mapping Forest Landscapes: Overview and a Primer . . . . . . . . . . . . . . . . . 1
Tarmo K. Remmel and Ajith H. Perera
Fuzzy Classification of Vegetation for Ecosystem Mapping . . . . . . . . . . . . . 63
F. Jack Triepke
Portraying Wildfires in Forest Landscapes as Discrete Complex
Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Tarmo K. Remmel and Ajith H. Perera
Airborne LiDAR Applications in Forest Landscapes . . . . . . . . . . . . . . . . . . 147
Connie Ko and Tarmo K. Remmel
Regression Tree Modeling of Spatial Pattern
and Process Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Trisalyn A. Nelson, Wiebe Nijland, Mathieu L. Bourbonnais,
and Michael A. Wulder
Mapping the Abstractions of Forest Landscape Patterns . . . . . . . . . . . . . . . 213
Evelyn Uuemaa and Tõnu Oja
Towards Automated Forest Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Lars T. Waser, Ruedi Boesch, Zuyuan Wang, and Christian Ginzler
Epilogue: Toward More Efficient and Effective Applications
of Forest Landscape Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Ajith H. Perera and Tarmo K. Remmel
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
ix
Contributors
Ruedi Boesch Swiss Federal Institute for Forest, Snow and Landscape Research,
Birmensdorf, Switzerland
Mathieu L. Bourbonnais Spatial Pattern Analysis and Research (SPAR) Laboratory,
Department of Geography, University of Victoria, Victoria, BC, Canada
Christian  Ginzler Swiss Federal Institute for Forest, Snow and Landscape
Research, Birmensdorf, Switzerland
Connie Ko Department of Geography, York University, Toronto, ON, Canada
Trisalyn A. Nelson School of Geographical Sciences and Urban Planning, Arizona
State University, Tempe, AZ, Canada
Wiebe  Nijland Spatial Pattern Analysis and Research (SPAR) Laboratory,
Department of Geography, University of Victoria, Victoria, BC, Canada
Tõnu Oja Department of Geography, University of Tartu, Tartu, Estonia
Ajith H. Perera Ontario Forest Research Institute, Ministry of Natural Resources
and Forestry, Sault Ste.Marie, ON, Canada
Tarmo K. Remmel Department of Geography, York University, Toronto, ON,
Canada
F. Jack Triepke USDA Forest Service, Albuquerque, NM, USA
Evelyn Uuemaa Department of Geography, University of Tartu, Tartu, Estonia
Zuyuan Wang Swiss Federal Institute for Forest, Snow and Landscape Research,
Birmensdorf, Switzerland
Lars T. Waser Swiss Federal Institute for Forest, Snow and Landscape Research,
Birmensdorf, Switzerland
Michael A. Wulder Canadian Forest Service, Natural Resources Canada, Pacific
Forestry Centre, Victoria, BC, Canada
xi
Abbreviations
0D Zero-dimensional
1D One-dimensional
2D Two-dimensional
3D Three-dimensional
ALS Airborne laser scanning
AVHRR The advanced very high resolution radiometer
BA Basal area
BRT Boosted regression trees
CART Classification and regression trees
CHM Canopy height model
CIR Color-infrared
DBH Diameter at breast height
DEM Digital elevation model
DSM Digital surface model
DTM Digital terrain model
DMSP Defense meteorological satellite program
FD Fractal dimension
GIS Geographical information systems
GLAS Geoscience laser altimeter system
GLM Generalized linear model
GPS Global positioning system
GSD Ground sample distance
ICESat Ice, cloud and land elevation satellite
IGBP International geosphere biosphere program
IMU Inertial measurement unit
IP Interpretation plot
ITC Individual tree crowns
k-NN k-nearest neighbor
LiDAR Light detection and ranging
L The travel time of a light pulse
t
MAUP Modifiable areal unit problem
xiii