Mapping Forest Landscape Patterns (eBook)

Tarmo K. Remmel, Ph.D. (University of Toronto), Associate Professor of Geography at York University: A GIScientist with over 10 years of experience teaching and conducting research involving remote sensing, GIS, and spatial statistics, Dr. Remmel focuses primarily on boreal forests, with a particular emphasis on wildfire disturbances and on the development of algorithms for measuring and assessing spatial patterns, planar shapes, and the quantification of spatial change and accuracy. A strong proponent of free and open source software tools, particularly within the R-project to facilitate implementation, his work integrates field-level data collection with remotely sensed imagery obtained from satellite, aircraft, and UAV platforms to characterize the effects of scale.Ajith H. Perera, Ph.D. (Penn State University), Senior Research Scientist and leader of the forest landscape ecology program at Ontario Forest Research Institute, Ontario Ministry of Natural Resources, adjunct professor at University of Waterloo, York University and University of Guelph: With over 25 years research experience in landscape ecology, Dr. Perera’s major focus is on quantifying and modeling spatio-temporal patterns in boreal forest disturbances. He has authored over 150 science publications, and been senior editor, co-editor and author of eight books on Forest Landscape Ecology.

PrefaceChapter 1: Mapping forest landscapes: overview and a primer1. Mapping forest landscapes: an introduction1.1 What is mapping? 1.2 What is a forest landscape? 2. Considerations in forest landscape mapping2.1 Describing spatial patterns2.2 Focus on boundaries2.3 Beyond 2D data3. Utility of forest landscape maps3.1 Map representations3.2 Morphological interpretations3.3 Map scale3.4 Error assessment and validation4. SummaryChapter 2: Fuzzy classification of vegetation for ecosystem mapping1. Introduction2. Overview of fuzzy systems2.1 Fuzzy systems – key concepts for mapping2.2 Mapping with fuzzy classifiers3. Fuzzy approaches for identifying and utilizing uncertainty3.1 Thematic uncertainty3.2 Spatial uncertainty3.3 Simultaneous considerations of thematic and spatial uncertainty3.4 Multiple outputs – fuzzy geodatabase4. Vertical structure mapping5. A look to the future6. SummaryChapter 3: Portraying wildfires in forest landscapes as discrete complex objects1. Introduction2. Wildfire initiation and anatomy2.1 Initiation2.2 Descriptors of footprints3. Wildfires as discrete and complex objects3.1 The outer edge of a wildfire is scale-dependent 3.2 Width of the ecotone3.3 Internal heterogeneity4. Standardized depiction of wildfires as discrete complex objects 5. The future of mapping wildfires5.1 Accuracy assessment in remote regions5.2 Landscape persistence5.3 Hierarchical data formats for capturing scale effectsChapter 4: Airborne LiDAR applications in forest landscapes1. Introduction1.1 Defining ALS LiDAR  1.2 Introduction to the three common LiDAR platforms1.3 Intensity, point density, and multi-spectral LiDAR2. Primary measurements2.1 Surface models (DEM, DSM, DTM, CHM)2.2 Canopy height models and detection and delineation of individual trees3. Secondary measurements3.1 Regression models and allometric equations3.2 Vertical profile for a single tree3.3 Classification of vegetation types3.4 Tree genus and species classification3.5 Case study: identifying potentially hazardous trees4. The future of LiDARChapter 5: Regression Tree modeling of spatial pattern and process interactions1. Spatial Pattern and Processes1.1 Describing spatial patterns1.2 Process complexity1.3 Data mining2. Methods2.1 CART models2.2 BRT2.3 RF models3. Case Study Context – Influence of beetle infestation spatial patterns on fire spatial processes3.1 Study area3.2 Spatial data4. Model evaluation4.1 CART4.2 BRTs4.3 RF models4.4 Comparing modeling approaches5. Interpreting regression tree results within the context of spatial pattern and process Chapter 6: Mapping the abstractions of forest landscape patterns1. Introduction2. Tools for evaluating landscape patterns3. Data preparation and uncertainties within metrics 3.1 Scale and classification issues4. Mapping different aspects of a landscape pattern4.1 Composition4.2 Configuration4.3 Criteria for selecting metrics5. Applications of forest pattern mapping5.1 Improving forest management5.2 Assessment of forest habitats5.3 Mapping landscape metrics by using GIS5.4 Using landscape metrics in modeling6. Future perspectives on mapping patterns6.1 3D landscape metrics6.2 4D landscape metrics7. ConclusionsChapter 7: Towards automated forest mapping1. Introduction1.1 Definitions1.1.1 Forest 1.1.2 Remote sensing for automated mapping of woodland and forest2. Data and Pre-processing2.1 Reference data 2.2 Remote sensing systems2.3 Processing of input data sets3. Mapping woodland 3.1 A hierarchical segmentation approach for mapping woodland 3.2 Individual tree and tree crown detection 3.3 Fractional tree cover approach4. Forest mapping4.1 Moving window approach4.2 Distance criterion approach5. Lessons learned6. Future perspectivesEpilogue: Toward more efficient and effective applications of forest landscape maps1. Background2. Goals of this chapter3. Considerations in forest landscape mapping3.1. The community of map developers and users is broad3.2. Maps are model outputs 3.3. Maps are probabilistic3.4. Maps contain errors3.5. Map contents are scale-related3.6. Map applications are scale-related3.7. Mapping methods are advancing rapidly4. A brief list of best practices for using forest landscape maps5. Conclusions