The Roles of Remote Sensing in Nature Conservation (eBook)

Dr. Ricardo Diaz-Delgado, CSIC-Estación Biológica de Doñana EBD, Sevilla, SpainProf. Richard Lucas, The University of New South Wales, School of Bio., Earth, Env. Science, Kensington, AustraliaDr. Clive Hurford, Natural Resources Wales, Pembroke Wales, United Kingdom

Acknowledgements 6
Contents 7
Contributors 9
Editors 9
Authors 9
Part I: An Introduction to Remote Sensing Tools for Habitat Mapping and Monitoring 11
Introducing the Book “The Roles of Remote Sensing in Nature Conservation” 12
The Need for a Book on RS and Conservation: Filling the gap 12
Book Structure 13
The Feeling of Managers: A Survey 15
References 19
Towards a Mature Age of Remote Sensing for Natura 2000 Habitat Conservation: Poor Method Transferability as a Prime Obstacle 20
Introduction 21
Transferability as a Critical Success Factor: The Case of Natura 2000 Habitat Monitoring 22
Possible Causes of Poor Transferability 23
Objects of Interest Differ (Even if They Carry the Same Name) 23
User Requirements Abound and Differ (Even if They Appear Similar) 24
Available Ground Reference Data Differ 25
Image Properties Differ 26
Real-World Examples 28
Transfer Test Case 1: To Later Dates 29
Study Site 29
Method Application 29
Transfer Test Case 2: To a Different Site – From Atlantic to Continental Heathland 32
Study Site 32
Method Application 33
Transfer Test Case 3: To a Different Ecosystem – from Heathlands to Coastal Dunes 34
Study Site 34
Method Application 34
Towards Better Transferability 37
Use More Widely Available Image Data 38
Limit the Dependence on Ground Reference Data 38
Exploit the Strengths of Remote Sensing 39
More Considerations for Repeatability and Transferability 40
Conclusions 40
Annex 41
References 44
Pre-processing of Remotely Sensed Imagery 47
Introduction 47
Geometric Correction of Airborne and Spaceborne Data 48
Optical Data 50
Radiance 51
At Sensor Radiance 53
Surface Reflectance 53
Empirical Line Calibration 54
Dark Object Subtraction (DOS) 55
Modeled Atmospheres 55
Light Detection and Ranging (LiDAR) 61
Overview of Products and Software 61
Classification of Point Clouds 62
Standard Raster Products 64
Interpolation of Elevation Surfaces 64
Derivation of Metrics for Vertical Vegetation Structure 65
Radiometric Correction 65
Standard Data Specifications 65
Synthetic Aperture Radar (SAR) 66
Overview of SAR 66
Geometric Correction 67
Defining Sigma Nought (?0) and Gamma Nought (?0) 67
SAR Image Filtering 68
Conclusions 68
References 69
Long-Term Ecological Monitoring at the Landscape Scale for Nature Conservation: The Example of Doñana Protected Area 72
Introduction 72
Long-Term Ecological Monitoring at Landscape Scale in the Doñana Protected Area 73
Ground-Truthing 74
Monitoring of Doñana Wetlands 74
Monitoring of Geophysical Processes 76
Monitoring of Terrestrial Plant Communities 77
The Pros and Cons of Integrating Remote Sensing in Nature Conservation 79
Conclusions 79
References 80
Part II: Habitat Case Studies 84
NILS – A Nationwide Inventory Program for Monitoring the Conditions and Changes of the Swedish Landscape 85
Introduction 86
Design of Monitoring in NILS 87
Inventory by Remote Sensing 89
Challenges in Using Remote Sensing for Monitoring 93
References 94
Mapping Coastal Habitats in Wales 97
Introduction 98
Background 99
Habitat Mapping with Remote Sensing 99
Image Analysis Techniques 100
Study Area 103
Methods 105
Data Collection and Pre-processing of Earth Observation (EO) Data 105
Very High Resolution Satellite Imagery 105
LiDAR 106
Data from Unmanned Aerial Vehicles (UAVs) and Aerial Photography 106
Field Data 107
Index Calculation 107
EODHaM System and ANOVA Separation 108
Detailed Habitat Mapping 111
Algorithm Selection 111
Input Feature Selection 112
Classification Structure and Validation 113
Results 115
Discussion and Conclusions 116
References 123
Integrated Monitoring for Biodiversity Using Remote Sensing: From Local to Regional – A Case Study from Norfolk 127
A ‘Multi-scale’ Approach with Ecological Input 128
Why Adopt a Multi-scale Approach to Mapping? 128
Mapping Approach 129
Principles Guiding Mapping at all Scales 130
Spectral and Textural 130
The Crick Framework 130
Mapping to ‘Real World’ Objects 132
Landscape Context: Confining the Search for Habitats 133
Placing Boundaries in Ecotones 133
Selecting Imagery for Classification 134
Local to Regional Scale Outputs in Norfolk 135
Sub Regional Mapping: Considerations and Outputs 137
Site Level Mapping: Considerations and Outputs 139
Towards a Living Map 141
Examples of the Use of the Mapped Outputs in Norfolk 142
Regional Scale and Sub-regional Mapping 142
Site Level Mapping 143
Conclusion 144
References 144
Sub-pixel Mapping of Doñana Shrubland Species 146
Introduction 146
Background of Plant Species Mapping Activities Using Imaging Spectroscopy 148
Airborne Imaging Spectroscopy 148
Field Spectroscopy 149
Spectral Unmixing 150
Plant Species Mapping 151
Field Data 153
Airborne Imaging Spectroscopy 154
Species Mapping Generation 155
Practical Case: Mapping Shrublands Species of Doñana National Park 155
Study Site 155
Shrub Communities 156
INTA AHS System 157
Doñana Shrub-Species Mapping 158
Results 161
Discussion 163
Conclusions 165
References 166
Part III: Species-Driven Remote Sensing and New Technologies Studies 169
Mapping the Distribution of Understorey Rhododendron Ponticum Using Low-Tech Multispectral UAV Derived Imagery 170
Introduction 171
Species Characteristics and Invasion 171
Remote Sensing Background 172
Study Area 174
Methods 175
Cameras, UAVs and Flight Design 175
Image Pre-processing: Distortion Removal 175
Image Mosaicking 175
Mosaic Alignment 176
Multispectral Image Generation 176
Spectral Checkerboard 177
Pixel-Based Digital Classification and Statistical Analysis 178
Results 179
Distortion Effects and Removal 179
Spectral Analysis 179
Digital Classification 180
Discussion 180
Feasability Study 180
Limitations and Cost-Benefit Analysis 181
Future Directions for Nature Conservation 183
References 183
The Potential of UAV Derived Image Features for Discriminating Savannah Tree Species 185
Introduction 186
Background 186
Study Area 187
Methods 188
UAV Imagery Acquisition 188
Ground Truth Data 189
RGB and NIR Spectral Indices 190
Image Texture 192
Canopy Height Model 193
Random Forest Classification 193
Results 194
Discussion 197
Conclusions 200
Practical Application for Nature Conservation 200
References 201
A Toolbox for Remotely Monitoring Large Carnivores in Sweden 204
Introduction 205
A Compensation System for Reindeer 205
Research, Monitoring and Management 206
Methods for Surveillance 207
Snow Tracking 209
Nest and Den Surveys 210
Sampling and Analysing DNA 211
Tagging with Biotelemetry Sensors 213
Automatic Cameras 215
Aerial Surveys 216
Reports from Moose Hunters 216
Reports from the Public 217
Attitude Surveys 217
Handling Loads of Data 218
The Scandinavian Carnivore Database Rovbase 218
The Wireless Remote Animal Monitoring Database E-Infrastructure 219
Geodata and Their Integration with Tracking Data 220
Summarizing Conclusions 221
References 221
Coupling Field Sampling with Earth Observation Increases Understanding of Tiger Movement and Behaviour 224
Introduction 225
Background 225
Why Are Tigers Endangered? 225
Determining the Distribution and Abundance of Tigers 227
Use of Remote Sensing Data for Tiger Habitat Mapping 228
The Indo-Bhutan Manas Tiger Conservation Landscape (IBMTCL) and Its Conservation Significance 229
Data Collected 230
Mammal Surveys 230
Remote Sensing and Other Spatial Data 232
Methods of Data Analysis 232
Analysis of Camera Trap Data 232
Spatial Analysis of Field Data 233
Land Cover Classification 234
Carnivore Distribution in Relation to Anthropogenic Disturbance 236
Results 236
Camera Trap Sightings 236
Ranging Pattern and Seasonal Home Range 237
Land Cover Classifications and Carnivore Preferences 239
Distributions of Tigers in Relation to Fires and Anthropogenic Disturbance 240
Discussion 241
The Role of Earth Observation 241
Distribution and Movement of Tigers 243
Final Conclusions 244
References 245
Improving the Accuracy of Bird Counts Using Manual and Automated Counts in ImageJ: An Open-Source Image Processing Program 249
Introduction 250
The Levels of Observer Variation Associated with Bird Counts 250
ImageJ Software 255
Case Study and Image Locations 256
Methods 256
Image Collection 256
Camera Equipment 257
Camera Settings 257
ISO Settings 257
The f-stop 257
Shutter Speed 258
Capturing the Images 259
Capturing Vertical Images of Feeding or Breeding Birds 259
Capturing Images of Flocks of Birds in Flight 260
Carrying Out Counts in ImageJ 260
Manual Counts 261
Manual Count Validation 261
Automatic counts 262
Automatic Count Validation 262
Results 262
A Manual Count of Guillemots at a Breeding Colony at Elegug Stacks, Pembrokeshire, UK 262
An Automatic Count of a Starling Murmuration at Aberystwyth, Ceredigion, UK 264
Starling Case Study Results and Validation 264
Validation 264
An Automatic Count of Wintering Greater Flamingos (Phoenicopterus roseus) in Flight over Doñana NP in South-­West Spain 267
An Automatic Count of Wintering Greater Flamingos (Phoenicopterus roseus) Feeding in a Lagoon at Doñana NP in Andalucia, Spain 267
An Automatic Count of a Flock of Linnets (Carduelis cannabina) in Flight over Stubble Fields at Angle, Pembrokeshire, Wales 271
An Automatic Count of Golden Plovers (Pluvialis apricaria) in Flight over Farmland Near Castlemartin, Pembrokeshire, UK 272
Discussion 274
The Opportunities for Counting Birds in the Future 275
Conclusions 275
References 276
Using UAVs to Map Aquatic Bird Colonies 277
Overview 278
Introduction 278
Background 279
Targeted Species: The Slender-billed Gull 280
Study Area: The Doñana Colonies 281
Methods 282
UAVs Flight Campaigns 282
Geometric Processing 283
Methods for Identifying Birds From UAV Imagery 284
Photointerpretation 284
Supervised Classification 284
Colony Monitoring 285
Results 285
Visual Identification of Birds 285
Image Automatic Classification 287
Colony Monitoring 288
Discussion 288
Conclusions 289
References 290
Part IV: Looking Ahead Through Current Research Projects and Expected Advances 292
Integrated Land Cover and Change Classifications 293
Introduction 293
Recognizing User Needs 295
The EODESM System 295
Our Unchanging World 296
Classification of Land Covers 298
Classification of Change 301
Causes and Consequences of Change 303
Concluding Remarks 304
References 305
Expected Advances in a Rapidly Developing Work Area 307
Global Driving Forces Over Protected Areas 308
Where to From Here? The Nature Conservation Perspective 309
Where to From Here? A Remote Sensing Perspective 310
Realistic Expectations 313
Concluding Remarks: What Has Remote Sensing Done for Us? 315
References 315