Forest inventory : methodology and applications

This book has been developed as a forest inventory textbook for students and could also serve as a handbook for practical foresters. We have set out to keep the mathematics in the book at a fairly non-technical level, and therefore, although we deal with many issues that include highly sophisticated methodology, we try to present first and foremost the ideas behind them. For foresters who need more details, references are given to more advanced scientific papers and books in the fields of statistics and biometrics. Forest inventory books deal mostly with sampling and measurement issues, as found here in section I, but since forest inventories in many countries involve much more than this, we have also included material on forestry applications. Most applications nowadays involve remote sensing technology of some sort, so that section II deals mostly with the use of remote sensing material for this purpose. Section III deals with national inventories carried out in different parts of world, and section IV is an attempt to outline some future possibilities of forest inventory methodologies. The editors, Annika Kangas Professor of Forest Mensuration and Management, Department of Forest Resource Management, University of Helsinki. Matti Maltamo Professor of Forest Mensuration, Faculty of Forestry, University of Joensuu. ACKNOWLEDGEMENTS

Preface

Acknowledgements

List of contributing authors

Part I Theory

1 IntroductionA. Kangas et al.

1.1 General

1.2 Historical background of sampling theory

1.3 History of forest inventories

References

2 Design-based sampling and inferenceA. Kangas

2.1 Basis for probability sampling

2.2 Simple random sampling

2.3 Determining the sample size

2.4 Systematic sampling

2.5 Stratified sampling

2.6 Cluster sampling

2.7 Ratio and regression estimators

2.8 Sampling with probability proportional to size

2.9 Non-linear estimators

2.10 Resampling

2.11 Selecting the sampling method

References

3 Model-based inferenceA. Kangas

3.1 Foundations of model-based inference

3.2 Models

3.3 Applications of model-based methodsto forest inventory

3.4 Model-based versus design-based inference

References

4 Mensurational aspectsA. Kangas

4.1 Sample plots

4.1.1 Plot size

4.1.2 Plot shape

4.2 Point sampling

4.3 Comparison of fixed-sized plots and points

4.4 Plots located on an edge or slope

4.4.1 Edge corrections

4.4.2 Slope corrections.References

5 Change monitoring with permanent sample plotsS. Poso

5.1 Concepts and notations

5.2 Choice of sample plot type and tree measurement

5.3 Estimating components of growth at the plot level

5.4 Monitoring volume and volume increment over two or more measuring periods at the plot level

5.5 Estimating population parameters

5.6 Concluding remarks

References

6 Generalizing sample tree informationJ. Lappi et al.

6.1 Estimation of tally tree regression

6.2 Generalizing sample tree information in a small subpopulation

6.2.1 Mixed estimation

6.2.2 Applying mixed models

6.3 A closer look at the three-level model structure.References

7 Use of additional informationJ. Lappi and A. Kangas

7.1 Calibration estimation

7.2 Small area estimates

References

8 Sampling rare populationsA. Kangas

8.1 Methods for sampling rare populations

8.1.1 Principles

8.1.2 Strip sampling

8.1.3 Line intersect sampling

8.1.4 Adaptive cluster sampling

8.1.5 Transect and point relascope sampling

8.1.6 Guided transect sampling

8.2 Wildlife populations

8.2.1 Line transect sampling

8.2.2 Capture-recapture methods

8.2.3 The wildlife triangle scheme

References

9 Inventories of vegetation, wild berries and mushroomsM. Maltamo

9.1 Basic principles

9.2 Vegetation inventories

9.2.1 Approaches to the description of vegetation

9.2.2 Recording of abundance

9.2.3 Sampling methods for vegetation analysis

9.3 Examples of vegetation surveys

9.4 Inventories of mushrooms and wild berries

References

10 Assessment of uncertainty in spatially systematic samplingJ. Heikkinen

10.1 Introduction

10.2 Notation, definitions and assumptions

10.3 Variance estimators based on local differences

10.3.1 Restrictions of SRS-estimator

10.3.2 Development of estimators based on local differences

10.4 Variance estimation in the national forest inventory in Finland

10.5 Model-based approaches

10.5.1 Modelling spatial variation

10.5.2 Model-based variance and its estimation

10.5.3 Descriptive versus analytic inference

10.5.4 Kriging in inventories

10.6 Other sources of uncertainty

References

Part II Applications

11 The Finnish national forest inventoryE. Tomppo

11.1 Introduction

11.2 Field sampling system used in NFI9

11.3 Estimation based on field data

11.3.1 Area estimation

11.3.2 Volume estimation

11.3.2.1 Predicting sample tree volumes and volumes by timber assortment classes

11.3.2.2 Predicting volumes for tally trees

11.3.3.3 Computing volumes for computation units

11.4 Increment estimation

11.5 Conclusions

References

12 The Finnish multi-source national forest inventory - small area estimation and map productionE. Tomppo

12.1 Introduction

12.1.1 Background

12.1.2 Progress in the Finnish multi-source inventory

12.2 Input data sets for the basic and improved k-NN methods

12.2.1 Processing of field data for multi-source calculations

12.2.2 Satellite images

12.2.3 Digital map data

12.2.4 Large-area forest resource data

12.3 Basic k-NN estimation

12.4 The improved k-NN, (ik-NN) method

12.4.1 Simplified sketch of the genetic algorithm

12.4.2 Application of the algorithm

12.4.3 Reductions of the bias and standard error of the estimates at the pixel level and regional level

12.5 Conclusions

References

13 Correcting map errors in forest inventory estimates for small areasM. Katila

13.1 Introduction

13.2 Land use class areas

13.3 Calibrated plot weights

References

14 Multiphase samplingS. Tuominen et al.

14.1 Introduction

14.2 Double sampling for stratification when estimating population parameters

14.3 Double sampling for regression

14.4 Forest inventory applications of two-phase sampling

14.4.1 Grouping method - two-phase sampling for stratification with one second-phase unit per stratum

14.4.2 Stratification with mean vector estimation

14.4.3 K nearest neighbor method with mean vector estimation

14.5 Multi-phase sampling with more than two phases

14.6 Estimation testing

14.7 Concluding remarks

References

15 SegmentationA. Pekkarinen and M. Holopainen

15.1 Introduction

15.2 Image segmentation

15.2.1 General

15.2.2 Image segmentation techniques

15.2.3 Segmentation software

15.3 Segmentation in forest inventories

15.4 Segmentation examples

15.4.1 General

15.4.2 Example material

15.4.3 Example 1: pixel-based segmentation

15.4.4 Example 2: edge detection

15.4.5 Example 3: region segmentation

References

16 Inventory by compartmentsJ. Koivuniemi and K.T. Korhonen

16.1 Basic concepts and background

16.2 History of the inventory method in Finland

16.3 Inventory by compartments today

16.3.1 The inventory method

16.3.2 Estimation methods

16.4 Accuracy in the inventory by compartments method and sources of error

References

17 Assessing the world's forestsA. Kangas

17.1 Global issues

17.1.1 Issues of interest

17.1.2 Forest area

17.1.3 Wood volume and woody biomass

17.1.4 Biodiversity and conservation

17.2 Methodology

17.2.1 Global forest resources assessment

17.2.2 Temperate and boreal forest assessment

17.2.3 Pan-tropical remote sensing survey

17.2.4 Global mapping

17.2.5 Forest information database

References

Part III Cases

18 EuropeT. Tokola

18.1 Sweden

18.1.1 Swedish national forest inventory

18.1.2 Inventory for forest management planning

18.2 Germany

18.2.1 National forest inventory: natural forests

18.2.2 Regional inventories

18.2.3 Forest management planning: compartment level inventory

18.3 Other European areas

References

19 AsiaT. Tokola

19.1 India

19.1.1 Forest cover mapping

19.1.2 Forest inventory

19.1.3 Trees outside the forest (TOF) and the household survey

19.1.4 Forest management planning

19.2 Indonesia

19.2.1 The national forest inventory

19.2.2 Concession renewal mapping

19.2.3 Forest management planning: compartment-level inventories ofnatural forests

19.2.4 Forest management planning: compartment-level inventories ofplantation forests

19.3 China

19.3.1 National forest inventory: natural forests

19.3.2 Forest management planning: compartment-level inventories

19.4 Other Asian areas

References

20 North AmericaT. Tokola

20.1 Canada

20.1.1 Provincial-level management inventories

20.1.2 National forest inventories, national aggregation

20.1.3 Industrial forest management inventories

20.2 The United States of America

20.2.1 The national forest inventory

20.2.2 Industrial forest management planning: stand-level inventory

20.2.3 Cruising, scaling and volume estimation

20.3 Mexico

References

Part IV Future

21 Modern data acquisitionfor forest inventoriesM. Holopainen and J. Kalliovirta

21.1 Introduction

21.2 Remote sensing

21.2.1 Digital aerial photos

21.2.2 Spectrometer imagery

21.2.3 High-resolution satellite imagery

21.2.4 Microwave radars

21.2.5 Profile imaging

21.2.6 Laser scanning

21.3 Use of modern remote sensing in forest inventories

21.3.1 Accuracy of remote sensing in forest inventories

21.3.2 Stand-, plot- and tree-level measurements on digital aerial photographs

21.3.3 Stand-, plot- and tree-level measurements using laser scanning

21.3.4 Integration of laser scanning and aerial imagery

21.4 Improving the quality of ground-truth data in remote sensing analysis

21.4.1 Development of field measuring devices

21.4.1.1 Terrestrial lasers

21.4.1.2 Laser-relascope

21.4.1.3 Digital cameras

21.4.2 Field data acquisition by logging machines

References

Available:*

On Order

Summary

Summary

Table of Contents