Global Forest Resources Assessment 2005

Professor Rod Keenan

The University of Melbourne

What are forests?

• Historically: an area of land reserved for the king for hunting

• Administrative/legal definitions • Production definitions• A biological community (ecosystem) of

plants and animals that is dominated by trees and other woody plants

Why do forests matter?

Livelihoods• 1.6 billion people

rely heavily on forests for their livelihood

• Woodfuel• Protein from

hunting and fishing• Medicines

Commodities and economic benefit

• Wood trade US$150B, 2% world trade

• Forest based enterprises provide up to 1/3 of non-farm employment

• Non-wood commodities also significant

Environmental services

• Clean water• Soil and erosion

protection• Carbon sequestration

Biodiversity

• Tropical forests cover 15% of land surface but over 50% of terrestrial species

Human values

• Aesthetic• Spiritual • Cultural• Recreational• Human health

FAO

• Rome based, regional offices, commissions

• Agriculture, forest production focus

• Direct technical support• ‘Normative’ functions • Information and statistics• Slow to react to changing

circumstances

The FRA 2005 Process229 countries and territories

10 regional review workshops

More than 40 variables on forest extent, condition, uses and values

229 detailed reports

3 points in time: 1990, 2000 and 2005

More than 800 specialists involved

Why undertake assessments?

Support implementation of sustainable forest management through the provision of sound information to inform policy and

management decisions

Total forest area

4 billion hectares of forest (40 million km2)

0.62 hectares per capita

30 percent of total land area

1.4 billion hectares of other wooded land

The world’s forestsThe Russian Federation, Brazil, Canada, USA and China together account for half the total

forest area

64 “low forest cover countries”

Forest change dynamics

FORESTReforestation

Natural regeneration

OTHERLAND

AfforestationNaturalexpansion

DeforestationNaturaldisasters

Deforestation and net forest loss

Net forest loss 2000-2005: 7.3 million ha/yr or 200 km2 per day

Deforestation: 13 million ha/yrNet forest loss 1990-2000: 8.9 million ha/yr

Mill

ion

hect

ares

Annual net change in forest area

Primary forests

36 percent of the world’s forests

Modified natural forests

53 percent of the world’s forests

Semi-natural forests

7 percent of the world’s forests

Forest plantation

4 percent of the world’s forests

Trends

Decrease:

6 million ha/yr

Increase:

2.8 million ha/yr

Forest health and vitality

Forests and climate change

Carbon in forest biomass: 283 GtAnnual change: -1.1 GtTotal carbon: More than in the atmosphere

Management,conservationanduse

Protection of soil and water

Production34%

9%

11%4%

34%

8%

Conservation of biodiversity

Recreation and education

Multiple use

Unknown function

Forest productsWood removals:

– 3 billion m3/yr– US$ 64 billion

NWFPs: Important, butdifficult to quantify

Industrial roundwood: 60%

Woodfuel: 40% Non-wood forest products

FRA 2005 is much more...OwnershipNumber of tree species

Growing stock composition

Vulnerable, endangered and critically endangered tree species

Growing stock and biomass

Employment

Synthesis Assessment

Science and policy: some fundamental differences

Scientists• Seeking truth• Long-term view• Open to many viewpoints• Focus on state of

environment or the ecosystem

• Ration and logic

Policy makers• Seeking a solution• Short-term view• Once policy is decided,

alternative viewpoints not encouraged or considered helpful

• Focus on public attitudes, perceptions and emotions

Telling the FRA story

• Headline variables• Composite indices

Headline indicators

• forest area• other wooded land area• Area of primary forest• area with designated primary function of biodiversity

conservation• productive plantation area• total growing stock commercial timber• total carbon stock• total volume and value of wood removals• total number of people employed in forests

Composite indices

• Economic indicators– GDP– Inflation rate– Consumer Price Index

Composite indices: desirable properties

• Clear message, easily understood and interpreted

• Strong scientific or theoretical basis• Policy relevance• Efficiently represent different elements

being integrated• Temporal relevance (capable of being

backdated, suitable for long-term measurement)

Composite indices

• environmental sustainability index (ESI)• ecological footprint• consumption pressure index• living planet index• green GDP

Purpose?

• Compare between countries• Trends across time• Regional analysis

Technical issues

• Variable selection• Scaling, ranking and standardising• Missing data• Aggregation• Weighting

Variable selection

Regional data availability

• High: 75–100 % of total forest area• Medium: 50–74 %• Low: 25–49 %• Less than 25 %: no data

Threshold of change

Global analysis, negatives

Global analysis, positives

Africa

Asia

Europe

North and Central America

South America

Oceania

Issues

• FRA variables do not address all aspects of sustainable forest management

• Low quality, out of date or missing data• Subjective variable selection that may not

relevant for all countries or at different scales• Value judgement on ‘positive’ or ‘negative’

trends – Forest area, wood removals, employment, area

conserved for biodiversity• Trends by region obscures positive or negative

trends in individual countries

Conclusions

• Most comprehensive assessment to date• 30% of the world’s land area is forest• Deforestation continues but net loss slowing

down • Forests are increasingly being conserved

and managed for multiple uses and values• New, comprehensive, unbiased information

for policy makers

Acknowledgements

• Country correspondents• FAO FRA team• Orjan Jonsson• Peter Holmgren• Mette Lloyche Wilkie• Lars Gunnar Marklund• Kailash Govil

What’s happening in Australia’s forests

Science for decision makers

Adam Gerrand

Manager, National Forest Inventory

How do we tell…

How we can improve our national forest monitoring

2www.brs.gov.au

Outline• Australia’s place in the worlds forests

• Why do we need to monitor our forests?

• Current forest data collection and reporting

• What would a national forest monitoring framework look like?

• Results from a pilot project in NE Victoria

• Conclusions ….possible next steps

3www.brs.gov.au

Why do we need to do it - Australia’s place in the worlds forests – the good news: mega biodiversity, unique flora & fauna

• Australia is 6th most forested country in the world (FAO)– 1. Russian Federation, 2. Brazil, 3. Canada,

4. U.S., 5. China, 6. Australia (4%)

• Australia has 164 million ha of forests• Australia is one of the world’s biodiversity “hotspots”

– In the top 17 countries with at least 17,000 recognised plant spp. and >4,000 known animal species classed as forest dependent

• We have domestic and international responsibilities to monitor and manage our forests and vegetation

4www.brs.gov.au

The bad news: deforestation is high= biodiversity loss, carbon emissions, etc

• 1788 est Australia = 33% forest, now down to 21%• Australia's annual deforestation was 500,000 ha/yr (-0.2%)

• Previously one of the highest in the world (top 10, now 15th)

• The good news is…..deforestation has halved from over 500,000 to 250,000ha/yr

• Government policies and action to halt land clearing are working

5www.brs.gov.au

Some justifications for improved forest monitoring• Our ability to operate/export (for many that means stay in business)

– Depends on the acceptance and social licence to operate– Public procurement policies are becoming more restrictive (EU)– Certification and monitoring of operations to demonstrate

sustainability is increasingly expected • Need to improve our data collection to fill data gaps & report on

TRENDS required to demonstrate sustainability• Applies to most countries…..but Australia is slipping behind

– Most developed countries already have sophisticated monitoring– Many developing countries are applying new techniques to

improve monitoring of their forests

6www.brs.gov.au

Demand for forest information is increasing

• Inform public debate• Inform policy development• Inform industry development• Evaluate management programs• Meet national & international reporting obligations

7www.brs.gov.au

What should we measure?The old paradigm was fairly simple:

Manage and monitor productive forests

ProductsSpecies

Productivity

Use

TIMBER

Volume

8www.brs.gov.au

Changing society values => forest and land management becomes more complex

Wildlife

WaterRecre-ation

Timber

AestheticsGrazing

Urban

SoilsManage

and monitorforest and landscape

Ecosystems for a range of values

New or future values (e.g. carbon)

Australia’s forests Australia has 164 million ha of forest

Brisbane

Sydney

Melbourne

Perth

Darwin

Adelaide

Hobart

CANBERRA

Tasmania has 3 million ha of forest

Most of Australia’s forests are low, sparse and remote

Not much is tall, dense and available for timber

12www.brs.gov.au

Australia’s forest reporting process (Howell et al.2005)

• 2003 State of the Forest Report (SOFR)

• Based on Montreal Process Criteria and Indicators

• First time all 74 national indicatorsreported against

• Some trends could be established for anumber of indicators based onAustralia’s 1998 SOFR

13www.brs.gov.au

Capacity for reporting

• Data not comprehensive for all indicators

• Of the 74 indicators in the 2003 SOFR

! 16 indicators - Good spatial and temporal data- Forest area and type

! 56 indicators - Limited coverage or data from case studies

! 2 indicators - No data - (eg. National soil erosion hazard)

14www.brs.gov.au

Availability and completeness of Australian forest data• Each indicated assessed eg.• Good data coverage,

currency and frequency for forest area

• Partial coverage and limited data for growth stage

• Soil erosion hazard– Poor coverage, currency

and frequency

15www.brs.gov.au

Australia

• SFM themes as reporting framework • BRS / NFI prepared Australia’s FRA report

– based on data from all States and Territories

– & Australia’s State of the Forest report

FAO Global FRA success

16www.brs.gov.au

Australia’s ability to complete FRA tables

xx1000 person-yearsEmployment in forestryT15

Xxx1000 USDValue of non wood forest products removalT14

xxxMass/unit/volNon wood forest products removalT13

Xxx1000 USDValue of wood removalT12

xxx1000 m3Wood removalT11

XxxMill. m3Growing stock compositionT10

xNumberDiversity of tree speciesT9

xx1000 haDisturbances affecting health & vitalityT8

XxxMill. tonnesCarbon stockT7

XxxMill. tonnesBiomass stockT6

XxxMill. m3Growing stockT5

xxx1000 haCharacteristics of Forest and OWLT4

xxx1000 haDesignated functions of Forest and OWLT3

xx1000 haOwnership of Forest and OWLT2

xxx1000 haExtent of Forest and OWLT1

200520001990UnitTitleTable

How we are working to improve our data:The current framework of data collection

• Compilation and mapping approach• Data from State agencies (variety of

sources and standards)– State and Territory inventories– Air photos, remote sensing etc

• OK for snapshots, BUT…

• Varying sources and methods= inconsistent

• Different scales and quality of data• Collected at different times

= CANT DO TRENDS!

a “patchwork quilt”

18www.brs.gov.au

A way forward: the proposed Continental Forest Monitoring Framework (CFMF)

• National consistency and trends through permanent plots• A grid framework rather than sampling according to one value• Core set of scientifically reliable data

– on baseline and trends in forest extent and condition– from repeated, consistent measures

• Accurate change estimates at 5 year intervals• 3 Tier structure integrates field data and remote sensing

technologies to build on relative strengths of each data source

1b

1a

2

3

Tier

Forest Extent Mapping – continental scale

Continental grid for national and regional reporting

still essential but linked to RS mapping for efficiency and change detection

+newExistingField Plots

High resolution remote sensing – regional to local

Lower resolution satellite imagery

CFMF Multi-tier approach

Tier 1a: Continental Forest Mapping• National consistency - baseline and trends • Cover the whole country• Periodic updates (potential every 6 to 24 months)• 25m – 12.5m resolution• Attributes measured

– Broad cover and forest extent - AGO land cover change (1972 – 2004) – Forest cover change

-Large scale disturbance

Canberra

2002 LANDSAT 2003

• 5 km by 5 km image plots

• 1m – 10m resolution data

• SPOT 5, lidar, air photos & Ikonos

• Repeat samples e.g. 5 years

• Archived library

• Attributes measured– Forest extent

– Forest type, cover, height

– Land use and change

– Forest health and disturbance

– Extent and condition rare forest types and linear features (eg riparian vegetation)

Landsat TM wall to wall coverage

5km

5km

Tier 2: Hi-resolution plot

SPOT 5 sample tile

22www.brs.gov.au

• Ground plots essential• Connected to remote sensing data

to enable scaling up– Species, height, cover, growth

stage by strata (upper, mid, lower)– Site, disturbance, health– Habitat values & woody debris– Tree map

Tier 3: Ground plots

30m

30m

90m

90m

SPOT 5

23www.brs.gov.au

Conclusions from NE Victorian CFMF pilot

• CFMF multi-tier connected layer design is robust– takes advantages of different technologies– provides high accuracy but at high cost – Cost estimated $14m over first 5 years, then $2m/yr ongoing

• Assess modifying original design– To reduce costs – investigate modelling and better use of existing data– Apply modular or staged approach

- eg vary grid spacing to 100km not 20km

• Investigate CFMF application across the landscape to incl. othervegetation and land uses (NRM regional clients to share costs)

24www.brs.gov.au

Need to link data and share knowledge better

• Need to link data and share knowledge with other systems & disciplines– For efficiency of cost and effort, reduce duplication– For consistency of results (less time debating data, more on decisions)– For improved capacity and sharing techniques

• Globally– FAO FRA www.fao.fra– Global scientific community (e.g. GEO / GEOS) – NGO’s such as WWF, Greenpeace– Even communities and individuals www.confluence.org

• Nationally - Other government Departments and agencies- AGO, Geoscience Australia, CSIRO, Universities, DEST (TERN)

• Regionally - NRM and catchment management agencies, CMA’s etc

• Area covered at sample site: 10 km x 10 km

➾ ~ 13 000 sampling sites (excluding poles and deserts)

➾ Sampling intensity: about 1 % ( 1.2 % of land surface)

• Area covered at sample site: 10 km x 10 km

➾ ~ 13 000 sampling sites (excluding poles and deserts)

➾ Sampling intensity: about 1 % ( 1.2 % of land surface)

Proposed FRA 2010 remote sensing Global Information frameworkConnecting the national data into the global

26www.brs.gov.au

www.confluence.org

• a bunch of amateurs? • Sample grid at 1 degree

intersections (approx 730 points for Australia)

the rest of the world really is watching• If you go into the woods (web) today, you’re in for a big

surprise….

Hands up if you have used Google Earth?Future data delivery?making data easily accessible

29www.brs.gov.au

Terrestrial Ecosystem Research Network (TERN)• DEST call for major national infrastructure projects (NCRIS)

– Nuclear research facility, nano-technology, etc– Also a Terrestrial Ecosystem Research Network (TERN)

• Up to $50 million to build infrastructure to collate, store and distribute data– better, more open access

• Connect remote sensing data to field data – RS frequent, consistent good spatial coverage and relatively cheap

• Connect detailed research LTER sites with broader landscape monitoring • Build cross-disciplinary research networks • Use the huge potential of information collected to build knowledge on the

sustainability of our forest and other ecosystems

30www.brs.gov.au

Next steps…?• The rest of the world is watching….• Recognise responsibilities and importance of forest monitoring• Build broad support and strong commitment• Refine options for design of an agreed integrated continental

forest and vegetation monitoring framework• Draw from and feed back into:

– Montreal Criteria and Indicator reporting for 2008 SOFR– FAO-FRA2010, esp. remote sensing component– Other processes eg. Global Earth Observation (GEO)

• Build into TERN design proposal and get funding support from key stakeholders, policy and decision-makers and just do it!

31www.brs.gov.au

Improving information on Australia’s forests

• Australia’s State of the Forests Report (400 pages)

• Short forest profile series (Eucalypts, Acacia…etc.)

• Forests at a glance (pocket size)

• Plantation reports

• Available free on internet:

– www.brs.gov.au/nfi

– www.brs.gov.au/plantations– www.brs.gov.au/stateoftheforests– www.daff.gov.au