delusional redd baselines

32
Delusional REDD baselines Britaldo Silveira Soares Filho MEASUREMENT, REPORTING AND VERIFICATION IN LATIN AMERICAN REDD+ PROJECTS CIFOR WORKSHOP MARCH 8-9, 2012 PETRÓPOLIS, BRAZIL Britaldo Silveira Soares Filho

Upload: center-for-international-forestry-research-cifor

Post on 21-May-2015

230 views

Category:

Education


0 download

DESCRIPTION

Credible baseline setting and accurate and transparent Measurement, Reporting and Verification (MRV) of results are key conditions for the success of REDD (Reducing Emissions from Deforestation and forest Degradation). In this presentation, Britaldo Silveira Soares Filho from Universidade Federal de Minas Gerais discusses the challenges inherent in setting baselines and applying modelling for REDD. Britaldo Silveira Soares Filho gave this presentation on 8 March 2012 at a workshop organised by CIFOR, ‘Measurement, Reporting and Verification in Latin American REDD+ Projects’, held in Petropolis, Brazil. The workshop aimed to explore important advances, challenges, pitfalls, and innovations in REDD+ methods — thereby moving towards overcoming barriers to meeting MRV requirements at REDD+ project sites in two of the Amazon’s most important REDD+ candidate countries, Peru and Brazil. For further information about the workshop, please contact Shijo Joseph via s.joseph (at) cgiar.org

TRANSCRIPT

Page 1: Delusional REDD baselines

Delusional REDD baselines

Britaldo Silveira Soares Filho

MEASUREMENT, REPORTING AND VERIFICATION IN LATIN AMERICAN REDD+ PROJECTS

CIFOR WORKSHOP – MARCH 8-9, 2012 – PETRÓPOLIS, BRAZIL

Britaldo S

ilveira S

oares

Filho

Page 2: Delusional REDD baselines

Main concepts

• Additionality: Proof that reduction in emissions from

REDD is additional to reductions that would occur if

initiative were not in place.

• Leakage: reduction in carbon emissions in one area

that results in increased emissions in another

• Permanence: long-term reduced emissions from

REDD. Depends on vulnerability to deforestation and/or degradation.

Concepts and methods borrowed from CDM projects

Britaldo S

ilveira S

oares

Filho

Page 3: Delusional REDD baselines

Measuring performance: the baseline approach

The CDM baseline

Historical approach: introducing an innovative process to an installed industrial plant.

Forward-looking approach (ex ante): the building of a new factory with a carbon neutral approach.

Depend strongly on the project and less on the context (Lesser risk of hot air).

Britaldo S

ilveira S

oares

Filho

Page 4: Delusional REDD baselines

REDD baselines

Forward-looking approach: Brazil overall GHG target of 39% reduction by 2020 (emissions from agriculture can increase from 480 to 627 MCO2)

Historical approach: Brazil´s National Climate Change Plan.

Reduction

6238

REDD6560

3805

18226

25396

27772

19014

14196

Hitorical

12911

7464

19600

Baseline15405

8935

0

5000

10000

15000

20000

25000

30000km2

Britaldo S

ilveira S

oares

Filho

Page 5: Delusional REDD baselines

What to do with countries or regions with low deforestation rates and large extent of forests

Acre Madre de Dios Pando

For more information: visit www.csr.ufmg.br/map

Britaldo S

ilveira S

oares

Filho

Page 6: Delusional REDD baselines

baseline

with project

-

2,000,000

4,000,000

6,000,000

8,000,000

10,000,000

12,000,000 C stocks

time

Forward-looking baseline

Project starts

Additionality = $$ credits

Britaldo S

ilveira S

oares

Filho

Page 7: Delusional REDD baselines

Project starts

Baseline is not solely based on the project itself, but on a reference region. Therefore, there is a need to understand the effect of the project on the deforestation regional trend.

Forward-looking baseline

Britaldo S

ilveira S

oares

Filho

Page 8: Delusional REDD baselines

1. understanding historical trends

2. identifying drivers of deforestation

3. modeling future scenarios

Solution: Deforestation model??

BASIC STEPS OF MODELING

A series of what if decisions !!! Brita

ldo Silveira

Soare

s Filho

Page 9: Delusional REDD baselines

• Historical fixed rates

• Historical trend

• Business-as-usual simulation (future scenarios)

Understanding historical trends Modeling approaches for BAU baselines

0

20000

40000

60000

80000

100000

120000

140000

160000

ha

time

average

projected

simulated (BAU)

ObservedWhich one ?

Britaldo S

ilveira S

oares

Filho

Page 10: Delusional REDD baselines

Modeling deforestation trajectories

Soares-Filho et al. 2010. PNAS.

2010 2007

Britaldo S

ilveira S

oares

Filho

Page 11: Delusional REDD baselines

Modeling scenarios

Soares-Filho et al. 2006. Nature

Nesptad, Soares-Filho et al. 2009. Science.

The End of Deforestation in the Brazilian Amazon

Trajectories can change drastically

Britaldo S

ilveira S

oares

Filho

Page 12: Delusional REDD baselines

Other aspects

• Delimiting the geographic reference region.

Reference region

Project

Britaldo S

ilveira S

oares

Filho

Page 13: Delusional REDD baselines

Other aspects

• Delimiting the geographic reference region.

Reference region

Project

road

Britaldo S

ilveira S

oares

Filho

Page 14: Delusional REDD baselines

Other aspects

• What if there are other projects???

Reference region

Project Project

Project

Madre de Dios, Peru, have 12 or more REDD projects

Britaldo S

ilveira S

oares

Filho

Page 15: Delusional REDD baselines

identifying drivers of deforestation deforestation up to 1994

deforestation up to 1999

deforestation from 1994 to 1999

(gray tone)

distance to main roads

buffer zone

overlay

cross-tabulation

deforested non-deforested total

----------------------------------------

distance > 10 km | 194187 1005296 | 1199483

distance < 10 km | 84228 199349 | 283577

----------------------------------------

total | 278415 1204645 | 1483060

They are in fact spatial determinants

Most models only incorporate spatial determinants!!!

Probability map of deforestation

Britaldo S

ilveira S

oares

Filho

Page 16: Delusional REDD baselines

modeling future scenarios • Spatially-explicit model of deforestation to quantify

where deforestation is likely to occur.

-09 50O `

-10 00O `

-10 10O `

-10 20O

`

-10 30O

`

-55 00O ` -54 50

O `

-09 50O `

-10 00O `

-10 10O `

-10 20O `

-10 30O

`

-55 00O ` -54 50

O `

-09 50O

`

-10 00O

`

-10 10O

`

-10 20O `

-10 30O

`

-55 00O

` -54 50O

`

Land cover map of 1986 Land cover map of 1994 Simulated land cover map of 1994

project project

deforestation is not random but occurs at locations that have a combination of bio-geophysical and economic attributes that are more attractive to the agents of deforestation

True or false?

Britaldo S

ilveira S

oares

Filho

Page 17: Delusional REDD baselines

Stochastic nature of deforestation

Performance x realism

Absolute frequency of transition probabilities of observed deforestation

scene 23167

0

500

1000

1500

2000

2500

3000

3500

0 50 100 150 200 250

probability value

num

be

r of

cells

Dinamica EGO simulates landscape structure

Half true, half false

Increasing spatial performance decreases realism

Britaldo S

ilveira S

oares

Filho

Page 18: Delusional REDD baselines

-09 50O `

-10 00O `

-10 10O `

-10 20O

`

-10 30O

`

-55 00O ` -54 50

O `

-09 50O `

-10 00O `

-10 10O `

-10 20O `

-10 30O

`

-55 00O ` -54 50

O `

-09 50O

`

-10 00O

`

-10 10O

`

-10 20O `

-10 30O

`

-55 00O

` -54 50O

`

Land cover map of 1986 Land cover map of 1994 Simulated land cover map of 1994

project project

Model validation

t1 t2 t3

training Validation only regarding location, but not spatial structure

Britaldo S

ilveira S

oares

Filho

Page 19: Delusional REDD baselines

Model performance assessment • Map comparison methods (how to lie with validation methods)

Label 1 2 3 4 5 6 7

map \ author

Costanza

(1989)

Power et al.

(2001)

Pontius

(2002)

Hagen

(2003)

Van Vliet

et al.

(2011)

Almeida et

al. (2008)1

Almeida et

al. (2008)2

degraded 5x5 0.98 0.83 0.65 0.70 0.66 0.83 1.00

degraded 9x9 0.97 0.81 0.48 0.47 0.48 0.62 0.75

degraded 15x15 0.96 0.81 0.34 0.24 0.34 0.46 0.55

degraded 31x31 0.95 0.80 0.18 -0.03 0.18 0.26 0.32

degraded 51x51 0.95 0.80 0.12 -0.13 0.12 0.19 0.23

random 0.90 0.84 0.002 1.00 1.00 0.04 0.05

line1 x line2 1.00 0.81 -0.02 0.59 -0.02 1.00 1.00

Soares-Filho et al. in review

And model comparisons are quite subjective (e.g. Vega et al. 2011)

Britaldo S

ilveira S

oares

Filho

Page 20: Delusional REDD baselines

Different models have different assumptions, geographic scales, spatial resolutions,

scopes, and, therefore, purposes.

Both simulations developed using Dinamica EGO, but none of them were designed to fix baselines

Thank you Reimer

Britaldo S

ilveira S

oares

Filho

Page 21: Delusional REDD baselines

This means trouble

Britaldo S

ilveira S

oares

Filho

Page 22: Delusional REDD baselines

What about leakage?

Leakage cannot be simulated, but can be evaluated. See Soares-Filho et al. PNAS. 2010

leakage In-out, out-out, diffuse (Fearnside, 2009)

Is modeling for dummies??? Model wizard for rapid

assessment

Effectiveness = Success rate – Leakage 50% = 70% - 20%

Representation of a fictitious software

Leakage Success Effectiveness

Enter value ------------ ------------

Britaldo S

ilveira S

oares

Filho

Page 23: Delusional REDD baselines

Good job &Business

• But!!!!

No ready solution for REDD (although vendors say so)

There is no magic solution, via software wizard…

MODEL must be built from the ground to incorporate local

knowledge…

Britaldo S

ilveira S

oares

Filho

Page 24: Delusional REDD baselines

In sum • Although various MRV methodologies for mapping carbon stocks and modeling

reductions as well as standards for international certification have been developed, the criteria for establishing baselines for crediting purposes are unsound; they do not take into account that forward-looking baselines are questionable because deforestation trajectories can alter drastically in response to changes in a complex set of circumstances (Nunes et al. 2012; Soares-Filho et al. IADB report, 2012).

• Furthermore, it is very difficult to isolate the local effect of a project from the overall trajectory of deforestation as well as to assess leakage arising from the establishment of projects, and none of those projects have attempted to perform such analyses (Soares-Filho et al. IADB report, 2012).

• As more specific standards and indicators are established, more flawed they become, because they are unfounded anyway (Personal Perception). And for financial crediting, every dollar counts. In addition, conflict of interest might lead to gamming (Fearnside, 2012).

Britaldo S

ilveira S

oares

Filho

Page 25: Delusional REDD baselines

So, how can we apply modeling to REDD?

Intervention

Information

Modeling Knowledge

Tool for policy design and planning

(let’s keep us in business)

Britaldo S

ilveira S

oares

Filho

Page 26: Delusional REDD baselines

Tool for REDD planning and management. Application to Acre

Dinamica EGO software

Simulation as a tool for regional planning www.csr.ufmg.br/map

Method adopted by SISA

Britaldo S

ilveira S

oares

Filho

Page 27: Delusional REDD baselines

Setting priority areas for REDD

• Index of threat (time dependent) (Soares-Filho et al. PNAS, 2010)

Priorização de áreas protegidas

Britaldo S

ilveira S

oares

Filho

Page 28: Delusional REDD baselines

Calculating costs of reduction

Amazon opportunity costs

Carbon Cost

0

5

10

15

20

25

30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

year

U$

Carbon cost along the deforestation frontier

year (Nespdad, Soares-Filho et al. 2009; Soares-Filho et al. 2010)

Britaldo S

ilveira S

oares

Filho

Page 29: Delusional REDD baselines

Assessing effectiveness and leakage over time

Only monitoring deforestation is not enough.

Soares-Filho et al. PNAS . 2010.

Odds ratio adjusted

Land use zones 1997 2000 2005 Mean

Urban and periurban areas 6.15 2.17 5.40 4.57

Mining concessions 1.26 1.62 1.51 1.46

Small-scale cattle ranching 1.42 1.20 1.28 1.30

Small landowner lots 1.02 0.99 1.55 1.19

Brazil nut concessions 0.76 0.67 0.49 0.64

Logging concessions 2.84 3.28 4.29 3.47

Native communities 0.77 1.20 1.35 1.11

Reforestation concessions 0.19 0.39 0.93 0.51

Conservation concessions 3.85 2.78 2.06 2.89

Ecotourism concessions 0.08 0.08 0.76 0.31

Protected areas 4.85 8.27 5.55 6.22

Indigenous reserves 0.20 0.02 x 0.11

Unassigned land use 0.96 0.94 1.17 1.02

Nunes, Soares-Filho et al. Env. Con. 2012.

Britaldo S

ilveira S

oares

Filho

Page 30: Delusional REDD baselines

Which REDD projects in Madre de Dios can make a difference?

Hajek et al. 2011

Nunes, Soares-Filho et al. Env. Con. 2012.

Britaldo S

ilveira S

oares

Filho

Page 31: Delusional REDD baselines

Channeling REDD+ investments

Nunes, Soares-Filho et al. Economic benefits of forest conservation: assessing the potential rents from Brazil nut concessions in Madre de Dios, Peru, to channel REDD+ investments. Environmental Conservation, 2012.

REDD funds invested to upgrade the production chain of non-timber forest products

Britaldo S

ilveira S

oares

Filho

Page 32: Delusional REDD baselines

Thank you/Obrigado/Gracias

Modeling in support of sound policy Dinamica EGO freeware can be downloaded at www.csr.ufmg.br/dinamica

For many more examples, please access http://www.csr.ufmg.br/dinamica/publications/publications.php

[email protected]

Britaldo Silveira Soares Filho (the culprit for developing deforestation models that are now being applied to REDD+)

Britaldo S

ilveira S

oares

Filho