Comments on Monitoring Methodologies:TanzaniaMoldova

Brazil

Winrock International

Training Seminar for BioCarbon Fund Projects

Monitoring Methodology for: TIST (Tanzania)

Purpose of Methodology

Designed for measurement of afforestation/reforestation activities where the project boundary delineates many discrete areas

Comment on Scope

The TIST project promotes development by giving an additional income stream to villagers• plus the incentive, instruction and

encouragement to improve practices Tree plantings are on the land of

individual villagers who are project participants

Monitors have limited education and are non-expert in forestry

Comment on Scope

Methodology should be limited to tropics

And to situations where the very limited income from carbon finance is sufficient to make a difference to individual small landowners

In more developed areas this methodology would not be the most efficient

Extraneous Inclusions

Criticism from A/R working group and expert reviewers for non-focused methodology• Inclusion of eligibility requirements• Inclusion of positive leakage (currently not

allowed under CDM)• Application to temperate scenarios• Application to scenarios where

measurements are of volume instead of DBH

Monitoring Baseline

Claim is of a zero carbon baseline• Argument of continued deforestation

in country Baseline is within project

boundaries So carbon in agricultural crops

and fallow vegetation should be measured

Measurement Methods – core methods

Two options for measuring project boundary• # of trees x spacing of trees

• Subject to HUGE error especially as spacing is only measured in one area of planting

• GPS tracking of boundary• Should be chosen as sole method with

methods to track error on GPS such as comparison of repeated measurements, comparison with topographic map

Non-expert monitor:• Counts number of trees• Measures circumference of 20 trees per planting,

systematically selected But

• All trees counted including trees below minimum DBH for measurement

• Factor applied to estimate this proportion but assumes linear relationship between proportion of trees non-eligible and proportion of biomass

• Mean DBH calculated and allometric equation applied to this

• Correct methodology is to calculate the biomass for each measured tree and average that – IS NOT A LINEAR RELATIONSHIP BETWEEN DBH AND BIOMASS

Measurement Methods – core methods

TIST-specific Issues

TIST unusual in: The large stake that landowners have in

a relatively few trees • Make it impossible to mark trees

• Will be treated differently by villagers

Many non-expert monitors produce data• Can’t select random trees at each planting at

each measurement to be measured• A more technical approach would make it more

difficult for non-experts to implement and more subject to errors

Reviewer Gonzalez simply wrong to state that trees have to be marked to monitor growth and to verify measurements

TIST should include examples to show that the level of effort is sufficient to record to within 10 % of the mean with 95 % confidence

TIST-specific Issues

Extraneous Methods

Calculation of biomass from biomass of just bole/stem• This method would be used by

foresters, it is unrealistic to assume that a forestry operation would fit under the TIST scenario

• Creates more complication and opens methodology up for more criticism

Allometric Equation

We suggested that TIST harvest trees where possible to create a region specific equation

As an alternative we provided an equation• TIST incorrectly listed the equation in the

methodology

DBH Correct Equation Equation Submitted to CDM10 42.5 1359109875.220 212.0 2718219750.430 543.1 4077329625.640 1058.5 5436439500.950 1776.2 6795549376.160 2711.2 8154659251.370 3876.6 9513769126.5

If TIST wants to use a generic equation it is necessary to verify its applicability

This can be done by harvesting a limited number of trees across the expected size range

Allometric Equation

Belowground Biomass

Criticism from reviewers for using single ratio factors for calculation of belowground biomass

Equations exist (Cairns et al 1997) that co-vary belowground biomass with aboveground biomass. • One equation is specifically for the tropics• Must not be used on a tree by tree basis but

on a per hectare carbon density

Summing through Time

TIST approach creates complications for calculating uncertainty

At each measurement a mean and a 95 % confidence interval

To calculate confidence interval on the increment will require Monte Carlo simulation • Time 1 and Time 2 are entirely dependent

Could use principle of conservatism instead

Leakage

Project proposes that there will be leakage associated with travel of participants-this is part of project activities and needs to be monitored (not leakage)

However no indication of measurements of potential for:• Increased income leading to more livestock and

more fertilizer• Displaced farmland – scrub/shrubs/trees cut to

replace farmland lost to afforestation These possibilities must be tracked or a

strong case made that they are not likely

Uncertainty

QA/QC plan exists includes blind resampling of 5 % of plantings• Difference between measurements

could be expressed as a % measurement error

95 % confidence intervals should be reported for all measurements

Conclusions

We would advise • Tightening of methods to include just the

specific conditions under which it will be applied

• Removal of extraneous details• Tightening of methods and illustration that

sampling level is enough to produce very high precision

It is ‘good practice’ and it greatly improves the chances of success if the IPCC Good Practice Guidance is repeatedly referred to

Monitoring Methodology for:Restoration of Degraded Lands through

Afforestation/Reforestation(Moldova Project)

Purpose of Methodology

Designed for monitoring of afforestation/reforestation activities:• Project boundary delineates many

discrete areas• Degraded lands – Baseline assumed to

be zero

Conditions of methodology

Land becoming more degraded over time – e.g. carbon stocks declining over time

Degradation permits only use of less intense land use to restore productivity

Project entities sufficiently organized

Overall Strengths and Weaknesses:

Can only be applied to areas of zero or negative baseline

Methodology is simple, easy to follow, relatively well organized

Brings up issue: Should NMM include step-by step methods, or just outline?

All leakage issues not addressed Suggested additions are in orange

Section B. Proposed new methodology:

Overall, good methodology The interaction between field

measurements and CO2FIX model unclear• Need to state that model is for projections only

Methodology steps:• Project area delineation• Field survey• Stratification• Sampling design• Selection of plot size and number• Monitoring and measuring• Quality assurance

Section B.2.2. Formulas to monitor actual net GHG removals by sinks:

PRCO2e = ∑ [{CPR * (44/12)}*An – PECO2e]

Where:

PRCO2e = CO2e GHG removals in the project scenario

PECO2e = CO2e Carbon emitted by project activities for total area44/12 = Factor to convert C to CO2e

An = Number of hectares of area

1…, n = Number of strata 1…, n

Section B.2.1 Actual net GHG removals by sinks data:

1. Monitoring frequency2. Project Boundary

• Discuss how to determine project boundaries• Can be complicated due to large number of

small project areas

3. Stratification• Aim of creating strata• Example of strata types

4. Sampling Design• Determination of plot locations

e.g. systematic, random, stratified random• Creation of permanent vs. temporary sample

plots

Section B.2.1 Actual net GHG removals by sinks data:

5. Plot number• Determination of number of plots required

– via collection of preliminary data• Discussion of accuracy + precision

required• Discuss variation of different pools,

degree of precision desired for each poole.g. size of variation vs. size of pool

6. Sample Frame, Nesting plot sizes• Remove detail on techniques, place later in

B.2.2.1. with other detail7. Data Collection

• Creation of Standard Operating Procedures (SOPs)

Section B.2.2.1 Description of formula + models used to monitor:

Straight forward methods Standard practices - IPCC Good Practice

Guidance Need to eliminate project specific

information Lacks

• Some detailed field and analysis methods• Uncertainty Analysis• Analysis at 2nd sampling time period• Expansion factors• Discussion of fire, disease, extreme events

Section B.2.2.1 Description of formula + models used to monitor:

1. Preliminary data to calculate plot number

2. Determination of nested plot sizes• ’10 stems’ is a good ‘rule of thumb’ to help

determine size of nested plots for trees.• Can use rectangular or circular plots

3. Creation of plots• Discuss methods used to create permanent plot (if

used)4. Tree

• Creation of allometric equations• Thinned trees - not best guidance

• Trees will all be small size class• Remaining trees may change growth pattern

• More explicit instructions on creation or verification of existing equations

Section B.2.2.1 Description of formula + models used to monitor:

5. Non-tree• More explicit on methodologies for measurement

6. Standing dead• Dead wood density classes – no description of

how collect

7. Lying dead wood• Mean density – no description of how to collect

8. Litter• Conflicting remarks – send to lab or assume

50%?

9. Soil10. Data Analysis and Uncertainty Analysis

Section B.2.2.2 Formula – Project Emissions

PE CO 2e = A * D * ED * + A * G * EG * 0.001 (11)

WhereA = Project area – units: haD = Quantity of diesel – units: L/haED = Emission factor for diesel (2.63 kg CO2/l)G = Quantity of gasoline - units: L/ha EG = Emission factor for gasoline (2.40 kg CO2/l)0.001 = Factor to convert kg to Mg

Definition of ‘project boundary’ used in strict sense.

Suggestion: put all vehicle emissions in this section

Section B.3. Leakage

1. Diversion of pre-existing A/R activities

2. Shifting of activities3. Market affects

Section B.4. Description of formula to estimate net GHG removals

Need to subract leakage from equation:

NR(t) CO2e = PR(t) CO2e - BR(t) CO2e - PLKCO2e

assumes BR(t) CO2e = 0 Where:

NR(t)CO2e = Net anthropogenic GHG removals in the monitoring period t(CO2e)

PR(t) CO2e = GHG removals in the project scenario in the monitoring period t (CO2e)

BR(t) CO2e = GHG removals in the baseline scenario in the monitoring period t (CO2e)

t = Indicator for the monitoring period

PLKCO2e = Leakage

Conclusions - Moldova

Overall good methodology Only minor improvements

need to made:• Expand field and analysis

methodology• Expand leakage monitoring

Discussion point: What level of detail should NMM contain?

Monitoring Methodology for:Reforestation Project Using Native

Species Around AES-Tiete Reservoirs(Brazil Project)

Status of Brazil

Rejected in first round and is presently under major revision

How many monitoring methodologies can there be? Class by sampling design (3)

• Single tree as in TIST• Systematic versus stratified random

•Readily stratified into homogeneous unitsE.g. plantations of native species on flat and sloped lands

•Highly heterogeneous system•Post stratifyE.g. planting in islands and natural regeneration

How many monitoring methodologies can there be? Class by type of sampling plots

• Systematic versus stratified random

•Permanent sample plots•Temporary sample plotsTOTAL NUMBER:

FOUR BASIC MONITORING PLANS

•Permanent sample plots

How many monitoring methodologies can there be?

Other pools can be included or not—make optional with some guidance• In most cases dead lying wood will be

very small pool and maybe not “worth” the effort (conservative is not monitored and counted)

Leakage monitoring can add some wrinkles

Possible sources of leakage-likely to be more project specific• Activity shifting—clearing land elsewhere

• Very difficult to assume any deforestation in region is due to project

• Depends on land tenure/ownership• Try to handle in project design by providing other sources

of livelihoods• Could monitor through rate of adoption of alternatives

• Market effects• Identify possible sources of this effect • Try to make strong case for limited to no leakage in PDD

so no need to monitor Common mistake in many PDDs —include GHG

emissions due to project as leakage because assumed to be outside “project boundary”—this is incorrect

Improve monitoring methodologies KISS principle and be generic—details

come in the application in the PDD Give step by step methods, following

those in the IPCC GPG Ch 4.3 but without great details• E.g. Step x. create a plan to locate plots

• Enough detail for a generic methodology?

Check all equations reported in document with data to make sure get reasonable answers—many NMM have sloppy mistakes and wrong conversion factors

Improve monitoring methodologies Don’t undermine the QA/QC plan—

develop SOPs and include in Appendices Work together and learn from each

other and pool expertise—this is not a competition

If English is not preparers first language, get someone with English to edit it—must communicate methodology succinctly