How to Make Mapping and Monitoring Zero-Deforestation Commitments Effective

Image: rogersmithpixCC BY-NC-ND 2.0
  • Experts have identified 12 attributes of effective zero deforestation commitment, or ZDC, mapping and monitoring systems.
  • The attribute framework can be used to guide the development of effective ZDC mapping and monitoring systems.
  • Having such a framework is important as there’s not a single one-size-fits-all ZDC mapping and monitoring system that will meet the needs of all users.

JAKARTA — Experts say that since there’s no one-size-fits-all solution to achieve effective zero deforestation commitment (ZDC) mapping and monitoring, they have created an attribute framework to guide the development of effective ZDC mapping and monitoring systems.

Effective mapping and monitoring systems are deemed integral to help companies around the world in eliminating deforestation from their agricultural production systems and global commodity supply chains, such as oil palm, beef, and soy.

To identify the attributes of effective systems, 17 experts from academia, the private sector and environmental NGOs, gathered in 2019 for a two-day workshop supported by the Science for Nature and People Partnership (SNAPP). SNAPP is a partnership of The Nature Conservancy and the Wildlife Conservation Society.

The experts identified 12 attributes of ZDC mapping and monitoring systems that contribute to their credibility, salience, legitimacy, and scalability — the four criteria that enable stakeholders to make better-informed decisions and act on their commitments to protect forests.

The attributes can be used to evaluate the strengths and weaknesses of existing or potential mapping and monitoring systems, and to identify gaps that could be filled by integrating multiple approaches.

“The goal of this effort was not to identify a single best ZDC mapping and monitoring approach, but to help supply chain actors and multi-stakeholder ZDC initiatives design mapping and monitoring systems that will meet user needs and contribute to reducing commodity-driver deforestation,” Kemen G. Austin, a senior policy analyst at NGO RTI and one of the report authors, told Mongabay by email.

The attribute framework is explained in detailed in a paper published in journal BioScience.

Cerrado clearing for agriculture. Image by Rhett A. Butler/Mongabay.


There are three attributes contributing to a credible mapping and monitoring system — technical rigor, consistency, and accuracy.

The technical rigor of a ZDC mapping and monitoring system represents the degree to which the system integrates best scientific practices into its protocols.

Consistent ZDC mapping or monitoring systems use the same — or comparable — methods and assumptions across space and over time, a necessary condition for identifying differences and trends.

Accurate ZDC mapping and monitoring systems correctly identify forest cover characteristics and deforestation occurrence at or exceeding a minimum level of confidence specified by the ZDC initiative in question.

Amazon rainforest canopy. Photo credit: Rhett A. Butler / mongabay
Amazon rainforest canopy. Photo credit: Rhett A. Butler / mongabay


Besides being credible, mapping and monitoring systems also have to be salient, meaning that their outputs are relevant to their users and address their needs.

The three attributes associated with saliency are geographic scope, monitoring frequency, and land cover categorical detail, according to the experts.

In terms of geographic scope, it is fundamental for mapping and monitoring systems to include the geographic areas where companies with ZDCs that use the systems source their deforestation-risk products.

As for monitoring frequency, it will depend on the need of the users. If the users want to be more responsive to problems and deforestation on the ground, then they’re likely need more frequent monitoring.

“Early detection on the order of days or weeks might be a priority for a company that wants to be able to act quickly if unexpected clearing is detected within their property,” the report states. “On the other hand, a system that provides less frequent but more accurate outputs may be preferable when identifying properties or producers that have not complied with ZDC criteria.”

Categorical detail refers to the ability of mapping and monitoring systems to distinguish land cover and land cover change relevant to and matching the definition of deforestation used by a given individual or collective ZDC.

Different mapping and monitoring systems may use different land cover categories and definition of deforestation, depending on their needs.

For instance, the definition of deforestation and land cover category used by federal government mapping and monitoring system, are different from the ones used by the High Carbon Stock Approach (HCSA).

PRODES maps only primary forest, and loss within primary forest, in the Legal Amazon (which includes all of the Brazilian portion of the Amazon biome). Once an area has been deforested, PRODES will no longer map loss within that area, even if secondary forest regrowth has occurred.

As a result, PRODES can’t be used to track compliance with other ZDCs that use a more liberal definition of forest.

On the other hand, HCSA defines several forest classes including high-, medium-, and low-density forests and young regenerating forests.

Terraces are cleared on a hillside in Malaysian Borneo to make way for an oil palm plantation. Image by Rhett Butler/Mongabay


ZDC experts define legitimate mapping and monitoring systems as ones that are fair and unbiased according to societal or ethical standards.

The three attributes contributing to a legitimate mapping and monitoring system are transparency, independence, and inclusivity.

Therefore, a legitimate mapping and monitoring system needs to be transparent, where users can access information on ZDC safeguarded areas and forest loss within these areas.

Its implementation also has to be independent from the influence of individual companies and commodity producers using the system..

Lastly, it has to be inclusive, which means it’s designed via the participation of all potential users and affected stakeholders, including those who have been traditionally underrepresented.

New oil palm planting near a protected area in Indonesia. Image by Rhett A. Butler/Mongabay.
New oil palm planting near a protected area in Indonesia. Image by Rhett A. Butler/Mongabay.


Scalability consists of three attributes: cost effectiveness, flexibility and sustainability.

In order for a system to be scalable and adopted by many users, it has to be cost effective at regional and national scales and across multiple years.

Lack of available data can increase the cost of a system significantly.

For instance, the overall cost of the mapping and monitoring system of the soy moratorium policy in Brazil is relatively low due to its reliance on free PRODES data. Furthermore, the Brazilian government had also mapped and registered property boundaries, further reducing the cost.

On the other hand, the HCSA mapping methodology requires intensive data collection and trained experts, and thus is relatively costly.

An assessment can cost up to $100,000, which often covers a single oil palm concession of around 10,000 hectares in Indonesia.

Besides being cost effective, a scalable system also have to be flexible, meaning it can be applied in other biomes, regions and countries while producing comparable results.

Lastly, it has to be sustainable, based on data that will be available and reliable over the long term. This is so that the system is available to users over time.

“There has been a substantial increase in freely available high temporal and spatial resolution satellite imagery, and accessible open-source software tools capable of classifying such imagery,” the paper states. “However, the production of consistent and comparable maps and analyses still requires a sustained commitment of resources.”

Soy field abutting against tropical forest. Degradation of land under intensive agriculture and deforestation are among other anthropomorphic influences that play a part in climate change. Image by Rhett A. Butler/Mongabay.


With 12 identified attributes that contribute to effective mapping and monitoring systems, there are bound to be some trade-offs.

The experts identified two of them: frequent detection or accurate monitoring, and local context dependence or large-scale consistency.

For the first trade-off, a mapping and monitoring system often has to choose between being highly accurate, which usually requires long-term data, or being near real-time, which is usually less accurate, particularly for small disturbances.

As a result, annual forest change monitoring system usually rely on data products with lower temporal resolution but higher accuracy, such as PRODES.

“In some cases this has resulted in a dual system, in which companies rely on accurate and vetted annual ‘definitive’ data for assessing and communicating effectiveness, but may also use near real-time data to support internal adaptive management and decision-making,” report author Austin stated.

This trade-off, however, can be addressed by improving the accuracy of near real-time forest disturbance detection. Once near real-time forest monitoring system achieves high accuracy, the need for multiple systems can be eliminated, Austin said.

She said there had been recent advances in near real-time change detection, including the RADD (RAdar for Detecting Deforestation) system. But notes that the tropical forests, which are the most under threat, come with special challenges.

“With an increasing number of satellite platforms with more diverse sensors (e.g., radar, optical) we expect that this tradeoff might become less stark,” Austin wrote. “However, clouds will always be a barrier to optical remote sensing in the humid tropics. Moreover, the very best annual products also incorporate field-based verification of violations – something that would be very costly to replicate for near real-time detection systems.”

The second trade-off, which might be more difficult to be addressed, is choosing between local relevance, inclusivity, and categorical detail on the one hand, and consistency, frequency, and sustainability on the other.

“Some users require local assessments that may be highly detailed and resource intensive, while others desire products that are regularly available, fairly low-cost, and consistent across large spatial scales,” Austin noted.

An example of a monitoring system that focuses on local assessments that are highly detailed is HCSA.

The platform builds credibility and legitimacy by aiming for high accuracy in specific geographies, integrating context-specific land-cover categories, and by including representation of community lands.

This approach might lead to differences in definitions across landscapes, and thus sacrificing consistency.

“HCSA guidelines are less scalable because of the relatively high cost associated with locally refined assessments and, as a result, will be less frequently updated and may exclude some small-scale producers,” the experts said in the paper.

Austin notes that these trade-offs show how there’s not a single one-size-fits-all ZDC mapping and monitoring system that will meet the needs of all users.

“The variety of users and user needs may mean that multiple systems are necessary,” she said. “Our characterization of mapping and monitoring system attributes can be used to evaluate where and how integrating multiple approaches might lead to more effective ZDC implementation.”

While using multiple systems can fulfill the needs of a variety of users, it can also lead to confusion, particularly in cases where there are conflicting results, according to Austin.

“For example, there are a number of high carbon stock (HCS) maps in development across spatial scales, from individual assessments to third party maps (e.g., Barry Callebaut) to government-led indicative HCS mapping efforts,” she said. “Groups that are developing these mapping and monitoring systems should carefully consider how they can best complement each other, without creating confusion among users.”

By Hans Nicholas Jong

Original Link: