- The U.N.’s Green Climate Fund (GCF) has approved the first proposal for REDD+ emissions reductions payments, totaling $96 million for around 19 million tons of emissions reductions.
- However, GFC board members and observer NGOs expressed concern over how the emissions reductions are calculated.
- A study published last month sheds light on the difficulty of accurately calculating changes in forest cover and calls for a more standardized approach.
The United Nations’ Green Climate Fund (GCF) has accepted the first proposal for REDD+ results-based payouts from Brazil, effectively paying the country for reducing its deforestation rates in 2014 and 2015, as compared to the 1996-2010 average. In return for around 19 million tons of emissions reductions, the GCF has agreed to pay Brazil $96 million, which the country says it will use to launch a program called Floresta+ aimed at ecosystem restoration, the provision of environmental services, and strengthening the country’s REDD+ strategy.
At the GCF board meeting held earlier this week during which the decision was made, board members reportedly alluded to another, larger proposal that may be submitted at future GCF meetings. This proposal is expected to request payment for a further 47 million tons of emissions reductions made in 2016, 2017 and 2018.
Many GCF board members are calling the move a “turning-point” for REDD+ and a historic moment for the conservation scheme that mitigates carbon emissions by paying countries to conserve their forests. However, some observing civil society organizations called to postpone the proposal. They drew attention to issues with how the baseline measurements against which deforestation reductions — called “forest reference emission levels,” or FREL — are calculated, whether the results will be permanent, and whether Brazil can adhere to social and environmental safeguards given the recent change in government.
While other countries gear up to submit their own proposals for portions of the $500 million allocated to the pilot payout program, a new study published in February sheds light on the difficulties of calculating the emissions savings of individual countries. Conducted by researchers from the University of Edinburgh and published by Environmental Research Letters in February, the study examines whether forest areas that historically have had high carbon emissions are targeted in the reference levels submitted by countries for REDD+ results-based payments.
The researchers found discrepancies between forest loss detected independently through the satellite-driven Global Forest Change v1.4 (GFC) data set and country-reported forest loss in seven countries in the Asia-Pacific region.
“The intention of REDD+ is to have countries reduce emissions. But the limited data and quality of data about forests might not be enough,” says the lead author of the study, Keiko Nomura, who previously worked for the U.N. REDD program in Myanmar, Sri Lanka and Cambodia. “We are looking at how we can make REDD+ more effective using more standardized tools like GFC to calculate actual forest change.”
Thirty-nine countries have submitted their reference levels to the U.N. Framework Convention on Climate Change (UNFCCC). These baselines are ultimately used to calculate payments from REDD+ if future monitoring can show a positive deviation from the reference level. Accurately calculating changes in forest cover is essential for understanding the carbon emissions from tropical forest loss, but has shown to be challenging for the implementation of REDD+.
“If there are biases in the setting of reference levels … then REDD+ will inevitably be less successful at reducing the rise in atmospheric greenhouse gases,” Nomura and her colleagues write in the paper. Successful implementation of REDD+ is essential to meeting the 2-degree-Celsius (3.6-degree-Fahrenheit) threshold outlined in the Paris climate agreement, according to many in the international community.
REDD+, enshrined in the Paris Agreement as the main financial mechanism to incentivize tropical countries to protect their forests, has been more than a decade in the making. While many previous studies have focused heavily on the policy and governance aspects of REDD+ programs, few studies have focused on its technical aspects, like forest reference emissions levels.
“There is very little interaction between policy and academia, which is why I decided to write this paper,” Nomura told Mongabay. “Ultimately the countries will get payment based on REDD+. We want to make sure that that payment actually reflects the reality on the ground.”
Defining a forest
The study compared forest area estimates submitted by seven countries in Asia and the Pacific with forest area estimates using the GFC satellite data set from 2000-2016. They looked at whether the “country-defined REDD+ forests” actually represent the main sources of emissions for Cambodia, Indonesia, Malaysia, Nepal, Papua New Guinea, Sri Lanka and Vietnam.
Previous studies have pointed to inconsistencies in deforestation data self-reported by individual countries. Complicating the designation of forest reference levels is the fact that there is no general definition of a forest.
“The REDD+ analysis of forest change and GFC is not an apple to apple comparison. It depends [on] how countries define a forest,” Nomura said. “The GFC data has one consistent and biophysical definition; it doesn’t discriminate as to whether a forest should be included or not.”
To estimate forest lost in the context of REDD+, countries must first decide what constitutes a tree, how many trees need to be grouped together to constitute a forest, and finally what activities are considered deforestation. Some countries, like Malaysia, include monoculture tree plantations in their forest analyses. Others, like Indonesia, only consider changes in natural forests and not plantations when defining deforestation.
Because legal definitions of forests don’t take into account all aboveground forest biomass where carbon is stored, reference levels don’t always capture the full emissions from changes in biomass due to changes in land use, the authors argue. Instead, the authors used satellite-based observations of tree cover change from the Global Forest Change data set, developed by the University of Maryland, to provide alternative estimates of deforestation rates.
Their analysis revealed significant differences between official reference levels reported by the seven countries they analyzed and satellite-detected tree cover loss in those countries. For example, Vietnam reported a net increase in forest cover between 2000 and 2010 while the GFC showed a net loss. The authors say this is due in part to the expansion of tree plantations in the country, as well as the inclusion of trees less than 5 meters (16 feet) in height, which gives the illusion of growing forests.
Defining deforestation
The study also shows that rates of forest area change vary considerably depending on the timing of the reference period. In Cambodia, for example, the reference period stopped at 2014, which yielded a deforestation rate of 2.9 percent. However, a 2010 end date for the reference period would have provided a 0.9 percent deforestation rate. The researchers write that this reference level overstates forest loss, resulting in the potential for over-claiming emissions reductions in the future.
But the issue isn’t just about overstating emissions savings. The authors are also concerned about the potential to miss emissions.
“There is a lot of forest still remaining in areas that are not politically forests, not considered in REDD+,” Nomura said. “In our research, we found … lots of remaining forest in those abandoned agricultural areas [that] could contribute to emissions reductions.”
Another issue raised by the researchers concerns minimum mapping units (MMUs), which are the smallest areas considered when analyzing forest change. They write that, in most cases, countries set different size thresholds when defining what counts as a forest and defining the areas that can be included in deforestation analyses. For example, Nepal sets its forest threshold at 0.5 hectares (1.2 acres) while it sets its MMU at 2.25 hectares (5.6 acres). This means that deforestation events in forests smaller than 2.25 hectares aren’t considered when the country evaluates its forest loss.
However, Nomura and her colleagues found that loss rates were up to 40 percent higher when MMUs were set to the same size as forest definitions. This, they write, is particularly important for countries where most deforestation happens on smaller plots. Such is the case in the Amazon Basin, where research indicates small-scale deforestation events — defined as clearances less than 1 hectare (2.5 acres) in size — are on the rise.
Brazil’s proposal that was accepted this week by the GCF similarly uses 0.5 hectares as the size threshold for defining a forest, but its MMU for what’s counted as forest loss is set at 6.25 hectares (15.5 acres).
“If you use much bigger mapping units then you miss a lot of deforestation,” Nomura said. “It doesn’t mean they are doing it maliciously, it’s more for consistency. In Brazil’s case it looks like they initially started to use a Landsat measurement of 6.25 and have continued to use that.”
Setting a precedent
Brazil’s forest reference emission level for its Amazon biome was the first reference level submitted to the UNFCCC for use in results-based payments. Now that Brazil’s proposal has been accepted, GCF board members suggested other countries might look toward it for the creation of their own forest reference emissions levels and submissions to the pilot program for REDD+ results-based payments.
“This is the living object upon which we all can learn for future projects in that area,” said board member Stefan Schwager.
Forest advocates and policy experts have previously expressed concern over how reference levels are calculated, and whether the emissions savings will be permanent.
This study falls in line with others that have called for the use of satellite-based forest cover change data to provide more accurate evaluations of deforestation and forest degradation rates.
“The key point of REDD+ is protecting forests under threat,” Nomura said. “In some countries’ submissions, the historical forest cover change doesn’t show signs of vulnerability. But when using GFC data, the forest cover in the same country can be decreasing. This is the mismatch; the planned REDD+ may not be targeting the right forests.”
During the GCF board meeting this week, members expressed a desire to see improvements in the way scorecards are assessed and forest reference emission levels are calculated, issues that will likely resurface again at the 24th board meeting in October 2019, when the board is also expected to ask for a replenishment of its funds.
Banner image: A waxy monkey frog (Phyllomedusa sauvagii), one of the many denizens of Brazil’s rainforests. Photo by Rhett A. Butler/Mongabay.
Citation: Nomura, K., Mitchard, E. T., Bowers, S. J., & Patenaude, G. (2019). Missed carbon emissions from forests: comparing countries’ estimates submitted to UNFCCC to biophysical estimates. Environmental Research Letters.
Feedback: Use this form to send a message to the editor of this post. If you want to post a public comment, you can do that at the bottom of the page.