Allocating the REDD+ national baseline to local projects: A case study of Cambodia

This study aimed to propose a set of decision support tools to allocate Cambodia's REDD+ national baseline or initial forest reference level (FRL) to local REDD+ projects, based on their forest cover and forest carbon stocks, and the historical deforestation trends in their reference regions. Our samples included 77 hypothetical REDD+ projects and five actual REDD+ projects. To identify reference regions for our samples, a cluster analysis of 127 districts in Cambodia was conducted using the Partitioning Around Medoids (PAM) algorithm (or k-medoids method). To calculate the baseline amount to be allocated to projects, four allocation methods were proposed. Two methods used ‘snapshot’ variables (i.e., [1] existing forest area or [2] forest carbon stock) and two used ‘change’ variables (i.e., [3] historical forest area change or [4] historical forest carbon stock change from 2006 to 2014). We weighted the baseline by the deforestation risk in 2014. We found that ‘snapshot’ methods tended to allocate more baseline if the project had more forest cover or forest carbon stocks. In contrast, ‘change’ methods tended to allocate more baseline if the project has more forest area loss or forest carbon stock loss between 2006 and 2014 in its reference region. These differences suggest that the ‘snapshot’ methods do not consider the inequity in the amount of baseline allocated to the projects. This inequity stems from the different degree of effort made by project stakeholders to protect their forests. The degree of effort varies depending on the historical deforestation trend in a project's reference region. However, the ‘change’ methods explicitly incorporate the trend into their calculations of baseline allocations. Technically, the ‘change’ method [4] seems to be the desired choice for allocating FRL to local projects; this is because it fits best with the idea of counter-factual thinking using a ‘reference period’ that a national baseline must have under the United Nations Framework Convention on Climate Change (UNFCCC). The ‘change’ methods are less likely to stimulate further deforestation, in a jurisdiction with high forest cover or forest carbon stocks that does not have a REDD+ project, compared with the ‘snapshot’ methods. However, the ‘change’ methods are more likely to face the challenge of political adjustment to balance the reduced emissions from some jurisdictions/projects with increased emissions from the other jurisdictions/projects, compared with the ‘snapshot’ methods. The findings are applicable not only in Cambodia but also in other REDD+ countries that include deforestation as one of the five REDD+ activities, within the scope of their national baseline submission to the UNFCCC. The study's outcomes will facilitate stakeholder dialogue on the proportions of national baseline allocated to sub-national programs/projects by showing the technical and political implications of the proposed allocation methods.