citation
Schneider, M., C. Winchester, E. Goldman, and Y. Shao. 2023. “Mapping cocoa and assessing deforestation risk for the cocoa sector in Côte d’Ivoire and Ghana.” Technical Note. Washington, DC: World Resources Institute. Available online at: doi.org/10.46830/writn.21.00011.
overview
The Cocoa Deforestation Risk Assessment (Cocoa DRA) is a map layer that identifies the risk of future deforestation events linked to cocoa in Côte d’Ivoire and Ghana. Cocoa and chocolate companies, as well as other stakeholders in this sector can use this map to identify and prioritize interventions in cocoa supply chains.<br>Multiple geospatial datasets were brought together to create the Cocoa DRA, which represent landscape features known to influence the probability of deforestation, including explanatory variables such as recent loss, terrain suitability for cocoa production, and human accessibility. We also included the West Africa Cocoa dataset as a variable, which we created by collecting and collating cocoa plot data contributed by 19 cocoa and chocolate companies. All variables were first analyzed at a 30-m resolution and then rescaled to a 1-km resolution to better capture broader landscape trends and ease visual interpretation. Finally, the results were reclassified to 5 priority classes which communicate the level of risk relative to the remainder of the study area.<br><br>1-km pixel values are encoded as follows:<br>1 = high risk, 2 = mid-high risk, 3 = medium risk, 4 = mid-low risk, 5 = low risk, 0 = no forest<br><br>The Cocoa DRA is not geared toward a specific type of infrastructure or level of traceability, meaning that it can be used to assess areas of interest as small as a cocoa plot or as large as a landscape. However, the type of areas of interest must be comparable; for example, comparing the risk associated with a cluster of farms to the risk associated with the sourcing area of a cooperative is an inappropriate use of the Cocoa DRA.<br>Users should note that for the Cocoa DRA, low risk does not immediately translate to low prioritization for interventions. It shows the likelihood of forest loss occurring, but on its own, does not show where forest loss has potential to do the greatest harm. Risk assessments should include additional data that show the presence of high-interest features such as primary forest or boundaries of protected areas. For example, Taï National Park in Côte d’Ivoire and Bia National Park in Ghana have experienced a low amount of recent tree cover loss and as a result deforestation risk is low within these forests. <br>The map layer currently on the Open Data Portal is version 2 and shows priority landscapes in 2023. The Cocoa DRA will be updated annually alongside updates annual tree cover loss data; for the 2023 version, recent loss as a variable includes forest loss that occurred between 2018 and 2022. Other inputs will be updated periodically as improved data becomes available. A full description of the methods can be found in this [technical note](https://www.wri.org/research/mapping-cocoa-assessing-deforestation-risk-cocoa-cote-divoire-ghana).