FAO.org

In May and June 2023, the CAFI/FAO Deforestation and Degradation Drivers project, in coordination with the Gabonese national project partner, AGEOS, and the Gabonese Agency for Spatial Observation, organized a series of online sessions to build capacity for generating high-integrity forest data using Open Foris platforms, including SEPAL and Collect Earth Online (CEO), reaching more than 300 participants, of whom 40% were women.

The interactive online sessions culminated in a three-day face-to-face workshop at AGEOS headquarters in Nkok, Gabon, where 25 participants from local government, NGOs, research organisations and universities directly put the skills learned during the webinars into practice. This included:

• the creation of cloud-free satellite image mosaics and their classifications to create land cover maps using SEPAL;
• the use  of CEO to create training data and for validation of land cover maps;
• a new ensemble approach to assess forest cover change through statistical sampling: Ensemble Sample Base Area Estimation (eSBAE); and
• how to access high resolution satellite imagery from Planet’s NICFI program using SEPAL.

The online sessions, conducted in French, were highly interactive; participants who encountered issues were encouraged to go to the instructors, share their screens and troubleshoot. All information and course materials and recordings are available here.

At the AGEOS workshop in Gabon, participants were introduced to the drivers' project objectives, followed by a day-long discussion on the validation methods, including how to recognize specific direct drivers in high resolution Planet data provided by the NICFI program. On the second and third day, participants validated nearly 3000 eSBAE points identifying deforestation, degradation and the associated direct drivers across Gabon.

The progress was tracked by Aurélie Shapiro, FAO Chief Technical Advisor, who also assisted with quality control – data points that needed to be rechecked were identified through live data analysis, and then reassessed by the entire group at the workshop.

“Participants learned how to assess live statistics and analysis derived from the data validated in CEO, including an alert system to identify how many points needed to be reviewed for quality control, how many points have been validated, and how many remain. Furthermore, participants were able to understand the ensemble sample-based area approach (eSBAE) and visualize the estimated area of deforestation and degradation over the 2015-2022 time period, which are re-calculated as points are validated, along with the confidence intervals – which are reduced as more points are assessed in CEO.” Aurélie explained.

This intense and collaborative 3-week process pioneers an approach that is now being replicated in the other countries, including Republic of Congo, Democratic Republic of the Congo, Equatorial Guinea, Cameroon and Central African Republic, to develop country capacity and implement statistically sound methods to produce high-integrity data to assess forest cover change over the period 2015-2022 and identifying direct drivers. 

“The capacity development of local experts proved to be particularly important for many of the Gabonese administrations that took part. It enabled the experts to discover and understand Open Foris solutions for processing geospatial data. Mastering tools such as SEPAL will eventually enable these authorities to access and process large volumes of data without the need for major IT infrastructures,” said Conan Vassily Obama, FAO’s national consultant for the project in Gabon.

This week, the same online session series will be offered in Spanish from July 24-29, 2023. Interested individuals can find relevant Information and the registration link here.

Read more about the results of the project here.

This article was originally published here.