Articles & Stories

SERVIR Applied Sciences Team helping African countries track and protect forests

REDD_carbon_screenzoom.jpg

With a growing population, increased demand for agricultural land and fuel wood are depleting Africa’s forests. Forested areas are important because they take in large amounts of carbon and release oxygen. When forests are removed or degraded, less carbon is taken from the atmosphere and carbon emissions increase. This is believed to hasten climate change and increase its impact.

Participants from DeKUT and DRSRS with M. Osunga of RCMRD
From Left: Professor Charles Mundia – Dedan Kimathi University, 
Christopher Amudavi – DRSRS, Michael Osunga – RCMRD  

SERVIR Applied Sciences Team PI Scott Goetz, Co-I Nadine Laporte, and Ned Horning of the American Museum of Natural History, together with the Regional Centre for Mapping of Resources for Development (RCMRD) -- the SERVIR-Eastern and Southern Africa hub institution -- are working in African countries to equip forestry technicians and geospatial analysts with tools and skills to monitor and document the extent of forests and where they are changing as a result of deforestation and forest degradation. Toward that end, the team has hosted several training sessions over the past few years.

The latest such event, the Tools for Carbon Emissions Estimation and Mapping in East Africa workshop, was held 22-24 March 2016 at RCMRD in Nairobi, Kenya. This training introduced standard and new methods for processing satellite data and estimating accuracy of land cover and land-cover change maps using open source software.

Maureen Kabasa, an intern at the Kenya Forestry Research Institute, attended the training.

“Where I work, GIS experts are few, so I will be able to help in terms of manpower and transfer of knowledge gained to the team to improve the department,” says Kabasa. “This data will become handy to inform decision making and have a clear representation of how the carbon stocks, land cover, emissions, and corridors are changing in terms of not only how they affect forests but also the other species.”

Participants from NFA, AWF and Zambia Ministry of Forestry with E. Ouko of RCMRD
From left: Edward Senyonjo – NFA, Cathy Kauda – Ministry of 
Forestry Zambia, Tom Kemboi – AWF, Edward Ouko – RCMRD  

Mapping land cover change and estimating the resulting greenhouse gas emissions are becoming increasingly important as tropical countries are participating in agreements with programs like REDD+ (Reducing Emissions from Deforestation and Forest Degradation). REDD+ operates under the United Nations Framework Convention on Climate Change (UNFCCC) to fund tropical countries in proportion to their efforts to protect their forests and reduce greenhouse gas emissions, particularly CO2 related to deforestation and degradation. Countries engaged with REDD+ require accurate information about their forests' extents and concentrations in order to assess their carbon stores and develop forest conservation strategies.

To implement REDD+, countries need to be supported in strengthening their technical capacities across a wide range of work areas, from forest inventories to land use mapping. Funds for monitoring forests and the carbon they contain are still limited, so gaining access to state-of-the-art open source software for spatial data processing for forest and carbon mapping is particularly important for these countries.

At the Nairobi workshop, participants learned about one such tool -- the “R” open source software package.  Attendees found the training very helpful. Edward Senyonjo, Remote Sensing Specialist at the National Forestry Authority in Uganda, says that for him, the SERVIR training “has shortened the mapping cycle from six years to about one year. …Time series analysis is possible and has contributed to the estimation of emissions and reference levels as well as the national communications to UNFCCC.”

Participants from Karatina University and DRSRS
From Left: Isaac Njoroge – Karatina University, 
Edward Juma – DRSRS, Christopher Amudavi – DRSRS  

Edward Juma, Technical Officer for Kenya’s Department of Resource Surveys and Remote Sensing (DRSRS), states “I am now able to do classification and change detection in land cover using R and wall to wall mapping of land cover types. This will help me improve on accuracy and speed of delivery in my related assignment at work.”

Christopher Amudavi, DRSRS GIS/Remote Sensing Technician, agrees, noting that the training resulted in “an improvement in terms of accuracy and time taken to carry out land cover classification, change detection, and emissions estimation.”

Although this is the final workshop of this SERVIR project, the team plans to continue to work in the region, advancing capabilities to map and monitor forests for conservation while also sustaining local livelihoods.

RCMRD’s Phoebe Oduor emphasizes the value of all that has been learned.

“We were able to obtain the Random Forest scripts during the training and have since customized and used them in our programs. For example, it was very useful for our Systems Land Based Emissions Estimation for Kenya (SLEEK) land cover change mapping program. The opportunity to use the project’s biomass product for our national land use emissions estimations has been a missing link. We can now use this as a starting point for further improvement."