What
Production of a land-cover map of the reservoir catchment area behind the dam.

Where
Nam Theun dam (450 sq.km), Khammouane province on the Nakai Plateau, Laos.

Keywords
Biomass, GHG emissions, forest.

Context
New dam projects, mainly located in tropical zones, require an accurate estimate of biomass in the future catchment area. Recent examples like the Petit Saut dam in French Guiana show that when a large amount of biomass is flooded, water quality can be adversely affected and huge volumes of GHG released.
Consequently, it is vital to characterize vegetation and quantify biomass in the area to be flooded.

Partners/clients
EDF (Electricité de France)

Objectives
Provide our client EDF with a land-cover classification for the future reservoir behind the Nam Theun dam, to help them assess the overall carbon masses involved.

Methodology/Results
Our task for this project was accomplished in two stages. First, the client furnished reference terrain data and photos of specific points matching the land-cover classes. We ran these data concurrently through two sequences of biophysical processing operations with, the first to characterize water bodies and the second to discriminate types of land vegetation. Classification rules were then implemented as a decision tree, using the reference data. Lastly, a classification of the zones in the dam’s catchment area was produced, with 8 classes. This classification was used by the domain expert commissioned for the project to estimate the biomass that would be flooded, and thus the risks of generating carbon emissions.

This study showed that the total organic carbon stock is estimated at 5.1±0.7 MtC. This is much lower than in other tropical reservoirs in South America, and the net greenhouse gas (GHG) footprint due to the emissions occurring in the first years is expected to be well below the one that would be produced by an equivalent thermal power plant.

more...

Data used
SPOT 5 10-m

Land-cover classification
Nine cover types were defined for the 450 sq.km of the future reservoir: dense, medium, light, degraded and riparian forests, agricultural soil, swamps, water and “others” (road, construction sites, etc.)