Greenhouse gas emission from tropical reservoirs : Spatial and temporal dynamics

Abstract: The emission of methane (CH4) and carbon dioxide (CO2) from reservoirs has been estimated to make up for about 1.3% of the global anthropogenic greenhouse gas emission. The impoundment of a river leads to the accumulation of sediment that is brought in from inflowing rivers, and the sediment organic matter is degraded to CH4 and CO2. CH4 is of particular concern as its global warming potential is 34 times stronger than that of CO2. In the tropics, high temperatures and high availability of fresh organic matter from high net primary production fuel CH4 and CO2 production. As the construction of hydropower plants is currently undergoing a boom, especially in the tropics, reservoir emission is probably bound to increase.The emission of CH4 and CO2 from reservoir surfaces is, however, highly variable, which makes current estimates uncertain. This thesis is built on the hypothesis that the spatial and temporal variability of greenhouse gas emission in tropical reservoirs, particularly of CH4 ebullition (the emission via gas bubbles), is so large that the sampling strategy affects whole-system estimates of greenhouse gas emission.This thesis shows that greenhouse gas emission from the four studied tropical reservoirs in Brazil varied greatly at different timescales – over 24 hours, between days and between seasons. Seasonal variability was identified as the most important temporal scale to be covered for CH4 ebullition inventories. In addition, the spatial variability of gas emission was large for all pathways. The variability of CH4 ebullition across space, for example, was estimated to be almost as large as its variability between seasons, and patterns of spatial variability in diffusive CH4 and CO2 emission differed between seasons. For both ebullition and diffusion, river inflow areas were prone to elevated greenhouse gas emission.This thesis shows that for retrieving solid emission estimates, there is no alternative to time-consuming measurements in the field. Measurements should be repeated at least once during each hydrological season (i.e. falling and rising water level). The seasonal surveys should cover space at a high resolution, including areas with and without river inflows, and different water column depths. CH4 ebullition made up for 60–99% of the total CO2-equivalent emission from the whole water surface of the studied reservoirs, with the highest contribution in the most productive reservoir. The most variable greenhouse gas emission pathway is therefore the most important one to be measured at appropriate resolution, particularly in productive reservoirs.