Inundation and greenhouse gas dynamics of the central Congo Basin wetlands

It was recently discovered that the Cuvette Centrale region in the Central Congo Basin is home to the largest peat swamp in the tropics, forming a major store of carbon. However, its land atmosphere dynamics have not yet been comprehensively assessed. Inundation dynamics are the primary control on greenhouse gas emissions from peatlands. In this thesis, I present an assessment of the spatial and temporal dynamics of inundation across the Cuvette Centrale using ALOS-2 PALSAR-2 remotely sensed radar data, together with meteorological data. This involved first determining a relationship between water level and radar backscatter using field collected measurements of water table depth. Correlative statistics were used to categorise locations as rainfed versus those which receive additional water inputs from river flooding. I found that most of the rainfed peatlands exist to the west of the Congo River. These categorisations determined which interpolation methods were applied to transfer the radar data to water level estimates, with the output being daily water level maps at 100 m spatial resolution. To understand how the peatlands interact with the atmosphere, I then assessed their carbon fluxes by deriving the relationship between field measured greenhouse gas fluxes and water table depth. Scaled up maps of methane fluxes were produced at the same resolution as the water level maps, and the total regional fluxes were calculated. There are seasonally inundated areas that do not harbour peat, but which have soils which are still likely to emit methane when inundated. It was important to include these areas in the final water level and methane flux maps to arrive at a more comprehensive understanding of the methane fluxes from forested regions of the Cuvette Centrale. A map of seasonally inundated areas was developed by applying a threshold to the SAR data. Anomalies were calculated between GEOS-Chem atmospheric model derived background methane concentrations and Tropomi satellite measured column-averaged concentrations of atmospheric methane. These were used to identify a mismatch in the timing between increases in the ground-estimated methane fluxes and those seen from space. The maximum satellite observed methane concentrations coincided with falling water levels. Carbon dioxide soil fluxes remain positive even as methane fluxes increase, suggesting that the temporal lag in maximum methane concentration may be attributable, in part, to competition between methanogenic bacteria methane production and methanotrophic bacteria methane consumption. It has previously been shown that the peatlands exist close to a climatic drought threshold, receiving just enough net water input to sustain them, and may therefore only reach sufficient levels of inundation for significant methane production to occur during and just following the wettest months. Additionally, I used the derived water level maps together with previously published swamp type distribution maps to determine differences between the inundation characteristics of the two main swamp vegetation types, palm, and hardwood swamp. Finally, I propose practical applications for the developed methods, including identifying domed peatlands, updating peatland extent maps, and identifying additional regions of the Congo Basin that would likely have had conditions conducive to peat development under historically wetter climate conditions. By improving our understanding of inundation, swamp vegetation type characteristics and soil methane flux dynamics, this research can contribute to assessing how future changes in environmental factors, including land-use and climate change impacts, could impact on future inundation patterns, swamp type distribution and carbon fluxes.