Using JERS-1 L-band SAR to estimate methane emissions from the Jau´ river floodplain (Amazon/Brazil)

Demonstrates how multi-temporal JERS-1 L_HH band SAR data can be used to derive detailed spatial and temporal information about wetland distributions and improve estimates of methane fluxes from these environments. Within the framework of the NASDA-led Global Rain Forest Mapping (GRFM) project, a part of the Jau´ river in the central part of the Amazon basin, was monitored during two full flooding cycles with JERS-1, and a spatial/temporal model of the inundation patterns along a 40-km long segment of the river was developed. Simultaneously with the SAR acquisitions, measurements of CH₄ emissions and flooding extent along a transect through the seasonally flooded forest in the Jau´ were performed. The ground measurements verify a clear relationship between the L-band backscatter and the status of inundation. A simple CH₄ flux model was developed and in combination with the JERS-1 inundation model, daily CH₄ fluxes could be estimated. Early results confirm significant CH₄ emissions from floodplain environments, indicating amounts up to 2000-4000 tons/year from the 1000 km² test area, where some 200 km² correspond to seasonally flooded forest. CH₄ fluxes were found to vary significantly between years depending on the river level, with the maximum flooded area being a key control parameter. Model results suggest that about 75% of the yearly fluxes in the basin occur within a period of about three months, when the river levels are receding