The Importance of the Spatial Density of Satellite Measurements for the Retrieval of Spatial Flux Patterns

Abstract:

Initial results from GOSAT flux inversions of column-integrated carbon dioxide suggest a significant redistribution of surface fluxes compared to inversions using only surface-based inversions as an observational constraint. New evidence suggests that this redistribution of fluxes is a robust feature, and is related to the increased spatial density of the measurements made available by remote sensing. However GOSAT's rather large measurement footprint and sparse sampling still provide poor coverage over many areas of the globe, particularly regions characterized by consistent cloud cover, such as the tropics, and all passive near-infrared sensors suffer from a seasonal sampling bias due to limited sunlight during high latitude winter. As such, errors in the pattern of retrieved fluxes may still be significant. Active sensors based on lidar do not suffer from the same seasonal (or diurnal) sampling biases, and their exceptionally small instantaneous field of view (~150 m) promises to greatly improve the spatial coverage of the measurements over partially cloudy regions. Using the case of MERLIN, a planned joint French-German lidar mission designed to measure XCH4, the implications of this increased spatial coverage is considered in an inverse modelling framework, and compared to presently available measurement coverage from the surface-based network and GOSAT. The gain in knowledge about the absolute size of the regional methane fluxes, particularly in currently undersampled regions such as the Arctic permafrost zones and tropical wetlands, is quantified.