Inverse modelling of CO2 sources and sinks using satellite data and the importance of transport model uncertainties
OCO and GOSAT will likely bring a wealth of accurate total column CO2 measurements that will allow the estimation of CO2 sources and sinks from space. Meanwhile even more advanced measurement concepts are being investigated for the next generation of instruments, such as the CO2 lidar A-SCOPE. Despite numerous attempts to simulate the benefit of remote sensing for the quantification of CO2 sources and sinks using theoretical Observing System Simulation Experiments (OSSEs) it is difficult to predict the performance of a real-world application. One of the potentially important shortcomings of the OSSEs, which have been reported so far, is the difficulty to account, in a realistic manner, for the impact of transport model uncertainties. As part of a preparatory study for the A-SCOPE mission we have tried to quantify these errors using a model inter-comparison experiment, including the IFS, LMDZ, TM3, and TM5 models. Synthetic A-SCOPE measurements were generated for each model using a common setup of CO2 fluxes and initial boundary conditions. The difference between the samples generated by any combination of models has been prescribed as pseudo measurements in CO2 inversions. The deviations of the retrieved fluxes from the true (common set-up) fluxes quantify the impact model errors provided that the ensemble of models can be considered a realistic representation of transport model uncertainty. This representativeness has been tested by comparing the ensemble performance against available total column CO2 measurements. The derived transport model uncertainties are put into perspective by comparison with the anticipated A-SCOPE measurement uncertainties.
EGU General Assembly Conference Abstracts
- Pub Date:
- April 2009