Investigating land-air carbon fluxes using a Lagrangian model and satellite retrieved carbon dioxide
thesisposted on 2011-01-10, 12:21 authored by Alan James Hewitt
The existing generation of satellite instruments (such as SCIAMACHY and AIRS) has allowed the retrieval of atmospheric mixing ratios of carbon dioxide. The feasibility of using these and later satellites (OCO-like or GOSAT) to investigate carbon fluxes between the terrestrial biosphere and the atmosphere, either alone or complemented by the high precision but low density network of surface measurement sites has been investigated. A methodology to investigate regional scale carbon budgets, based on the UK Met Office Lagrangian trajectory model NAME (Numerical Atmospheric-dispersion Modelling Environment), has been developed and demonstrated. A forward modelling methodology was developed, where top-down surface flux information from CarbonTracker was combined with the background CO2 mixing ratio to obtain an atmospheric concentration. Synthetic testing of the initialisation method demonstrated that a strong correlation coefficient (R2 ≈ 0:9) between the forward modelled and satellite observed atmospheric CO2 fields can be achieved. Forward modelled CO2 concentrations using CarbonTracker fluxes were demonstrated to be moderately correlated with the SCIAMACHY-retrieved CO2 field (R2 varies by month, from 0.4 to 0.8). An inverse modelling methodology was developed, where the change in carbon mass between the satellite-retrieved CO2 columns and the background concentration was combined with the surface residence time from the NAME model. Synthetic testing of the inversion method has shown that the a posteriori flux covariance scaled linearly to the satellite-retrieved error covariance and inversely to the NAME residence time of the ecosystem. On the regional scale, this method could improve on the carbon flux estimates from CarbonTracker and an equivalent Eulerian method using GOSAT. This thesis also presents the first carbon fluxes inverted from satellite retrieved CO2 columns, which captured the seasonality of the carbon fluxes of the vegetation and negligible ocean fluxes.