 |
| Titel |
Inverse model algorithms for GOSAT L4 regional carbon flux product |
| VerfasserIn |
S. Maksyutov, Y. Koyama, V. Valsala, T. Oda, D. Belikov, M. Saito, A. Ito, P. K. Patra |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250026496
|
|
|
|
|
|
| Zusammenfassung |
| An inverse modeling system for retrieval of the regional fluxes of carbon dioxide from
GOSAT data is being developed and tested. The system is based on tracer transport
models for ocean and atmosphere, observation optimized process models of the
carbon cycle in the atmosphere and ocean, inventories of the anthropogenic and
natural flux components for which process models are not available (fires, fossil fuel
emissions). Offline atmospheric tracer transport model used in the inversion has been
modified to employ mass conservative flux-form transport algorithm on a reduced
grid, resulting in better vertical profile simulation and more consistent regional flux
distribution simulated with inverse model. Northern extra-tropical land sink simulated
with the model is now close to the range supported by independent observations.
Global atmospheric transport model is also enhanced with a Lagrangian particle
diffusion model running in backward mode, than can simulate realistic fine resolution
observation footprints and better synoptic scale variability. Fossil fuel emissions are
improved by using databases of large point sources and weather-dependent electricity
demand parameterization. The parameterization is based on monthly energy statistics
collected mainly for North American regions. The terrestrial biosphere is modeled with
a process based VISIT model at a daily time step. It has been validated against
atmospheric observations. Although the model phenology is driven by meteorology alone
without use of remote sensing products it is reproducing atmospheric CO2seasonal
cycle phase and amplitude. A merged precipitation dataset based on JCDAS and
NCEP reanalyses is used to compensate modeled precipitation biases. Oceanic
surface pCO2 is simulated with an ocean tracer transport model based on reanalyzed
currents coupled with simple ocean biogeochemistry model (after McKinley et al),
which is further adjusted to observed surface pCO2 data from LDEO and NIES
databases using 4-D variational assimilation technique. A spatially varying constrain on
pCO2 is used to allow for large interannual variations of air-sea pCO2 fluxes in the
regions sensitive to climate variability. In the most of the globe the ocean model flux
seasonality show good correlation with observation based climatology. Inverse model
of atmospheric transport solves for monthly mean fluxes at 64 regions globally,
with possible increase in number of regions and time resolution in near future. |
|
|
|
|
|
|