HR: 1340h
AN: A13D-1514 [Abstracts]
TI: On the utility of GOSAT column average CO2 observations in reducing uncertainties of CO2 flux estimates
AU: Kadygrov, N
EM: nikolay.kadygrov@nies.go.jp
AF: National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, 305-8506, Japan
AU: * Maksyutov, S
EM: shamil@nies.go.jp
AF: National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, 305-8506, Japan
AB:
The utility of upcoming GOSAT column CO2 observations in surface CO2 flux estimations was evaluated. The
main aim was to estimate the CO2 flux uncertainty reductions contributed by GOSAT data. To simulate the global
distribution of total column CO2, we used the NIES tracer transport model at 2.5x2.5 resolution. Observation
errors on 151 surface and aircraft observation sites used in this work were assigned according to (GLOBALVIEW-
CO2, 2006) data product and had seasonal variations. All column CO2 data were aggregated to 7.5x7.5 degree
grid cells. GOSAT observations were simulated with transport model. To simulate the spatial and temporal
distributions of total column CO2 observation frequency and uncertainty of aggregated monthly mean
observations, we used orbit parameters and the monthly mean total cloud coverage from JCDAS reanalysis (to
estimate clear-sky frequency). Total column CO2 error is combined of systematic retrieval bias and random error
adjusted by number of successful observations for each aggregated grid cell within a month, assuming a single-
shot retrieval error of 5 ppm. We applied 66-region time-dependent Bayesian inverse model in cyclo-stationary
mode and estimated the regional flux uncertainties of monthly mean regional CO2 flux with and without satellite
data. Our results show that GOSAT data with 1.5ppm total error (for monthly mean, observation over land-only)
achieve same mean flux uncertainty as the observation data from the existing surface network, with assumed
systematic error contributing significantly to the total error in this case. Our estimation shows that the relative
reduction in surface flux uncertainties is about 50-60 % for some regions. Expectedly, most of reduction in
uncertainties occurs in the regions with low density of surface observation
DE: 0322 Constituent sources and sinks
DE: 0368 Troposphere: constituent transport and chemistry
SC: Atmospheric Sciences [A]
MN: 2007 Fall Meeting