Estimating North American CO2 Fluxes within a Geostatistical Inverse Modeling Framework: Preliminary Pseudo-data Results
Sharon
Gourdji, University of Michigan, sgourdji@umich.edu
(Presenting)
Adam
Hirsch, National Oceanic & Atmospheric Administration, adam.hirsch@noaa.gov
Kim
Mueller, University of Michigan, kimlm@umich.edu
Deborah
Huntzinger, University of Michigan, dnhuntzi@umich.edu
Arlyn
Andrews, National Oceanic & Atmospheric Administration, arlyn.andrews@noaa.gov
Anna
M
Michalak, University of Michigan, amichala@umich.edu
A geostatistical inverse modeling analysis is currently underway to estimate CO2 fluxes at a finer spatial (i.e., 1°x1°) and temporal resolution (i.e. sub-monthly) than previous studies, providing better understanding of the spatio-temporal distribution of North American sources and sinks for 2004 and 2006, with possible extension to 2005 and 2007. The flux distribution will be estimated using continuous atmospheric CO2 concentration measurements from 10 tall towers located across the continent, along with an atmospheric transport model driven by high-resolution meteorology (STILT/ WRF) and environmental auxiliary information (i.e. precipitation, air temperature, etc.). This poster will show results from a pseudo-data inversion for June 2004 estimating fluxes at various temporal resolutions with minimal use of auxiliary environmental data within the statistical model of the trend. The purpose of the pseudo-data study is to addresses the potential biases associated with resolving (or not resolving) the diurnal cycle, and the optimal concentration averaging and flux resolution in order to estimate an 8-day average flux with maximum precision. The study also evaluates the ability of the transport model to correctly represent planetary boundary layer dynamics near tall tower measurement locations.
NASA Carbon Cycle & Ecosystems Active Awards Represented by this Poster: