Moist synoptic transport of CO2 along midlatitude storm tracks, transport uncertainty, and implications for flux estimation
Date
2011
Authors
Parazoo, Nicholas C., author
Denning, A. Scott, advisor
Randall, David, committee member
Maloney, Eric, committee member
Kawa, Randy, committee member
Paustian, Keith, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Mass transport along moist isentropic surfaces on baroclinic waves represents an important component of the atmospheric heat engine that operates between the equator and poles. This is also an important vehicle for tracer transport, and is correlated with ecosystem metabolism because large-scale baroclinicity and photosynthesis are both driven seasonally by variations in solar radiation. In this research, I pursue a dynamical framework for explaining atmospheric transport of CO2 by synoptic weather systems at middle and high latitudes. A global model of atmospheric tracer transport, driven by meteorological analysis in combination with a detailed description of surface fluxes, is used to create time varying CO2 distributions in the atmosphere. Simulated mass fluxes of CO2 are then decomposed into a zonal monthly mean component and deviations from the monthly mean in space and time. Mass fluxes of CO2 are described on moist isentropic surfaces to represent frontal transport along storm tracks. Forward simulations suggest that synoptic weather systems transport large amounts of CO2 north and south in northern mid-latitudes, up to 1 PgC/month during winter when baroclinic wave activity peaks. During boreal winter when northern plants respire, warm moist air, high in CO2, is swept upward and poleward along the east side of baroclinic waves and injected into the polar vortex, while cold dry air, low in CO2, that had been transported into the polar vortex earlier in the year is advected equatorward. These synoptic eddies act to strongly reduce seasonality of CO2 in the biologically active mid-latitudes by 50% of that implied by local net ecosystem exchange while correspondingly amplifying seasonality in the Arctic. Transport along stormtracks is correlated with rising, moist, cloudy air, which systematically hides this CO2 transport from satellite observing systems. Meridional fluxes of CO2 are of comparable magnitude as surface exchange of CO2 in mid-latitudes, and thus require careful consideration in (inverse) modeling of the carbon cycle. Because synoptic transport of CO2 by frontal systems and moist processes is generally unobserved and poorly represented in global models, it may be a source of error for inverse flux estimates. Uncertainty in CO2 transport by synoptic eddies is investigated using a global model driven by four reanalysis products from the Goddard EOS Data Assimilation System for 2005. Eddy transport is found to be highly variable between model analysis, with significant seasonal differences of up to 0.2 PgC, which represents up to 50% of fossil fuel emissions. The variations are caused primarily by differences in grid spacing and vertical mixing by moist convection and PBL turbulence. To test for aliasing of transport bias into inverse flux estimates, synthetic satellite data is generated using a model at 50 km global resolution and inverted using a global model run with coarse grid transport. An ensemble filtering method called the Maximum Likelihood Ensemble Filter (MLEF) is used to optimize fluxes. Flux estimates are found to be highly sensitive to transport biases at pixel and continental scale, with errors of up to 0.5 PgC/year in Europe and North America.
Description
Rights Access
Subject
tracer transport
CO2 inversion
synoptic meteorology
biogeochemistry
carbon cycle