Estimating scalar sources, sinks and fluxes in tropical forests using inverse models
Mario
B.
Siqueira, Nicholas School of the Environment and Earth and Oceans Sciences, Duke University, mbs4@duke.edu
(Presenting)
Humberto
Ribeiro da
Rocha, Departamento de Ciências Atmosféricas, Universidade de Sao Paulo, humberto@model.iag.usp.br
Michael
L.
Goulden, Department of Earth System Science, University of California Irvine, mgoulden@uci.edu
Scott
Dennis
Miller, Department of Earth System Science, University of California Irvine, sdmiller@uci.edu
Renato
Ramos da
Silva, Department of Civil and Environmental Engineering, Duke University, renato@duke.edu
Katul
G.
Gabriel, Nicholas School of the Environment and Earth and Oceans Sciences, Duke University, gaby@duke.edu
The estimation of scalar source and sink distribution and vertical fluxes within canopies is critical to quantifying biosphere-atmosphere mass and energy exchange rates. Direct measurements of the source/sink strength are impractical at scales larger than a single leaf. On the other hand, measurements of scalar concentration profiles are routinely performed in many field experiments. Since scalar source/sink distributions are directly related to scalar concentration by continuity and turbulent transport equations, “inverse modeling” can be used to infer these distributions from the routinely measured mean scalar concentration profile. Over the past decade, several inverse models have been proposed and utilized Lagrangian, Eulerian, and hybrid approaches. Although these models have been successfully used in temperate and boreal forests, they were not tested for tropical forests known to have complex canopy morphology and strong density gradients within the canopy. Measurements suggest that in closed tropical forests, a persistent stable layer near the forest floor exists. The dynamics of this layer presents a major challenge for these inverse models given that they were originally derived for neutral atmospheric stability with semi-empirical corrections to boundary conditions for non-neutral stability. This shortcoming was addressed theoretically via a revised Eulerian approach for heat exchange; however, this approach has not been extended to other scalars. Here, we propose to extend and test this Eulerian method to water vapor and CO2 using data collected at LBA FLONA Tapajos Logged Forest Tower Site in Santarem, PA, Brazil. This site was chosen because of the availability of long-term temperature, water vapor, and CO2 concentration profiles along with fluxes of CO2, latent heat and sensible heat.