Close Window

CD-07 Abstract

High Resolution Carbon Exchange Over Large-Scale Amazonia Based on Modeling and GOES Satellite-Derived Radiation Inputs

Harry J. Cooper — Florida State University (US-PI)
Pedro Leite Silva Dias — USP - Universidade de Sao Paulo (SA-PI)
Eric A. Smith — NASA/GSFC (US-PI)


Photosynthesis and respiration by vegetation and soil play a vital role in the carbon

cycle over the rain forest in the Amaz˘nia region. Among various key factors in

determining these physiological processes by models are photosynthetically active

radiation (PAR), and the canopy and soil temperatures (Tc , Ts). For example, it has been

suggested that the CO2 uptake rates over the Amazon forest may be highly

sensitive to radiation and temperature. That the system may change from a sink to a source

in response to reduced radiation levels due to cloud cover or an average temperature rise

of a degree or less. However, our knowledge of the space-time distribution of PAR and Tc

is too incomplete to facilitate large-scale comprehensive modeling studies. The data

currently available for Amaz˘nia come from only a few sites, and thus past studies of

this kind have been restricted to specific time periods and areas. Moreover, extrapolation

of datasets to larger spatial scales and longer time periods may not be appropriate

because of discontinuous processes controlling PAR and Tc.

Scientific Objectives

This proposal address these concerns with the following four main scientific


1. Retrieval of the surface radiation budget over the large-scale Amazon basin,

including retrieval of PAR, Tc, and Ts, based on GOES-8 satellite measurements;

2. Examination of net CO2 exchange between the atmosphere and the

soil-forest ecosystem as a function of land type and land-use based on a biosphere model

forced with satellite-derived inputs;

3. Analysis of the basin-wide spatial and temporal variability of net CO2


4. Quantification of the effects of distribution of environmental variables such as

cloudiness, boundary layer thermodynamics, and aerosol concentration on scaling up both in

space and time of seasonal and basin-wide CO2 exchange.


Retrieval of the surface total solar fluxes and PAR fluxes is to be carried out with a

physical algorithm designed for GOES satellite measurements which has been successfully

applied to the BOREAS project. For the long-wave fluxes, we will use a new algorithm based

on the idea of energy balance for a surface layer. To obtain the canopy and soil

temperatures, we begin by retrieving surface radiometric skin temperature (Trs) from the

GOES split window channels, then employ an optimization procedure in conjunction with the

biosphere model to separate the Tc and Ts temperatures. The resulting datasets including

the SRB fluxes, PAR, Trs, Tc, Ts, and the CO2 flux will be calculated over a 4

x 4-km grid at 30-minute time resolution. With these space and time resolutions, the

datasets are expected to play a vital role in various modeling studies to be undertaken

within the LBA-ECO project.

The CO2 flux modeling is to be accomplished by forcing the biosphere model

with the above-derived radiation and temperature variables. The model will incorporate one

of the following physiological parameterizations, depending on available data: (1) scaling

leaf photosynthesis, respiration, and stomatal conductance to the canopy if canopy-level

data are not available; (2) canopy-level parameterization based on light-use-efficiency if

canopy-level data are available. The main results from the modeling efforts will be

detailed time series of the various CO2 flux components at the high space-time


Scientific Importance

This will be the first detailed, quantitative study of CO2 exchange over

large-scale Amaz˘nia. The study is designed to take advantage of recent advances in

physiological and biosphere modeling, and the intrinsic high space-time resolution

capabilities of the GOES Imager. One of the motivations of this study is to determine the

space-time sampling requirements needed to obtain representative estimates of CO2

exchange over the Amazon basin. This process variability is controlled by the

heterogeneity of land cover, changing environmental factors in the planetary boundary

layer, and the highly variable cloud systems and aerosol plumes affecting tropical South


It is our hypothesis that representative carbon budgets cannot be obtained by

monitoring at a limited number of point sites, even if all major land classifications are

sampled, because the processes controlling boundary layer thermodynamics and cloud

formation are not ergodic and thus require frequent sampling to ensure accurate

integration. Currently, we do not quantitatively understand the sources of carbon budget

variability in the rain forest, nor the best strategy to integrate up to basin scales from

the limited amount of information currently on hand. In conducting this study over the

large-scale basin, we expect to determine the major factors controlling CO2

flux exchange variability, which will help shed light on how changing climatic conditions

would be expected to alter carbon sequestering and its release in the Amaz˘nia rain


Value of Results to LBA-ECO

The results are expected to contribute to the LBA-ECO scientific goals in several

ways. With high space-time resolution CO2 fluxes, we shall be able to estimate

the net CO2 fixed by different land covers on different time scales by

integrating various components of the CO2 fluxes in time at selected sites.

Similarly, by integrating the CO2 fluxes in space, we shall be able to estimate

contributions by different land covers to basin-wide carbon accumulation. In addition,

with the high resolution radiation and temperature inputs, which are greatly affected by

cloudiness, boundary layer temperature-moisture, and aerosol concentration, we will be

able to quantify the effects of the variation of these environmental variables on scaling

up to seasonal and basin-wide scales.

Research Team Responsibilities

  • Eric Smith: Radiative Transfer, Biosphere Modeling, Surface Radiation Budget,

    Atmospheric Remote Sensing in Optical-Infrared-Microwave Spectrums, Analysis of Large

    Satellite Datasets

  • Jiujing Gu: SRB/PAR Retrieval from GOES/AVHRR Satellite Measurements, Solar

    Radiative Transfer, Satellite Data Management

  • John Norman: Biophysical Research on Plants and Environment, Carbon Budget,

    Biosphere Modeling, Surface Remote Sensing in Optical-Infrared Spectrums

  • Harry Cooper: Biosphere Modeling, Cloud Processes, Mesoscale Modeling

Last Updated: October 1998

Close Window