Modeling Carbon Sequestration over the Large-Scale Amazon Basin, Aided by Satellite Observations. Part I: Wet- and Dry-Season Surface Radiation Budget Flux and Precipitation Variability Based on GOES Retrievals

Abstract

In this first part of a two-part investigation, large-scale Geostationary Operational Environmental Satellite (GOES) analyses over the Amazônia region have been carried out for March and October of 1999 to provide detailed information on surface radiation budget (SRB) and precipitation variability. SRB fluxes and rainfall are the two foremost cloud-modulated control variables that affect land surface processes, and they require specification at space-time resolutions concomitant with the changing cloud field to represent adequately the complex coupling of energy, water, and carbon budgets. These processes ultimately determine the relative variations in carbon sequestration and carbon dioxide release within a forest ecosystem. SRB and precipitation retrieval algorithms using GOES imager measurements are used to retrieve surface downward radiation and surface rain rates at high space-time resolutions for large-scale carbon budget modeling applications in conjunction with the Large-Scale Biosphere-Atmosphere Experiment in Amazônia. To validate the retrieval algorithms, instantaneous estimates of SRB fluxes and rain rates over 8 km ϫ 8 km areas were compared with 30-min-averaged surface measurements obtained from tower sites located near Ji-Paraná and Manaus in the states of Rondônia and Amazonas, respectively. Because of large aerosol concentrations originating from biomass burning during the dry season (i.e., September and October for purposes of this analysis), an aerosol index from the Total Ozone Mapping Spectrometer is used in the solar radiation retrieval algorithm. The validation comparisons indicate that bias errors for incoming total solar, photosynthetically active radiation (PAR), and infrared flux retrievals are under 4%, 6%, and 3% of the mean values, respectively. Precision errors at the analyzed spacetime scales are on the order of 20%, 20%, and 5%. The visible and infrared satellite measurements used for precipitation retrieval do not directly detect rainfall processes, and yet they are responsive to the rapidly changing cloud fields, which are directly associated with precipitation life cycles over the Amazon basin. In conducting the validation analysis at high space-time scales, the comparisons indicate systematic bias uncertainties on the order of 25%. These uncertainties are comparable to published values from an independent assessment of bias uncertainties inherent to the current highest-quality satellite retrievals, that is, from the Tropical Rainfall Measuring Mission. Because precipitation is a weak direct control on photosynthesis for the Amazon ecosystem, that is, photosynthesis is dominated by the strong diurnal controls of incoming PAR and ambient air-canopy temperatures, such uncertainties are tolerable. By the same token, precipitation is a strong control on soil thermal properties and carbon respiration through soil moisture, but the latter is a time-integrating variable and thus inhibits introduction of modeling errors caused by random errors in the precipitation forcing. The investig...

A coupled 1-D atmosphere and 3-D canopy radiative transfer model for canopy reflectance, light environment, and photosynthesis simulation in a heterogeneous landscape

“…The spectral composition of diffuse PAR also changes (Dye, 2004;Min, 2005). These changes can be partially or fully considered in existing atmospheric radiative transfer models for PAR estimation (Frouin et al, 1989;Eck & Dye, 1991;Gu et al, 2004;Kobayashi et al, 2004;Liang et al, 2006;Pinker & Laszlo, 1992;Van Laake & Sanchez-Azofeifa, 2004). However, due to their simplicity, these approaches do not take into account the angular variability of the incident diffuse PAR, while incident diffuse light has directionality due to the strong forward scattering properties in aerosols and clouds.…”

Estimation of incident photosynthetically active radiation from Moderate Resolution Imaging Spectrometer data

“…Satellite observations acquired at the top of the atmosphere (TOA) contain information of both atmosphere and surface. The current methods [e.g., Carder et al , 2003; Frouin et al , 2000; Gu et al , 2004; Pinker and Laszlo , 1992; Pinker et al , 2003] assume that either the atmospheric information is available from other sources (e.g., the ISCCP PAR product with atmospheric climatology data as input) or that water surface reflectances are known (e.g., SeaWiFS and MODIS PAR products over ocean).…”