A new approach for remote sensing of canopy-absorbed photosynthetically active radiation .1. Total surface absorption

Thr canopy-absorbed photosynthstically active radiation APAR(CAN) is the solar energy consumed in the canopy photosynthetic process. Due to the difficulty of acquiring extensive ground-based observations, increasing efforts are being devoted to estimate APAR(CAN) from optical satellite measurements. So far, APAR(CAN) has been obtained from the downwelling PAR at the surface (SFC), PAR(SFC)down arrow, and the fraction of PAR absorbed by a canopy, FPAR. This study proposes a new approach which defines APAR(CAN) as the product of APAR(SFC) and RPAR. APAR(SFC) is the total PAR absorbed by all surface materials including canopy, soil, litter, etc., while RPAR is the ratio of the PAR absorbed by the green canopy only, to APAR(SFC). The advantage of this approach is that APAR(SFC) can be determined more accurately and readily than PAR(SFC)down arrow, while the determination of RPAR is as accurate as that of FPAR with the same difficulties. The whole approach is introduced in two parts. Part I, as presented in this article, deals with the retrieval of APAR(SFC). Using a complex atmospheric radiative transfer model, APAR(SFC) is related to the upwelling PAR reflected at the top of the atmosphere (TOA), PAR(TOA)up arrow. The relationship is independent of cloud parameters and surface conditions, and moderately dependent on ozone amount and aerosol optical properties. A parameterization teas developed to estimate APAR(SFC) from PAR(TOA)up arrow, which is inferred from satellite measurements in the visible bands. Error analyses were made using data from both model simulations and field observations. The parameterization is valid to within 5 W m(-2) compared to the results of detailed radiation model simulations. A preliminary comparison against FIFE ground observations showed a bias error of - 2.7 W m(-2) and a standard error of 21.9 W m(-2) for the instantaneous estimates of APAR(SFC).