A Geometric Ultraviolet-B Radiation Transfer Model Applied to Vegetation Canopies

Many radiative transfer models for the shortwave band have been developed and used to understand and The decrease in stratospheric ozone (O3 ) has prompted continued simulate the radiation environment of vegetative canoefforts to assess the potential damage to plant and animal life due to enhanced levels of solar ultraviolet (UV)-B (280–320 nm) radiation. pies. Smith (1983), Goel (1988), and Myneni et al. (1989) The objective of this study was to develop and evaluate an analytical have reviewed these models and studies. Most of these model to simulate the UV-B irradiance loading on horizontal belowradiative transfer models apply to homogeneous canocanopy surfaces, as influenced by vegetation. The UV-B irradiance pies of a large horizontal extent and are one-dimenabove canopy and transmitted to below-canopy points was measured sional (1-D) models (deWit, 1965; Monsi and Saeki, in a widely spaced orchard and in a closely spaced maize (Zea mays 1953; Cowan, 1968). Many important canopies, howL.) crop during cloud-free days, with solar zenith angle ranging from ever, are extremely variable spatially and cannot be 20 to 80 . The sky view fraction was typically 0.59 for the orchard treated by 1-D models. For instance, tree canopies often and 0.28 for the maize canopy. Transmitted irradiance fractions were have large natural openings between crowns, incomsimulated and compared to measured fractions. Measured and simuplete row crops have big spaces between rows of vegetalated values of UV-B canopy transmittance generally agreed well both for points in locations shaded by plant crowns and for points below tion, and urban scenes are complex three-dimensional the top of the canopy that were not shaded. The model had mean (3-D) arrangements of trees and buildings. bias errors of 0.04 and 0.03 for the orchard and maize canopies respecSimulation of UV-B irradiance above and in canopies tively, and the root mean squared error of the model was 0.08 for differs from total shortwave or photosynthetically active orchard and 0.06 for maize. The model can serve as a much-needed radiation in that the fraction of global irradiance retool to examine UV-B irradiance loading of organisms below tree ceived as diffuse irradiance from the sky is much larger, canopies and of sensitive plant surfaces in and below tree and vegetafrequently exceeding 50% for midlatitudes during much tion canopies. of the day. Consequently, particular effort must be made in the modeling treatment of the sky-diffuse irradiance. Simulation of the UV-B irradiance on surfaces below T has been growing concern about the possior in the canopy also requires detailed knowledge of ble impact of ozone layer depletion because the UV-B irradiance penetration of the direct and diffuse stratospheric ozone column is one of the primary attenuirradiance through the canopy. Surfaces of potentially ators of solar ultraviolet (UV)-B radiation (280–320 UV-B-sensitive plant parts (like young leaves and inflonm). A decrease in this ozone column would lead to rescences) are frequently present in canopies before increases in UV-B irradiances reaching the earth’s surcanopy closure or in the higher part of the canopy where face. The most important wavelengths for assessing poit is relatively open. Open canopies typically have large tential plant damage due to increased UV radiation are discontinuities that give large views of the sky and its in the UV-B band (Caldwell, 1971; Caldwell et al., 1998; diffuse irradiance (a large portion of the total UV-B Madronich et al., 1998). The effect of UV-B enhanceirradiance) and that also provide paths for the transmisments on plants includes reduction in grain yield, altersion of direct irradiance under the appropriate sun ation in species competition, decrease in photosynthetic angle. The 1-D models assume a homogeneous canopy activity, susceptibility to disease, and changes in plant that cannot simulate an open canopy where there is structure and pigmentation (Tevini and Teramura, 1989; large spatial variation in leaf area in the horizontal plane Bornman 1989; Teramura and Sullivan, 1991). Some and important anisotropic distributions of the incident plant species show sensitivity to present levels of UV-B irradiance at the canopy top as is the case with UV irradiance while others are apparently unaffected by sky radiance distributions. An advanced 3-D radiation rather massive UV enhancements (Becwar et al., 1982). model that considers anisotropic sky radiance penetratTo make matters more complicated, there are reports ing through heterogeneous canopies (such as row crops of equally large response differences among cultivars before canopy closure) is needed to evaluate UV-B of a species (Biggs et al., 1981; Teramura and Murali, irradiance loading in many plant canopies. Such a 3-D 1986). About two-thirds of some 300 species and cultimodel is most useful for canopies that contain dense vars tested appear to be susceptible to damage from grouping of leaves within subcanopies, or crowns, that increased UV-B irradiance. are widely separated. When dimensionality increases in radiation models, more canopy structure information is needed as input to the model. W. Gao, USDA UV-B Radiation Monitoring and Res. Progr., Coop. Inst. for Res. in the Atmos., and J.R. Slusser, USDA UV-B Radiation This paper describes the testing of a 3-D UV radiation Monitoring and Research Progr., Nat. Resour. Ecol. Lab., Colorado transfer (UVRT) model that can be used to calculate State Univ., Fort Collins, CO 80523; R.H. Grant, Dep. of Agron., canopy transmittance (Tcanopy, irradiance below canopy/ Purdue Univ., West Lafayette, IN 47906; and G.M. Heisler, Northeastern Res. Stn., USDA Forest Serv., Syracuse, NY 13210. Received 21 Abbreviations: 1-D, one dimensional; 3-D, three dimensional; MBE, Aug. 2000. Corresponding author ([email protected]). mean bias error; RMSE, root mean squared error; Tcanopy, canopy transmittance; UV, ultraviolet; UVRT, ultraviolet radiation transfer. Published in Agron. J. 94:475–482 (2002).