Source/sink distributions of heat, water vapour, carbon dioxide and methane in a rice canopy estimated using Lagrangian dispersion analysis

Source distributions for heat, water vapour, CO2 and CH4 within a rice cano py were derived using measured concentration profiles, a prescribed turbule nce field and an inverse Lagrangian analysis of turbulent dispersion of sca lars in plant canopies. Measurements were made during IREX96, an internatio nal rice experiment in Okayama, Japan. Results for the cumulative fluxes of heat, water vapour and CH4 at the canopy top were satisfactory once their respective concentration profiles were smoothed using simple analytic funct ions. According to the inverse analysis, water vapour was emitted relativel y uniformly by each of five equi-spaced layers within the canopy, whereas s ensible heat fluxes were small (<100 W m(-2)) and of either sign. Methane f luxes were predicted to be emitted most strongly in the lower 50% of the ca nopy, as expected from the distribution of micropores along leaves and leaf sheaths, the major pathway for CH4 loss from the soil-crop system. No smoo thing was required for CO2 concentration profiles and the inverse analysis provided close correspondence between the turning point in the concentratio n profile is the changeover from respiration by the soil/paddy water and lo wer canopy to net photosynthesis by the upper canopy. These results could o nly be obtained by including both the near- and far-field contributions of sources to the total concentration profile. Neglect of the near-held contri bution in the inverse analysis led to spurious source distributions. Excell ent agreement was obtained between cumulative fluxes of heat, water vapour, CO2 and CH4 at the top of the canopy from the inverse analysis and direct eddy covariance measurements when the friction velocity u(*)>0.1 ms(-1), an d atmospheric stability was approximately neutral. Nocturnal fluxes of CO2 and CH4 from the inverse method exceeded micrometeorological measurements a bove the canopy by a factor of 2-3 when u(*) < 0.1 m s(-1) and stable atmos pheric conditions prevailed within and above the canopy. Neglect of these s tability effects will lead to an underestimate of the dispersion coefficien ts (dimension of resistances) in the transport model and hence an overestim ate of the fluxes. Further work is required to establish the correct proced ure for incorporating stability effects into the inverse analysis. (C) 2000 Elsevier Science B.V. All rights reserved.