Measurement and development of mathematical model for dry deposition of ozone and sulfur dioxide

Abstract

The aims of this work were to conduct an experiment for the measurement of dry deposition velocity of ozone and sulfur dioxide, and to develop a corresponding mathematical model for this process. Three areas in Japan with different types of land used were selected as modeled study: (i) agricultural area, (ii) pine forest and (iii) city (urban) area. In agricultural area, the dry deposition velocities of both ozone and sulfur dioxide in bean fields were higher than in wheat fields. This was because the dry deposition velocity was found to depend linearly on Leaf Area Index (LAI), and LAI was larger in the bean than in the wheat cropping areas. In addition, measurement revealed that dry deposition velocity was independent of wind speed and air temperature, but depended on solar radiation and canopy conductance. The relationship between dry deposition velocity and solar radiation was best described by a logarithmic function, whilst a linear function wasfor the dependency of dry deposition velocity and canopy conductance. The empirical model was found to give good estimates of dry deposition velocity with approximately 30 to 40% deviation from measurement compared with 90-150% from the Wesely model. A modified Wesely model (with correlation coefficient) was found to give a slightly better results than the Wesely model (with 30-80% error), but still could not give the same level of accuracy as the empirical model. The complicated mechanisms of chemical and biochemical reactions were believed to be the main factors that prevented accurate measurement of dry deposition velocity in the pine forest. In this area, some pollutants could emit to atmosphere instead of deposit onto the earth surface. This caused a variation in the ground-level concentrations of pollutants and made the measurement of the dry deposition velocity difficult. However, the relationship between dry deposition velocity and canopy conductance in certain periods of summer time, where weather conditions were relatively stable, could still be established. The associated errors from the Wesely model in predicting ozone deposition velocity was found to be quite high, 50-350%. The introduction of correlation coefficients into the Wesely model could bring the error down to 40-50% in summer period (June-July). In the case of sulfur dioxide, it was not possible to establish any dry deposition velocity model. The study in the city area focused only on dry deposition velocity of ozone. No mathematical relationships between the dry deposition velocity and meteorological factors in any form could be formulated. Therefore the empirical model could not be established. The modified Wesely model with appropriate correlation coefficients were found to give good estimates of dry deposition velocity during summer season.