A system for monitoring SO2 dry deposition fluxes to low vegetation
The improvements of the monitoring system which has been used since 1987 are discussed. The performance of the system in due form, and the reliability and accuracy of the fluxes were evaluated by comparing values to those obtained with a similar dry deposition system for forests, which was tested next to the low vegetation system during the first half of 1992 at the heathland site Elspeetsche Veld.
The evaluation showed that the total uncertainty in the hourly average SO2 dry deposition fluxes derived from the measurements is 25% for stable, 50% for neutral (which occur most of the time) and 300% for unstable conditions, while the total systematic error is in the range of 10 to 20%. Since fluxes cannot always be directly calculated from measured data due to limitations of the gradient method, a parameterization of the SO2 surface resistance Rc was developed based on measurement results and on literature.
It was found that 44% of the variance in parameterized versus measured vd values was accounted for, with no systematic differences. By combining the surface resistance parameterization for rejected periods (periods, which did not fulfil the demands for application of the gradient method) with measured values during selected periods, yearly average SO2 fluxes and deposition velocities can be obtained with uncertainties in the order of 20-30%.
At Elspeetsche Veld, the yearly average flux of SO2 was found to be 250+-100 mole ha-1 y-1 and the deposition velocity 1.0+-0.1 cm s-1. From June 1994 until January 1996, the system was operated at the semi-natural grassland site Oostvaarderplassen. Unfortunately, most of the data were lost due to frequent technical failure of the power generator used at this site. Using the Rc parameterization for this site, an average deposition velocity of 0.7+-0.6 cm s-1 was estimated from the technically validated measurements. However, since these measurements covered only 15% of the total period, this deposition velocity cannot be considered a yearly average. In future applications, the system should therefore always be connected to the mains electricity supply.
|Author(s)||Mennen MG ; Erisman JW ; Putten EM van ; Hellemond J van ; Hogenkamp JEM ; Zwart HJMA ; Elzakker BG van|