Development and Application of Disjunct Eddy Covariance Techniques for the Measurement and Interpretation of Fluxes of Volatile Organic Compounds from Urban and Rural Canopies.

Langford, Ben (2008) Development and Application of Disjunct Eddy Covariance Techniques for the Measurement and Interpretation of Fluxes of Volatile Organic Compounds from Urban and Rural Canopies. PhD thesis, UNSPECIFIED.

PDF (11003731.pdf)
11003731.pdf - Published Version
Available under License Creative Commons Attribution-NoDerivs.
Download (14MB)

Abstract

Two disjunct eddy covariance systems for the measurement of volatile organic compound (VOC) fluxes were developed. The first, disjunct eddy covariance (DEC), was validated against the standard eddy covariance (EC) technique, in a study of CO2 and H2O fluxes from a grassland field (Easter Bush, Edinburgh, Scotland). The comparison convincingly showed fluxes measured by the DEC technique to be comparable to those measured using the EC technique. A second, simplified approach, virtual disjunct eddy covariance (vDEC), was developed and compared against standard DEC during the CityFlux project, where measurements of VOC fluxes were made from Portland Tower in Manchester. Averaged daily fluxes measured by the vDEC system typically ranged between 19 and 90 microg m-2 h-1 for individual VOC species and were comparable to those measured by the DEC system, but were typically 19% higher than the latter. The discrepancies between the two methods were thought to relate to both the reduced response time of the DEC system which attenuated higher frequency flux contributions and the high level of noise in the covariance function which may have led to a systematic overestimation of the flux. The vDEC technique was subsequently deployed on the Telecom Tower in central London to give very detailed flux information on seven VOC species. Individual average fluxes ranged between 5 and 100 microg m-2 h-1 and were well correlated with traffic density. Fluxes of benzene were extrapolated to give an annual emission estimate for the city, which was found to be 1. 8 times lower than that suggested by the National Atmospheric Emission Inventory. Finally, two vDEC systems, one using a high sensitivity (HS) proton transfer reaction mass spectrometer (PTR-MS) and the other a standard model (Std), were used alongside each other to measure biogenic VOC fluxes from macchia vegetation at the Castelporziano nature reserve near Rome, Italy. The two systems compared well, although the HS system appeared to give fluxes with greater amplitude than the Std model. This highlighted the importance of the allocation of correct lag times when using vDEC, particularly at night. Fluxes of isoprene and monoterpenes were compared with the Guenther algorithm of 1995 and showed excellent agreement between the modelled and measured values. The results presented in this study have convincingly demonstrated the capacity of the DEC and vDEC techniques to give very detailed VOC flux information over a range of non-ideal canopies, which can be used to both validate and constrain "bottom-up" style emission inventories.