Main content area

Reducing the discrepancies between the Aerodynamic Gradient Method and other micrometeorological approaches for measuring fumigant emissions

Anderson, Ray G., Yates, Scott R., Ashworth, Daniel J., Jenkins, Dennise L., Zhang, Qiaoping
The Science of the total environment 2019 v.687 pp. 392-400
aerodynamics, eddy covariance, emissions, environmental protection, fumigants, fumigation, meteorological parameters, models, pesticides, public health, uncertainty, California
Observations of fumigant and pesticide emissions are needed for multiple public health and environmental protection mandates. The aerodynamic gradient method (ADM) is commonly used to measure fumigant and pesticide emissions. However, the ADM may over estimate emissions compared to other micrometeorological and modeling approaches, which would increase uncertainty over the true flux estimate. Different studies with ADM have also used multiple differing transport functions that relate concentration gradients to emissions. Therefore, we tested different and more recent transport functions to try to correct the anticipated observed higher values with ADM using observations from two sites in California, USA. We evaluated different transport functions against eddy covariance observations and found that using the functions developed by Högström (1996) corrected the ADM values to be in line with other observational methods. For the Fresno experiment, cumulative emission masses from the ADM- Högström functions were within 7% of other approaches while the Pruitt function was >15% higher. Applying the Högström functions to a series of previous fumigation experiments in California saw reductions in the ADM observations of >25% for cumulative mass emissions. The results indicate that the Högström functions should be used for future ADM experiments in the absence of more robust transport factors for local meteorological conditions. The results also illustrate how previous ADM observations could be corrected to reduce uncertainty in flux emissions estimates.