Effect of point source emission height on the error of the flux estimation through eddy covariance

A challenge commonly associated with eddy covariance is that real measurement sites are rarely homogeneous. Therefore, scientists had to identify a footprint area or “effective upwind source area sensed by the observation” in order to make sense of the measurements. This led to the development of footprint functions weighting the respective contribution of each element of the surface to the measured vertical flux. A promising use of footprint models would be to extend the use of eddy covariance to quantify point source emissions, such as methane emissions from livestock or emissions from vents in geothermal areas. However, most footprint models consider the source at ground level which may not be a valid hypothesis for all point sources as for example livestock.
The present work aimed to assess the effect of neglecting the source height on the accuracy of the gas emission prediction. The first step of the work focused on the validation of the footprint model FIDES which take in account the source height. The validation was based on experimental measurements with an artificial source at two heights (0.05 & 1.3m) located at two distances (14 & 24m). For both distances and sampling heights the footprint model provided satisfactory results. Then the second step used FIDES model to perform a sensitivity analysis of the flux prediction error in respect with the source height and the distance from the mast. This analysis showed that the error mainly occurs close to the mast. Indeed, below a critical distance if the source height is not known, assuming it at ground level may induce large error on the flux measurements. The critical distance can be expressed in respect with the mast and the source heights. For example, neglecting the source height for a source lower than 0.15 of the mast height leads to relative error smaller than 10% when the point source was located further than 8 times the mast height. These results can be used to design future experiments in order to minimizing the flux measurements error when the height of the source may not be exactly known by sampling within a defined range from the mast in which the measurement errors remains low