Dealing with ultra-low spray deposits in experiments

In spray drift experiments, measuring spray deposits far downwind may yield deposits values that approach or go below the level of detection. Particularly, when deposits are corrected for background concentrations of whatever origin, in principle values at or even below zero may result. A conventional way to deal with those data is to adjust them to zero or a ‘very low but positive’ value. Alternatively, zero and negative deposits are simply removed from the data set. In either way, such adjustments may affect the results in a way the researcher may not have foreseen or intended. In this study different common methods of dealing with ultra-low deposition values are compared. It is discussed how these methods may affect the results and their interpretation. An alternative method is introduced and discussed which may reduce the negative implications of the common methods. The study is supported by various computational examples.