Predicting ground boom spray drift

Multiple regression procedures were used to develop models to predict spray drift from ground, boom sprayers. These models were developed to study the effect of independent variables on drift deposits. The Missouri model was developed from single nozzle tests and contained six significant variables. Spray drift deposits were most strongly related to three variables (i.e., common lograrithm of corrected downwind distance, wind speed, and nozzle he ight; in that order). Neither volume median diameter or the '% of the spray , volume less than or equal to 105 mu m' were significantly related to drif t deposits. The Illinois model (i.e., developed from six nozzle spray boom tests) included four significant variables. The most important var0iable fo r this model was the 'common logarithm of the corrected downwind distance'. The common logarithm of the corrected downwind distance, nozzle pressure, and dry bulb temperature were included in both of the models. The common lo garithm of the corrected downwind distance was clearly the most import vari able in both models as indicated by sensitivity analyses, The results from the model verifications indicated that the predicted and measured deposits, were in very good agreement.