Wind tunnel experiment and regression model for spray drift

Abstract: With greater environmental awareness, the movement of pesticides within and off of a spray target area is a critical public concern. Ideally, all of the material applied should be deposited within the targeted swath on the intended pest or plant. But realistically, a portion of the spray remains airborne and is carried downwind to non-target areas. Airborne spray leaving the targeted area reduces the applied dosage, and could cause damage to neighboring plant and water source or other detrimental environmental impacts. To study the influences of nozzle type, spray mixture and wind speed on spray drift, experiments were conducted using a wind tunnel. Spray drift risk was assessed by adding a tracer to the spray mixture and measuring the quantities of spray deposited downwind from the nozzle on horizontal polythene lines with 2 mm diameter perpendicular to the wind direction in a vertical and a horizontal array. At a distance of 2 m downwind from the static nozzle, five collector lines (V1 to V5) were positioned one above the other at the spacing of 0.1 m to provide an estimate of the spray still airborne through this vertical profile. An additional five sampling collector strings (H1 to H5) were placed in a horizontal array with one-meter horizontal spacing at 0.1 m height to determine the fallout volumes and gradients of the spray from 2 to 6 m downwind. A water-soluble fluorescent tracer was dissolved into tap water as the spray liquid, and after the experiments, the collecting lines were washed with deionized water to measure deposit and drift. The results indicated that deposits on sampling collector decreased with increased vertical elevation and horizontal distance. Average fallout and airborne deposit resulting from the different spray applications were shown in the paper. These figures showed the expected fallout and airborne profiles for all tested nozzle types and sizes. The highest fallout deposits were measured at a position closest to the nozzle (H1) with a systematic decrease with the distance from the nozzle. The highest airborne deposits were found at the lowest sampling collector (V1) with a systematic decrease with increasing height above the wind tunnel floor. Airborne spray drift was affected by wind speed. At all sample positions, deposits on collectors were reduced at lower wind speed. Nozzle's structure was also found to influence droplet's size, so injector/pre-orifice nozzle produced coarser droplets and reduced spray drift. The amount of spray recovered is based on the amount of active ingredient of spray mixture within each droplet rather than the total droplet volume. On that basis, a multiple non-linear model for statistical drift prediction including four independent, non-correlated variables (target distance, wind speed, nozzle type and chemical type) was established. The regression model provided a drift evaluation approach, and it was important in the interpretation of wind tunnel data for different nozzle types, chemical types and sampling methodologies.