Influence of a Young Pinus radiata Canopy on Aerial Spray Drift

Abstract

Abstract. An experiment was undertaken to test the hypothesis that AGDISP, a mechanistic model that simulates the landing position of droplets aerially released for pesticide application, overestimates downwind airborne spray concentrations by not accounting for the turbulence generated within the canopy sublayer. AGDISP is used by pesticide regulators in many parts of the world to quantify exposure from aerial pesticide spraying as part of an assessment of risks to public health and the environment. Using a novel design, airborne spray drift was measured in a 3.2 m high radiata pine plantation, 65 m downwind from the aerial spray flight line, using a small circular sampling array. The spray sampling array consisted of a central 10 m high mast plus a set of 16 secondary masts distributed evenly around a circle with a radius of 5 m from the central mast. All spray applications were made with a Bell Jet Ranger using a spray mix of water plus a fluorescent tracer and a droplet spectrum with a volume median diameter of 101 µm. A single treatment comprised six passes along the flight line, and there were 28 replications. Airborne spray was sampled with both rotorods and artificial foliage placed at four heights on the central mast, ranging from approximately half canopy height to 3 times canopy height. The secondary towers sampled airborne spray using the same collectors placed at just above canopy height and at half canopy height. Other measurements included leaf area distribution, prevailing meteorological conditions, and turbulence using sonic anemometers. Backscatter LiDAR was used in an attempt to quantify the height and density of the spray cloud in the vicinity of the central mast. Rotorod and artificial foliage deposition values were highly correlated, giving confidence in the sampling approach. Neither wind speed nor release height had a significant effect on airborne spray flux in the 0 to 10 m zone, but results were complicated by a spurious negative correlation between wind speed and spray release height. The lack of dependence of deposition on wind speed in this experiment is surprising but is explained both by the variance in the data as well as the physics of small droplet movement in the atmosphere. At the individual sample level, measurements of airborne spray using either rotorods or artificial foliage were poorly correlated with AGDISP predictions. However, when data were averaged across replications, there was a very strong correlation between measurements and AGDISP data, even though in absolute terms, and as hypothesized, AGDISP significantly overpredicted the quantity of airborne spray. As indicated from a comparison with measured deposition, the backscatter LiDAR appeared effective in describing the spray cloud profile, and it suggested a total spray cloud profile that was more complex than the AGDISP model prediction. Keywords: Aerial application, AGDISP, Canopy deposition, Pesticides, Spray drift.