WTDISP – adapting a Lagrangian ground sprayer model using wind tunnel data

The ground boom model (AGriculture DISPersion) or AGDISP calculates the deposition of agrichemicals both in the area sprayed and more importantly downwind on neighbouring properties. However modelling the breakup and evaporation of the spray droplet cloud over the first few metres is very difficult. Wind Tunnel Dispersion or WTDISP arose to overcome this difficulty. Measurements of droplet spectra and fluxes were made 2 m downwind of several nozzles in a wind tunnel. WTDISP originated as this source was input into the frame work of the ground boom model AGDISP. AGDISP calculates the deposition by ensemble averaging the droplet paths to the ground and uses a statistical distribution of the likely paths about these averages to calculate the deposition on the ground. The nozzles selected for the wind tunnel were those used in the field study trials undertaken by Wolf and Caldwell (2001) so that WTDISP could be used with this study as a basis for comparison with AGDISP. Measurements were undertaken at 4.5 m/s wind speed, at 0.1 m vertical intervals from 0.1 m to 0.6 m and at 0.08 m horizontal intervals from the centre of the nozzle at each vertical level, (Hewitt, 2008). The analysis calculated the total flux at each level, together with the droplet spectra, to input into WTDISP in order to calculate the deposition downwind for one nozzle. This flux was matched to the field situation to give the deposition. Figure 1 shows the result for the first trial of the Wolf & Caldwell (2001) data.(Graph Presented) The results show a considerable improvement for WTDISP over AGDISP for this trial. Overall the 5 trials analysed, WTDISP comparison with field measurements halved the difference between (and also the standard deviation) of the modelled/measured ratio. The paper below also outlines possible ways to improve the technique to model field and wind tunnel measurements, including turbulence scaling effects, the orientation of the nozzle fan to the wind direction and field measurements. It also discusses the drop size distribution from both models at 2 m downwind leading to improvements to the evaporation/breakup algorithms used in AGDISP and gives recommendations for experiments to improve the model results.