This article reviews approaches to analytical pesticide deposition modeling for aerial, ground boom, and orchard airblast sprayer applications, and describes modeling approaches including the Lagrangian model commonly used for aerial spraying, which tracks the mean position and standard deviation of spray material released into the ambient environment. This review also describes the approaches used in mechanistic ground sprayer modeling, including near-nozzle effects and spray sheet obstruction of the ambient flow. The random-walk approach for ground spraying and the significant computational fluid dynamics (CFD) problem associated with modeling orchard airblast spraying are also described. These mechanistic approaches predict the spray deposition downwind from an application area for any set of initial conditions, and can recover results consistent with field datasets. Incorporation of spray deposition modeling into a Geographic Information System (GIS) platform is also discussed.

The AGDISP Aerial Spray Simulation Model is used to predict the deposition of spray material released from an aircraft. The prediction is based on a well-defined set of input parameter values (e.g., release height, and droplet size) as well as constant data (e.g., aircraft and nozzle type). But, for a given deposition, what are the optimal parameter values? We use the popular Genetic Algorithm to heuristically search for an optimal or near-optimal set of input parameters needed to achieve a certain aerial spray deposition.

In this paper, we describe several genetic algorithm methods to deal with the spray parameter optimization problem, and compare them with the original heuristic method SAGA.

AGDISP (Aerial Spray Simulation Model) is used to predict the deposition of spray material released from an aircraft. Determining the optimal input values to AGDISP in order to produce a desired spray material deposition is extremely difficult (NP hard). SAGA, an intelligent optimization method based on the simple genetic algorithm, was developed to solve this problem. In this paper, we apply several nature inspired heuristics to this problem. The first method still uses the genetic algorithm, but changes its type, selection method, crossover and mutation operator. The second method applies a neural network to improve the initial population, crossover and mutation. The third method uses GADO, a general-purpose approach to solving the parametric design problem. The fourth method applies simulated annealing to this problem. Finally, we compare their performance with SAGA and discuss their applications to the aerial spray deposition problem.

The Spray Treatment Planner is an intelligent decision support system for productivity and efficiency evaluation in an aerial spray treatment project. It can be used as a tool to plan the schedule for spraying pesticides aerially. STP schedules the spraying operation of selected blocks from selected airports using single or multiple aircraft(s). The scheduling is done to maximize the spray efficiency and spray productivity while minimizing the total time and distance flown. It uses heuristics to obtain a near optimal solution. Aerial spray applicators can use STP to estimate total time required for a treatment operation. Forest managers can use STP to determine an efficient and productive treatment plan.

Determining the parameter value settings to use as input to AGDISP (aerial spray simulation model) in order to produce a desired spray material deposition is considered an instance of the parametric design problem. SAGA (spray advisor using genetic algorithm) was developed to solve this problem. In this paper, we describe several approaches to improve the performance of SAGA. First, we replace the original generational genetic algorithm with a steady-state genetic algorithm, and the original roulette wheel selection with tournament selection. We call the new system SAGA2. Second, we apply a neural network to improve the initial population, crossover, and mutation. We call this version SAGA2NN. Then, we apply GADO, a general-purpose approach to solving the parametric design problem. The integrated GADO version is called SAGADO. Finally, we compare the performance of SAGA, SAGA2, SAGA2NN and SAGADO.

The United States Department of Agriculture — Forest Service (USDA-FS) has been involved in the development of computer models to simulate deposition from aerial pesticide spraying since the early 1970s. Originally, this work was driven by the need to improve the percentage of aerially sprayed material that actually deposited on a target area. The amount of on-target deposition is a primary factor in determining the level of pest control achieved. A second focus of this modeling work that has become the objective in much of the recent work is to use modeling to determine the amount of sprayed material that does not land on the target area and is defined as “drift”. It is assumed that drift causes unintended environmental consequences and is a form of environmental pollution.

Improving aerial spray application results is a major concern for the USDA Forest Service and Environmental Protection Agency. The AGDISP Aerial Spray Simulation Model is used to predict the deposition of spray material released from an aircraft. The prediction is based on a well-defined set of input parameter values (e.g., release height, and droplet size) as well as constant data (e.g., aircraft and nozzle type). But, for a given deposition, what are the optimal parameter values? This problem is considered to be a parametric design problem or more generally a configuration problem. Attempting to optimize a configuration based on some set of constraints is known to be extremely difficult (NP-Hard). We use the popular Genetic Algorithm to heuristically search for an optimal or near-optimal set of input parameters needed to achieve a certain aerial spray deposition. Having this knowledge can benefit forest managers substantially, especially regarding such issues as cost, environmental safety, and forest

The AGDISP Aerial Spray Simulation Model is used to predict the deposition of spray material released from an aircraft. The prediction is based on a well-defined set of input parameter values (e.g., release height. and droplet size) as well as constant data (e.g., air craft and nozzle type). But, for a given deposition, what are the optimal parameter values? We use the popular Genetic Algorithm to heuristically search for an optimal or near-optimal set of input parameters needed to achieve a certain aerial spray deposition. Having this knowledge can benefit forest managers substantially, especially regarding such issues as cost, environmental safety and forest treatment accuracy.

GypsEX provides knowledge-based decision support for two aspects of aerial application of pesticides against gypsy moth: calibration and spray timing. Calibration provides advice on setting up an aircraft's spray system for a desired flowrate of pesticide. Spray timing determines the optimal date for aerial application of Bacillus thuringiensis based on weather forecasts, simulated gypsy moth-host tree phenology, and expert heuristics. Knowledge diagrams for the heuristics of each module were developed from interviews with experts and technical references, then converted into computer code with a frame-based expert system shell. Prototype versions of the system were used for further knowledge elicitation both in the office and under field conditions. The Calibration module of GypsEX was validated in the field through comparison of its advice and logic with that of three experts on the calibration of aerial application systems. Future refinements and the incorporation of GypsEX into a decision support system for several aspects of gypsy moth management are discussed.

The AGDISP Aerial Spray Simulation Model is used to predict the deposition of spray material released from an aircraft. The prediction is based on a well-defined set of input parameter values (e.g., release height, and droplet size) as well as constant data (e.g., aircraft and nozzle type). But, for a given deposition, what are the optimal parameter values? This problem is considered to be a parametric design problem or more generally a configuration problem. Attempting to optimize a configuration based on some set of constraints is known to be extremely difficult (NP-Hard). We use the popular Genetic Algorithm to heuristically search for an optimal or near-optimal set of input parameters needed to achieve a certain aerial spray deposition. Having this knowledge can benefit forest managers substantially, especially regarding such issues as cost, environmental safety, and forest treatment accuracy.