Harnessing atmospheric dispersion models for predicting long-distance dispersal of small pest insects

• Atmospheric dispersion models (ADMs) are designed for abiotic matter dispersion. • Our method allows existing ADMs to be parameterized for small insect species. • Parameterization is made based on insect biological traits and flight aerodynamics. • Our method enables researchers to adapt existing ADMs to study insect dispersal. • Our method enhances ADMs’ accuracy and realism in simulating insect dispersal. Atmospheric dispersion models (ADMs) are state-of-the-art numerical simulation models for investigating atmospheric transport processes, thereby being often used to study microinsect (small insect) pest invasions via aerial pathways. However, ADMs are designed for non-living matter (e.g., gaseous and particulate pollutants). These tiny aerosols are usually represented by computational particles in 3D dispersion simulations, where they are passively dispersed. In contrast, living organisms such as airborne microinsects transported downwind have relatively large bodies. Their aerodynamic capacities depend largely on their biological characteristics (e.g., anatomical, behavioural, morphological, and physiological adaptations). Directly using ADMs without considering these characteristics could compromise their simulation realism and accuracy in predicting the long-distance atmospheric dispersal (LAD) of microinsects. To address such issues, we propose a generic method to enhance the simulation realism and accuracy by taking insect biological traits and flight aerodynamics into account. The method first allows insect species to be precisely defined and quantified as aerodynamically small enough to be passively transported in ADMs. Then it translates this species into their aerodynamic equivalence of computational particles, which can be readily parameterised for ADMs. To demonstrate and validate the method, the LAD of bird cherry-oat aphids, Rhopalosiphum padi (Linnaeus) (Hemiptera: Aphididae), from an area of irrigated pastures to the Wheatbelt at a large spatial scale in the south-west of Western Australia was simulated with 15 million aphids released. Compared with other conventional setting approaches, only our method successfully enabled ADMs to simultaneously reproduce multiple field observations, which led to biologically realistic and robust results that are essential for biosecurity management.