Characterizing animal anatomy and internal composition for electromagnetic modelling in radar entomology

The use of radar as an observational tool in entomological studies has a long history, and ongoing advances in operational radar networks and radio-frequency technology hold promise for advances in applications such as aerial insect detection, identification and quantification. Realizing this potential requires increasingly sophisticated characterizations of radio-scattering signatures for a broad set of insect taxa, including variability in probing radar wavelength, polarization and aspect angle. Although this task has traditionally been approached through laboratory measurement of radar cross-sections, the effort required to create a comprehensive specimen-based library of scattering signatures would be prohibitive. As an alternative, we investigate the performance of electromagnetic modelling for creating such a database, focusing particularly on the influence of geometric and dielectric model properties on the accuracy of synthesized scattering signatures. We use a published database which includes geometric size measurements and laboratory-measured radar cross-sections for 194 insect specimens. The insect anatomy and body composition were emulated using six different models, and radar cross-sections of each model were obtained through electromagnetic modelling and compared with the original laboratory measurements. Of the models tested, the prolate ellipsoid with an internal dielectric of homogenized chitin and hemolymph mixture best replicates the measurements, providing an appropriate technique for further modelling efforts.