Hyperspectral lidar monitors insect diversity in southern Sweden

13 Jul 2023

NKT Photonics and Lund University platform discriminates between free-flying species.

Monitoring insect populations is an important part of understanding natural ecosystems, especially when those populations appear to be declining through pesticide use or climate change.

One basic method involves trapping live insects for identification, a relatively time consuming and expensive process, but more advanced photonics techniques are now being applied to the task.

Denmark's NKT Photonics and Lund University in Sweden have studied how a hyperspectral lidar platform can be used to count insects, measure the frequency of their wing beats, and resolve coherent scattering from their wings to distinguish different species.

Published in Advanced Science, the findings "could revolutionize insect surveillance," according to the project.

The NKT/Lund platform employs a technique termed elastic hyperspectral Scheimpflug lidar (EHSL). The Scheimpflug principle, originally derived from aerial photography, involves the lens and detector being tilted with respect to each other, increasing the focal depth that the system can achieve.

In a lidar system with the image sensor orientated according to the Scheimpflug principle, backscatter from a laser beam directed into the atmosphere can arrive on a tilted sensor with all backscattered echoes in focus simultaneously, a route to theoretically infinite focal depth with a large optical aperture.

For trials in southern Sweden, NKT and Lund University arranged its device so that nocturnal insects passed through a stationary laser beam that recorded the distance to the insect and the reflected spectrum.

"Insect wings are thin membranes," said the project. "When the laser light hits the insect wing, some light is reflected from the first surface, and some is reflected from the second surface after passing through the wing. This creates patterns called wing interference patterns."

Different insect species have specific wingbeat frequencies and wing interference patterns, and capturing data from both characteristic properties could make it possible to differentiate hundreds of free-flying insect species in their natural habitat.

Remote measurements with nanometer precision

The EHSL device employed continuous wave Scheimpflug lidar alongside an inelastic hyperspectral version of the same architecture, a version which had previously been developed and employed in fluorescence mode. Adding spectroscopy to lidar allows retrieval of tiny features over far distances, according to the project team.

"Wing flashes have previously been retrieved remotely by polarization lidar and dual-band lidar," said the project in its published paper. "But until now no remote instrumentation has been able to capture such flashes, resolve them spectrally and uniquely determine the wing thickness."

The project demonstrated that its method successfully captured signals from insects in flight when it was deployed in-field at night, and offered the possibility of remotely acquiring insect wing thicknesses with nanometer precision.

"Until now, lidar techniques for detecting free-flying insects have relied mostly on frequency analysis of wingbeat patterns, which requires multiple wingbeats during the beam transit," said the team. "The EHSL technique could in principle determine the wing thickness from a single microsecond flash."

Other forms of optical remote sensing or environmental monitoring could also make use of devices based on the same principle. The project noted that hyperspectral lidar for vegetation canopies could improve tree species classification and report on leaf moisture, fertilization or internal leaf structures, while atmospheric sensing of greenhouse gases and other molecules could also be possible.