We present a dual-wavelength polarimetric measurement method to distinguish species and sexes of disease transmitting mosquitoes in flight. By measuring co- and de-polarized backscattered light at 808 and 1550 nm, the degree of linear polarization, wingbeat frequency, reflectance, spectral ratio and glossiness of mosquitoes can be retrieved. Body and wing contributions to these signals can be separated. Whereas the optical cross section is sensitive to the aspect of observation, thus the heading direction of the insect in flight, we demonstrate that polarimetric- and spectral-band ratios are largely invariant to the aspect of observation. We show that wing glossiness, as well as wing- and body-spectral ratios are particularly efficient in distinguishing Anopheles coluzzii and Anopheles arabiensis, 2 closely related species of malaria vectors. Spectral and polarimetric ratios relate to microstructural and melanization features of the wing and body of these species. We conclude that multiband modulation spectroscopy is a useful expansion of the parameter space that can be used to improve the specificity of entomological lidars.

Abstract-Assessment of biodiversity of pollinators on the landscape scale or estimation of fluxes of disease-transmitting biting midges constitutes a major technical challenge today. We have developed a laser-radar system for field entomology based on the so called Scheimpflug principle and a continuouswave laser. The sample-rate of this method is unconstrained by the round-trip time of the light, and the method allows assessment of the fast oscillatory insect wing-beats and harmonics over kilometers range, e.g., for species identification and relating abundances to the topography. Whereas range resolution in conventional lidars is limited by the pulse duration, systems of the Scheimpflug type are limited by the diffraction of the telescopes. However, in the case of sparse occurrence of the atmospheric insects, where the optical cross-section oscillates, estimation of the range and spacing between individuals with a precision beyond the diffraction limit is now demonstrated. This enables studies of insect interaction processes in-situ. REMOTE OPTICAL IN-SITU INSECT MONITORINGAlthough tiny in size, the massive number of insects makes them play a key role in most eco-systems around the globe. While the Western world experiences significant economic losses in agriculture due to lack of biodiversity and the colony collapse disorder of pollinators [1], disease vectors and pests are feared and, to a great extent, blamed for stagnating the development in tropical parts of the world.Whereas birds can be ring marked or tracked via GPS or sun loggers, only the very largest and least abundant insects can be equipped with electronic tags [2,3]. While research in the area of radar entomology has been conducted over several decades and numerous interesting applications have been described [4], laser radar (light detection and ranging; lidar) systems in the optical regime have the potential of achieving a far better sensitivity and address and classify even the tiniest insects, simply because most insects are much smaller than the wavelengths of microwaves used in radars but larger than the wavelengths of light. Further, optical off-the-shelf components allow spectral-and polarimetric target classification, providing molecular as well as microstructure information [5,6]. Along these lines our group has previously demonstrated lidar remote detection of insects labeled with fluorescent powders, e.g., for assessing dispersal rates on a landscape scale [7,8].Today a major limitation in ecological entomology is that insect abundance assessment is based on sweep nets, light-, pheromone-or CO 2 -traps. Placing and emptying the traps are tedious operations and constitute a major effort, and the results are known to be biased with respect to the species, sexes and age groups caught. Although trapping allows precise studies with microscopes, mass spectrometry

In recent years, the field of remote sensing of birds and insects in the atmosphere (the aerial fauna) has advanced considerably, and modern electro-optic methods now allow the assessment of the abundance and fluxes of pests and beneficials on a landscape scale. These techniques have the potential to significantly increase our understanding of, and ability to quantify and manage, the ecological environment. This paper presents a concept whereby laser radar observations of atmospheric fauna can be parameterized and table values for absolute cross sections can be catalogued to allow for the study of focal species such as disease vectors and pests. Wing-beat oscillations are parameterized with a discrete set of harmonics and the spherical scatter function is parameterized by a reduced set of symmetrical spherical harmonics. A first order spherical model for insect scatter is presented and supported experimentally, showing angular dependence of wing beat harmonic content. The presented method promises to give insights into the flight heading directions of species in the atmosphere and has the potential to shed light onto the km-range spread of pests and disease vectors.