Entomological Lidar : Target Characterization and Field Applications

Sammanfattning: This thesis treats entomological lidar from various angles: laboratory reference work on insects of interest, methodological development of lidar- and data processing methods, as well as field implementations of lidar techniques for entomological research. Insects are crucial components of ecosystems and are currently in a global decline. In this thesis, insects are mainly studied in their roles as disease vectors and food sources for vertebrates. However, several other feasible application avenues of entomological lidar exist and are touched upon briefly. Entomological lidar is an optical remote sensing technique in which the light scattered by insects is recorded by a sensor and the distance to each insect is derived. In classical lidar, ranging is achieved through time-of-flight detection. In this thesis the Scheimpflug- and passive lidar methods have been used, in which ranging is achieved through triangulation and geometrical optics.In the laboratory reference work, the light-scattering properties of insects were investigated. Considerable effort has been put into the study of ex-vivo and in-vivo malaria mosquitoes in search of optical properties that may enable remote species classification. These species otherwise require capture and microscope analysis by an expert to distinguish. Dragonflies have the narrowest spectral bands so far observed in nature. In this PhD work, their scattering properties were investigated and give hints regarding possible uses of these narrow bands.Methodological development has been pursued for improvement and optimization of instrumentation through simulation and laboratory reference measurements. Hyperspectral images of insects were used to motivate laser wavelength selection based on signal strength, information yield and laser availability. Raytracing was used to devise a passive lidar scheme and to optimise the geometry of Scheimpflug lidars. Data processing techniques for robust and accurate calibration of sizes, wing-beat frequencies with associated modulation spectra, flight headings and dispersal of insects in lidar data were developed.Entomological lidar techniques were applied in a number of field settings around the world. In Sweden, insect swarms at the nacelle of a wind farm were observed post sunset in weather conditions associated with high bat mortality through collision with wind farms. In China, increased insect activity was observed at the onset of heavy rain. The main crepuscular activity peak of insects was observed in the short time window with decreased predation pressure around sunset, when neither birds nor bats were active. In Africa, an extra activity peak was observed at noon among mosquitoes and other crepuscular insects during a solar eclipse. Male mosquito mating swarms were observed with consistent timing and location each day, and a highly directional dispersal of mosquitoes into a village was observed every evening.In this thesis work, peak numbers of more than a thousand insects per minute have been observed, resolved temporally and spatially at μs and cm scales, respectively, which is inconceivable with conventional entomological methods. Laboratory reference work and methodological development allow the quantification and classification of insect signals in-situ. Thereby, questions of significant ecological importance could be answered.