[A-CG31] Aircraft and UAV Observations for Earth-planetary sciences

[ACG31-05] Observation of red tide in an estuary by a fixed-wind type unmanned aircraft

*Soichi Yamaguchi¹, Shota Koide¹, Takayuki Mine², Takuya Iwanaga² (1.Kyushu university, 2.Saga Ariake Fisheries Promotion Center)

Keywords:red tide, fixed-wind type unmanned aircraft, spectral radiation, Ariake Sea

In order to elucidate the actual condition of red tide, it is important to observe the series of processes from the early stage of its development to the decay stage through the proliferation stage with high resolution in space and time. One of the methods for obtaining the concentration of sea surface chlorophyll a (Chl-a: an index of phytoplankton community) is observation by ship, but it requires a great deal of labor and cost to secure spatiotemporal resolution. Sea-color remote sensing using artificial satellites, which is another method, enables spatial monitoring, but in the orbit of the satellite, weather condition (clouds), and sea turbidity and seaweed aquaculture in the inner part of the Ariake Sea, which is the target sea area of this research, makes it difficult to obtain data with sufficient spatiotemporal resolution. Observations using unmanned aerial vehicles (UAVs) have been in the limelight in recent years because an aircraft observation has a high degree of freedom in observation such as immediacy and directness, and can perform observations with high spatiotemporal resolution that cannot be performed by satellite or ground observation. In this study, we aimed to establish a spatiotemporal high-resolution sea surface Chl-a concentration remote sensing method using a fixed-wing unmanned aircraft equipped with a spectroradiometer, and attempted to detect red tide in coastal waters. A handheld Spectra PAR Meter (PG200N by UPRtek corp.) was mounted on the UAV to estimate the Chl-a concentration in the sea surface. The PG200N can measure spectral data of light in the visible light region (350 nm to 800 nm) with a wavelength resolution of 1 nm, and can estimate the concentration using the optical characteristics of the Chl-a dye possessed by phytoplankton. It is small and lightweight, suitable for installation on UAV. In the observation, the irradiance incident on the sea surface from the sky and the radiance emitted from the sea surface were measured. In estimating the Chl-a concentration from the obtained spectral data, the reflectance of each wavelength (= upward radiance from the sea surface / downward irradiance from the sky) was calculated, and applied to a three-waveband model (Dall'Olmo, G. and Gitelson, A.A, 2003). Field observations were conducted on December 7 and 18, 2000 and January 5, 2021, and for fixed-wing type unmanned aircraft, OPTiM Hawk V2 (manufactured by OPTiM Corporation) has two PG200N sensors inside the aircraft body, so that it faced directly above and below and measured the spectral data of light from the sky and the sea. The flight speed of OPTiM Hawk V2 was set to 70km / hour, and the spectral data was acquired every 5 seconds, resulting in the spatial resolution of the data to about 100 m. A high correlation was obtained as a whole between the Chl-a concentration estimated by the 3-waveband model and the measured value by the field observation. However, the estimation error tended to increase in the sea area where the Chl-a concentration was high. Generally, when the concentration of phytoplankton is increased, the "packaging effect" that change in the light absorption coefficient per unit Chl-a concentration due to the overlapping of pigments occurs, and the Chl-a concentration estimated by sea color remote sensing is underestimated or overestimated. Therefore, analysis of covariance was performed to estimate the threshold of Chl-a concentration at which the packaging effect appears. Based on this threshold value, we succeeded in developing an algorithm that can estimate Chl-a concentration with high accuracy even in the high Chl-a concentration sea area. In the three observations, a high-concentration local region, which is considered to be the initial occurrence of red tide, was detected. Therefore, this observation using UAV has sufficient performance to grasp the spatiotemporal high resolution of red tide in the coastal area.