[P-PS03] 太陽系小天体：太陽系の形成と進化における最新成果と今後の展望

[PPS03-P02] Parameter Estimation of Thermophysical Properties of Asteroid Ryugu observed by TIR onboard HAYABUSA2

*荒井 武彦¹、出村 裕英²、千秋 博紀³、坂谷 尚哉⁴、嶌生 有理⁴、神山 徹⁵、関口 朋彦⁶、金丸 仁明⁷、石崎 拓也⁴、田中 智⁴、岡田 達明⁴ (1.前橋工科大学、2.会津大学、3.千葉工業大学、4.宇宙航空研究開発機構、5.産業総合技術研究所、6.北海道教育大学、7.東京大学)

キーワード：はやぶさ２、中間赤外カメラ、リュウグウ、熱慣性、表層ラフネス

The thermal infrared imager (TIR) onboard the Hayabusa2 spacecraft observed brightness temperature emitted from the surface of asteroid Ryugu during 2018 and 2019. TIR estimates the surface thermophysical properties to understand the surface thermal evolution of Ryugu and the origin of their accumulation process. Thermal inertia is an essential index for deriving surface thermophysical properties, but surface roughness confuses that estimation. Since the observed surface brightness temperature was globally flattened due to the roughness, we required a surface thermophysical model using the detailed shape model, including the roughness model. We constructed the thermal inertia map using the roughness model for a one-rotation period of Ryugu (7.63 hours). The value of the thermal inertia depends on the observation period and observer position. Therefore, we determine the accurate thermal inertia using other observed data of TIR. However, the data analysis undergoes high calculation costs of time. We develop the parameter estimation method of the thermophysical properties using one observed image of Ryugu to reduce the computation time. We used the shape model of Ryugu, whose number of polygons is 196,608. The model, including the surface roughness model, is developed to compare the observed image with anticipated images derived from the thermophysical model. The thermophysical model includes a view factor calculation using the ray tracing method to calculate the re-radiation process from neighboring surfaces by a collision detection algorithm. This computation cost is enormous. Therefore, we use a parallel computing method using OpenMP. We use a few tricks, such as fixing the pair of view factors and limiting calculation areas of the shape model in visible areas of TIR. We assume the roughness model using inverted triangle pyramids dividing the polygon facet of the shape model of Ryugu. The height of the inverted pyramid depends on an RMS slope as the roughness index. We made an anticipated image derived from the thermophysical model with the roughness model. Using the least-squared fitting for one observed image, we searched the best-fit value of the thermal inertia and the RMS slope as the free parameters. We could obtain the thermal inertia value, but the parameter depended on the roughness parameter of the RMS slope. In this study, we resolve the dependency by a maximum likelihood estimation.