11:00 〜 13:00
[PEM12-P03] Automatic derivation of ionospheric electron density profile by VIPIR2 ionosonde observation and its validation
キーワード：イオノゾンデ、電子密度高度分布、機械学習
An ionosonde transmits high-frequency radio waves towards the ionosphere while sweeping frequencies and measures delay time until the transmitted radio waves are reflected to the ionosonde by the ionosphere. Normally, the observation data is recorded as “ionogram” in which the signal strength is indicated with the horizontal axis of the frequency and the vertical axis of virtual height. The virtual height is derived by multiplying the speed of light by flight time and dividing by two. The virtual height is different from true height since the propagation speed of the transmission radio wave is a function of the electron density distribution and the geomagnetic field. In order to obtain the electron density distribution with respect to the “true height” from the ionogram, a procedure is required in which an echo trace is performed to read the frequency and delay time of the reflected echo one by one, and the propagation speed is considered and converted to the electron density with respect to the true height.
National Institute of Information and Communications Technology has been conducting routine ionosonde observations for many years. In this project, some parameters such as critical frequencies and virtual heights of the E- and F- layers are automatically derived, however, the echo trace has not been conducted yet. Recently, we developed a technique to trace ionospheric echoes using machine learning. The traced echoes were subjected to a procedure of deriving electron density profile against true height using the POLynominal Analysis program (POLAN). In this presentation, we will overview our derivation technique of ionospheric electron density profile. We also show some results of its validation using density profile data observed with an incoherent scatter radar and LEO satellite occultation data.
National Institute of Information and Communications Technology has been conducting routine ionosonde observations for many years. In this project, some parameters such as critical frequencies and virtual heights of the E- and F- layers are automatically derived, however, the echo trace has not been conducted yet. Recently, we developed a technique to trace ionospheric echoes using machine learning. The traced echoes were subjected to a procedure of deriving electron density profile against true height using the POLynominal Analysis program (POLAN). In this presentation, we will overview our derivation technique of ionospheric electron density profile. We also show some results of its validation using density profile data observed with an incoherent scatter radar and LEO satellite occultation data.