Detecção e análise de manchas solares utilizando técnicas de inteligência artificial e big data

• Ano de defesa: 2022
• Autor(a) principal: Justino, Jean Carlo Costa
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Cidade de São Paulo; Brasil; Pós-Graduação; Programa de Pós-Graduação Mestrado em Astrofísica e Física Computacional; UNICID
• Palavras-chave em Português: Manchas solares; Inteligência artificial; Big Data; Penumbra; CNPQ::CIENCIAS EXATAS E DA TERRA::ASTRONOMIA
• Link de acesso: https://repositorio.cruzeirodosul.edu.br/handle/123456789/4139

In operation since February 2010, the “Solar Dynamics Observatory” (SDO) obtains daily images of the Sun using the “Helioseismic and Magnetic Imager” (HMI) equipment. This equipment produces magnetograms and visible light images of the entire solar disk. Analyzing the images, it is possible to identify the sunspots. Sunspots represent darker regions in white light images, this region has a larger magnetic field relative to the Sun. These areas have different structures, which are divided between: umbra and penumbra. The umbra is a central and punctual region, with a darker shade in the stain, while the penumbra surrounds the umbra and has a lighter shade. The hue is directly related to the temperature, the darker the temperature is, the lower. The objective of this work is the statistical analysis of the structures of the sunspots mentioned above, in an automated way. It makes use of Artificial Intelligence and Big Data computing techniques to identify and calculate the physical characteristics of the umbras and penumbras, such as: area, temperature, magnetic field and location. Through the applicability of these techniques, a database was obtained that guides the analyses. This database reflects the 11-year period of daily maps, between May 1, 2010 and May 10, 2021. Analyzing a total of 6961 images, 12562 umbras, 16169 penumbras were detected, with longitudes between -40◦ and 40 ◦. As a final result, the correlations between the characteristics of the umbras and penumbras were analyzed, proving the different behavior between them. A formula and an Artificial Intelligence algorithm were also created to infer the maximum magnetic field of sunspots, based only on the area and temperature of the umbras and penumbras, thus ruling out the use of magnetic sensor data (magnetograms).