Indian researchers have successfully used Artificial Intelligence to digitize hand-drawn solar observations from the Kodaikanal Solar Observatory (KSO), creating a consistent record of magnetic activity from 1916 to 2007. According to the Ministry of Science and Technology, this machine-learning approach allows scientists to better understand space weather risks that can affect technology on earth.
How AI Transforms Historical Solar Records
For more than a hundred years, solar astronomers recorded features like sunspots, filaments, and plages—bright, magnetically active regions—by hand on paper grids. These KSO ‘suncharts,’ spanning from 1904 to 2022, represent a collection of solar behavior. However, inconsistencies in drawing styles and the physical degradation of paper records previously hindered their use.
A research team led by Dibya Kirti Mishra, involving the Aryabhatta Research Institute of Observational Sciences (ARIES) and collaborators from the Indian Institute of Space Science and Technology, the Southwest Research Institute, and the Indian Institute of Astrophysics, addressed these challenges using a supervised machine learning model called U-Net. According to the study published in The Astrophysical Journal, the model performed two critical tasks: it first standardized the orientation and size of the sun’s disk in each image, then automatically identified and traced plage regions across nine solar cycles.
Why Long-Term Solar Data Matters for Earth
Understanding the sun’s rhythmic cycles is essential for protecting technology. Solar flares and eruptions can disrupt satellites, navigation, and power on earth. By converting these historical drawings into machine-readable data, researchers have generated a “butterfly diagram” that illustrates the solar cycle.

According to the Ministry of Science and Technology, this data allows scientists to compare the strength and structure of different solar cycles. This historical context is vital for improving reconstructions of how the sun’s energy output and magnetic influence have changed in the past. The study confirmed that plage areas derived from the AI-processed drawings align closely with KSO’s full-disk observations, validating the reliability of the historical archives.
Future Trends in Solar Research
The success of the KSO digitization project signals how historical archives are utilized in space physics. Research focuses on applying U-Net models to historical records to improve long-term solar data.
Pro Tip: To learn more about how solar activity impacts our planet, explore the NOAA Space Weather Prediction Center, which provides real-time monitoring of solar events.
Frequently Asked Questions
What are plages on the sun?
Plages are magnetically active patches on the sun, considered a reliable “fingerprint” of the sun’s magnetism.
How does AI help in solar research?
AI models, such as U-Net, can identify and trace features in historical hand-drawn records to create machine-readable data from otherwise inconsistent archives.
Why is 1916–2007 data important?
This period covers nine solar cycles. By analyzing this timeframe, scientists can connect today’s space-age measurements with what the sun was doing decades earlier.
Have questions about how solar cycles affect our modern technology? Share your thoughts in the comments below or subscribe to our newsletter for the latest updates on space science and technology.


