PIB TrackerAI reads a century of hand-drawn Sun charts from Kodaikanal to map the solar cycle
Researchers used machine learning to convert 100 years of hand-drawn Sun records from the Kodaikanal Solar Observatory into usable data, tracing bright magnetic 'plage' regions from 1916 to 2007 and rebuilding the Sun's activity cycle.
What happened
- Researchers led by Dibya Kirti Mishra of the Aryabhatta Research Institute of Observational Sciences (ARIES), an autonomous institute under the Department of Science and Technology (DST), used a supervised machine-learning model (U-Net) to turn about 100 years of hand-drawn 'suncharts' from the Kodaikanal Solar Observatory (KoSO) into machine-readable data.
- KoSO — run by the Indian Institute of Astrophysics (IIA) — holds daily suncharts from 1904 to 2022 on which sunspots, plages, filaments and prominences were drawn on a standard grid, one of the world's longest continuous solar records.
- The model first located the Sun's disk (centre, size, tilt) in each scan, then traced plages — bright, magnetically active patches — across nine solar cycles from 1916 to 2007.
- This built a 'butterfly diagram' showing how magnetic activity shifts with latitude and solar-cycle phase, and the plage areas matched those from KoSO's Ca II K full-disk photographs, showing the drawings can fill gaps in long-term data.
- Consistent long records let scientists compare the strength and structure of different solar cycles and better understand space-weather risks to satellites, navigation and power grids on Earth; the work appeared in an Astrophysical Journal publication.
For Prelims
- Kodaikanal Solar Observatory (KoSO): Established in 1899 in the Palani Hills, Tamil Nadu, and run by the Indian Institute of Astrophysics (IIA); it holds one of the world's longest continuous digitised records of the Sun.
- Indian Institute of Astrophysics (IIA): A Bengaluru-based autonomous institute under the Department of Science and Technology (DST) that operates KoSO.
- ARIES: Aryabhatta Research Institute of Observational Sciences, Nainital — an autonomous institute under DST; the study was led from ARIES by Dibya Kirti Mishra.
- Plages: Bright, magnetically active patches in the Sun's chromosphere — a reliable 'fingerprint' of solar magnetism, used as a proxy for the solar cycle. (Best seen in Ca II K, a calcium spectral line.)
- Solar cycle & butterfly diagram: The Sun's magnetic activity rises and falls roughly every 11 years; plotting where features appear by latitude over time makes a wing-shaped 'butterfly diagram' of the cycle.
- Space weather: Solar flares, sunspots and eruptions can disturb satellites, navigation, communications and power grids on Earth — long, consistent solar records improve forecasting of these risks.
- Don't confuse: Plages are bright chromospheric regions, distinct from dark sunspots (cooler photospheric regions) and from filaments/prominences; and the AI here read digitised historical drawings, it did not take new observations.
For UPSC: Machine learning turned a century of hand-drawn Kodaikanal suncharts into usable data, mapping solar plages from 1916 to 2007. Anchor KoSO (est. 1899, run by IIA under DST) as one of the world's longest solar records, and the value of long time-series plus AI for solar-cycle science and space-weather forecasting. Revise the solar cycle, sunspots vs plages, DST's research institutes (IIA, ARIES), and India's astronomy and space-science ecosystem.
What it is NOT: The AI did not make new solar observations — it digitised and read a century of existing hand-drawn records to build consistent data. The study maps the solar cycle from historical drawings; it is not a real-time space-weather forecasting service, and plages are not the same as sunspots.
For Mains
Syllabus: GS3.13 · GS3.11 · Linkage L2
Anchor
Value of long-term scientific archives plus modern AI — turning a century of hand-drawn Sun records into consistent data for solar-cycle science and space-weather understanding.
Substantiation (data)
U-Net machine learning applied to ~100 years of Kodaikanal suncharts (daily records 1904-2022); plages traced across nine solar cycles, 1916-2007; results matched KoSO's Ca II K photographs; KoSO established 1899, run by IIA under DST.
Exemplification
Cite KoSO's digitised suncharts and the AI-built butterfly diagram as a model for reviving legacy scientific data, alongside India's DST institutes (IIA, ARIES).
Problematisation
Inconsistent old records (drawing styles, paper ageing, scan quality), the need for expert-labelled training data, and translating archival science into operational space-weather forecasting remain challenges.
Way-forward
Extend AI-based digitisation to other historical archives, integrate long solar records into space-weather models, and sustain open, digitised scientific datasets.
Position
Preserving and mining long-term scientific archives with AI strengthens both fundamental science and preparedness for space-weather risks.
Deploys into: AI/machine learning applied to scientific archives & space-weather science (Kodaikanal solar records, DST institutes) · India's astronomy and space-science ecosystem (GS3.13 science & technology — IT/space/bio/IPR, GS3.11 applications of S&T in everyday life).
Ministry of Science & Technology · 2026-07-01 · PRID 2279849 · PIB source ↗