Cosmic Data Science:
Our Sun
This page holds materials related to a proposal for the next phase of CosmicDS: Cosmic Data Stories, based at the Smithsonian Astrophysical Observatory.
Once this program is deployed, the kinds of materials shown on this page will be moved to pages learners will see before, during, and after they undertake their exploration of the Solar Cycle. (The page as-is is being used by the development team for planning purposes, and it is only accessible by those who know its URL.). In general, the final structure of materials will be:
-
BEFORE" A quick introduction to sunspots and solar brightness measurements (TSI). Sample viideos included below that give away the Solar Cycle punchline will NOT be included in the introduction.
-
DURING An page with embedded LLM and visualization tools, used in the data exploration and visualization activities that allow learners to "discover" the Solar Cycle themselves.
-
AFTER An assortment of follow up videos (drawn from the samples below), explaining how researchers, going all the way back to Galileo, have studied the Sun's activity, and Cycle.
See a short example run-through here!
Explanatory Video
What's "Solar Maximum"?
Why the connection to aurorae, and why does this matter in our daily lives?
Video credit: Goddard Space Flight Center (source)
Explanatory Video
How much do sunspots effect the total brightness of the Sun?
A measurement called "TSI" (of "Total Solar Insolation" tells us how bright the Sun appears. It varies due to various factors, including sun spots. Learners might wonder how important the sunspots are to changing TSI...
Video credit: Laboratory for Atmospheric and Space Physics (source)
Curated videos will be offered to learners in a custom "channel" like what is shown here, once they have "discovered' the Solar Cycle on their own
Solar Cycle
Solar Cycle
NASA | The Heliophysics Program
How To Track The Solar Cycle
NASA | Solar Cycle
For demonstration purposes only
Sunspot Cycle Data Exploration by an Expert, using an LLM
The summarized discussion between Prof. Alyssa Goodman & ChatGPT 4o, immediately below, shows what AI could do as of June 2025, in the hands of an expert. Using just one data set, about an historical record of sunspots over about 150 years, tremendous insight into both the Solar cycle, and the limitations of "real-life" data are made clear.
In designing Cosmic Data Science, we surely do NOT expect learners to instantly acquire the level of sophistication to ask the questions that allowed an expert to make these graphs in under an hour. BUT, we will test--in our pilot project-- how various elements of what ChatGPT helped Goodman do can be deployed in Cosmic Data Science.
Below the summary discussion, a more sophisticated visualization, which is interactive, shows the sunspot and TSI data together, overlain in way that facilitates comparison and measurement.
Even that interactive visualization used only "vibe coding," which refers to an expert using plain English to iterate with an LLM, where only the LLM actually writes the code.
It took just over an hour to produce the interactive graphic. It is shown on this site as a Solara app using plot.ly within the app. The app itself is hosted on a GitHub page. Note, though, that non-technical visitors to this site just see a cool interactive graphic, and do not care about how the modular architecture of modern web tools allows Wix (which is used to create this website), Solara, plot.ly, and GitHub to show them what ChatGPT, which uses Python plotting libraries like matplotlib, can do in the hands of an expert.
And, finally, here's a Sunspot Cycle Data Exploration, Interactive Graphic, created by visualization expert using only ChatGPT, as a demonstration.
This Interactive plot shows the daily Sunspot number (in blue) and Total Solar Irradiance (in orange) for 1945-2018, with lines indicating the beginning of each solar cycle.