In January 2026, a coronal mass ejection sent a magnetic shockwave through the heliosphere. Within days, a muon detector at Star Safari Observatory in the Wairarapa recorded a clear, measurable drop in cosmic ray flux — then a slow recovery lasting nearly a week.
This protocol uses the real data from that event to run a genuine scientific investigation in any classroom in New Zealand. Students describe the baseline, identify the anomaly, quantify the change, and construct a causal explanation working from the solar eruption through to the detector reading on the ground.
The curated dataset is embedded on this page and downloads in one click. No specialist equipment or site visit required. Spaceward Bound facilitators Hari and Sam are available for in-school delivery where schools want that option.
Data collected at: Star Safari Observatory, 1169 Ponatahi Road, Carterton, Wairarapa — star-safari.nz
Spaceward Bound: Hari (astrobiologist) and Sam (astrophysicist) are Field-Based STEM facilitators available for in-school delivery of this investigation. Contact [email protected] to discuss options.
Cost: Dataset free. In-school facilitation costs on application.
Full dataset: The complete raw data from the Star Safari muon detector is also available by contacting [email protected].
Global Muon Network — Singapore:
http://131.96.55.92:8501/?country=Singapore&city=Bras+Basah&site=Singapore+Management+University
Global Muon Network — Sweden (Lund):
http://131.96.55.92:8501/?country=Sweden&city=Lund&site=BMSL
Oulu Cosmic Ray Station — neutron monitor:
https://cosmicrays.oulu.fi/
1. Describe the baseline behaviour of the data before the event.
2. Identify the anomaly: onset, minimum, recovery.
3. Quantify the change: estimate the percentage decrease from baseline to minimum.
4. Characterise the timescale: how rapidly did the drop occur, and how long was the recovery?
5. Propose and refine an explanation using physical principles. Strong responses cite specific features of the graph with quantitative estimates and link cause to effect at each step.
Describe the data and identify that a change in muon count rate has occurred. Name the approximate dates of onset and recovery.
Explain that solar activity affects the number of cosmic rays reaching Earth and link this to the observed decrease in muon count rate. Connect cause to effect with reference to the data.
Construct a coherent causal chain: a solar eruption (CME) disturbs the heliospheric magnetic field, reducing the flux of galactic cosmic rays reaching Earth, leading to fewer atmospheric interactions, reduced muon production, and a measurable decrease in detector count rate. Quantitative estimates and specific graph references required.
These prompts anchor AI use in the real data students have already worked with. Students bring their observations, measurements, and explanations to the tool; AI helps them extend, interrogate, or articulate their thinking. All prompts assume the investigation is complete.
Tell the AI what you saw in the graph: "The numbers went down around [date] and then came back up. Here is what I think that means..." Ask it to help you explain what you noticed in simpler words a friend who wasn't there could understand.
Tell the AI: "The Sun sent out a big burst of energy called a coronal mass ejection. Can you explain what that is using words a Year 5 student would understand?" Then decide whether you agree with its explanation.
Ask the AI to help you describe the journey of a particle: from outside the Solar System, through the heliosphere, into Earth's atmosphere, down to the detector. What happens at each step? Does the AI's description match what the data shows?
What do you still want to know after looking at the data? Ask the AI your question. Then think: could you answer it by looking at more data? What data would you need?
Give the AI your baseline estimate, the onset date and time, the minimum value, and the recovery period from your own measurements. Ask it to help you write a precise scientific description of the event, then check whether every claim is justified by your data.
Tell the AI: "I want to explain why the muon count dropped. Here are the steps I think are involved: [your steps]." Ask whether the chain is complete and whether any step is in the wrong order or missing a link.
Share your percentage change calculation with the AI. Ask: "Is this calculation method appropriate? What assumptions am I making? What would change my answer?" Use the response to check and refine your working.
After examining the Global Muon Network data, tell the AI what you found: "The NZ signal looked like [X]. The Singapore signal looked like [Y]. Here is what I think that difference means." Ask whether your interpretation is physically reasonable.
Draft your full causal explanation (CME → heliospheric magnetic disturbance → galactic cosmic ray flux → atmospheric interactions → muon production → detector count rate). Give it to the AI and ask: "What are the weakest links in this chain? What evidence from the data would make each link stronger?"
Share your baseline statistics and percentage change estimate. Ask the AI to help you think about uncertainty: what noise is present in the baseline, how confident can you be in your percentage estimate, and how would you report this result in a scientific context?
After comparing global detector sites and the Oulu neutron monitor, describe your findings: onset timing, signal magnitude, recovery rate. Ask: "What physical factors could explain any differences between detectors? What would you need to know to test that hypothesis?"
Ask the AI: "What can a muon count rate time series tell us about a Forbush decrease, and what can it not tell us? What additional instruments would a scientist use alongside this data?" Use the response to write a limitations section for your investigation report.
| Level | Years 0–6 | Years 7–10 | Years 11–13 |
|---|---|---|---|
| 1 | Student identifies that the muon count changed at a specific time and can point to where on the graph this happened. Recognises that the data came from a real detector in New Zealand. | Student identifies the onset, minimum, and recovery of the Forbush decrease and can state approximately when each occurred with reference to the data. | Student identifies the key features of the event (onset date and time, minimum value, recovery duration) and calculates an approximate percentage change from baseline. |
| 2 | Student explains that something from the Sun caused the muon count to drop and draws or describes a simple sequence connecting the Sun to the detector reading. | Student explains that a CME reduced the number of cosmic rays reaching Earth and links this to the observed decrease in muon count rate. Uses graph features as evidence. | Student constructs a causal chain from CME through heliospheric magnetic disturbance to reduced cosmic ray flux to reduced muon production. Uses data explicitly as evidence for each link. |
| 3 | Student compares the NZ data with what they would expect to see from a detector in another country, making a prediction before looking and explaining their reasoning. | Student accesses one additional detector from the Global Muon Network and makes a direct comparison with the NZ signal, noting similarities and differences with reference to the data. | Student compares data across at least two global detector sites and the Oulu neutron monitor, identifies whether onset timing and signal magnitude are consistent, and proposes an explanation for any differences. |
| 4 | Student explains what real detector data from Star Safari adds that a simulation or textbook diagram could not: it shows an actual event, in actual time, with actual numbers collected in New Zealand. | Student articulates what direct measurement via a muon detector provides that secondary sources cannot: independently verifiable, time-stamped, locally collected evidence of a solar event affecting Earth. | Student discusses the limits of muon count rate as a proxy for cosmic ray flux, identifies what complementary instruments would add, and frames the investigation within a broader space weather monitoring context. |
| 5 | Student generates a question they want to investigate further: for example, whether bigger solar events produce bigger drops, or whether the same pattern has appeared in the data before. | Student formulates a testable research question arising from the investigation (e.g., Is the recovery rate related to the size of the initial drop? Does detector latitude affect signal magnitude?) and identifies what data would be needed. | Student designs a follow-up investigation proposal: states a research question, identifies data sources, describes the analysis method, and acknowledges key uncertainties. Suitable as a Level 3 Physics practical investigation design. |