Combining a Geomagnetic Precursor with Dynamic Time Warping for Solar-Cycle Peak and Trajectory Forecasting

The Sun is not a steady, unchanging lamp; it is a restless star that breathes in a rhythm of roughly eleven years. This heartbeat, known as the solar cycle, dictates the number of dark spots that appear on the Sun's surface and drives the intensity of the space weather that surrounds our planet. When the Sun is active, it can send powerful bursts of radiation and magnetic fields toward Earth, potentially disrupting satellites, radio communications, and power grids. For scientists and engineers who must protect our technology, predicting when this cycle will reach its peak and how it will behave is a critical challenge. The difficulty lies in the timing: near the very beginning of a new cycle, there is very little sunspot data to work with, making it hard to guess how strong the coming storm will be. Conversely, as the cycle matures, the pattern of spots becomes clearer, but the early clues about the total strength of the storm are often lost or forgotten.

Two researchers, Varun Soman and Amaranathan Sabarinath, have developed a new way to bridge this gap by combining two different types of historical clues. One clue comes from the Earth's magnetic field, which acts as a sensitive detector for the Sun's activity even before the new cycle of sunspots fully develops. The other clue comes from looking at the past: finding previous solar cycles that look very similar to the one currently unfolding and using their future paths as a guide. By weaving these two sources of information together, the team created a forecasting method that is more accurate than using either clue alone. Their work, tested against the last four completed solar cycles, suggests that this hybrid approach can predict the peak strength of a solar cycle with significantly less error than previous methods, while also providing a more realistic picture of how the cycle will rise and fall over time.

The core of their method relies on a technique called dynamic time warping, which is essentially a flexible way of comparing shapes. Imagine trying to match the outline of two different hills; one might be tall and narrow, while the other is shorter and wider. A rigid comparison might fail to see that they are similar in shape because their peaks happen at different times or over different durations. Dynamic time warping allows the computer to stretch or compress the timeline of one hill to fit the other, finding the best possible match between their curves without forcing them to be the same length. In this study, the researchers used this tool to compare the early, partial shape of a developing solar cycle against a library of complete, historical cycles. They did not force the historical cycles to be shortened to fit the new one; instead, they kept the historical cycles at their full, original length and let the matching algorithm find the best alignment. This specific choice turned out to be the most important factor in improving the accuracy of the forecast.

Before this new approach, researchers often shortened the historical cycles to match the length of the data they had at the moment of prediction. The new study found that this practice of resampling actually distorted the comparison, leading to less accurate predictions. By keeping the historical cycles intact and using the flexible matching tool, the researchers reduced the average error in predicting the future path of the sunspots by a noticeable margin. When they tested their method on Solar Cycles 21 through 24, the new technique produced a forecast trajectory with an average error of about 20 smoothed sunspots, compared to an error of roughly 26 when using the older, resampling method. This improvement was not due to a more complex mathematical calculation, but simply because the method preserved the true length and shape of the historical examples, allowing the algorithm to select better matches.

The second major innovation was how the team combined the two different types of predictions. The geomagnetic precursor, derived from the Earth's magnetic field, is excellent at giving an early estimate of how strong the cycle will be, but it does not tell us much about the shape of the curve. The analog method, which looks for similar past cycles, gets better at predicting the shape as more sunspot data becomes available, but it can be uncertain about the total strength early on. The researchers created a system that weighs these two estimates based on how much historical uncertainty each one has. When the cycle is just starting, the geomagnetic clue carries more weight because it is the only source of information about the peak strength. As the cycle develops and more sunspot data accumulates, the weight shifts toward the analog method, which then provides a more detailed picture of the cycle's evolution. This dynamic balancing act resulted in a combined peak prediction with an average error of about 16 smoothed sunspots, a significant improvement over the 25-sunspot error of using the geomagnetic clue alone.

Another practical problem the team solved involved the "declining phase" of the solar cycle. When a forecast is adjusted to match the latest observed data, simple mathematical shifts can sometimes push the predicted sunspot numbers below zero during the waning years of the cycle. Since it is impossible to have a negative number of sunspots, this would create an artificial, flat line at zero that does not reflect reality. The researchers introduced a rule to prevent this, ensuring that the forecast trajectory remains physically possible by adjusting the math only when necessary to keep the numbers positive. This change did not necessarily make the historical predictions more accurate, but it ensured that the forecasts remained realistic and free of impossible artifacts.

The study also examined how well the method handles uncertainty. By generating thousands of possible future paths based on historical errors, the team created "envelopes" that show the range of likely outcomes. They found that while these new envelopes are generally narrower and more precise than those from older methods, they are not perfect. For some cycles, like Solar Cycle 22, the method struggled to capture the rapid decline, leading to a wider gap between the prediction and reality. This suggests that while the method is a robust tool, it is not a crystal ball; the behavior of the Sun can still surprise us, particularly when a cycle behaves differently from the historical examples available. The researchers emphasize that their uncertainty bands should be viewed as empirical measures of spread rather than guaranteed limits, as the small number of past cycles available for testing makes it difficult to be certain about every possible outcome.

Looking ahead, the researchers have set a strict plan for applying this method to the upcoming Solar Cycle 26. They have decided to fix all the rules and parameters of the model before the cycle begins, ensuring that no changes are made after the outcome is known. This approach will provide a true test of the method's ability to predict a new cycle in real time. Until then, the method has been applied to Solar Cycle 25, which is currently underway. The analysis of this cycle shows how the combined estimate evolves, starting with a geomagnetic guess and gradually refining itself as more sunspot data becomes available. The study concludes that by combining early magnetic clues with a flexible search for historical twins, scientists can now offer more reliable forecasts of the Sun's behavior, helping to better prepare for the space weather that affects our modern world.