When the Sky Goes Dark, Traffic Goes Quiet
On August 12, 2025, a total solar eclipse traced a path across the North Atlantic and into mainland Europe, crossing Iceland, northern Spain, and Portugal just before local sunset. It was the first total eclipse to touch the European mainland in two decades, drawing crowds outdoors and, as Cloudflare Radar data shows, pulling internet traffic down with remarkable precision.
Comparing HTTP request volume in five-minute intervals against a three-week baseline, the data reveals a clear pattern: as the moon covered the sun, online activity dipped. The moment of maximum eclipse in each country aligns almost perfectly with the deepest trough in traffic, shown as dark red in the heatmap below. In regions where the eclipse was only partial and shallow—like Sweden, Denmark, and Poland—the effect was negligible or absent. But along the path of totality and in areas with deep partial coverage, including Iceland, Ireland, the UK, France, Spain, and Portugal, traffic fell into a sharp, short-lived trough that rebounded within minutes as the sun re-emerged.
To confirm the eclipse was the cause rather than coincidence, a scatter plot of each country’s peak solar obscuration against its traffic change in the 15-minute window around maximum eclipse shows a strong inverse relationship. Nations in the path of totality saw traffic drop by roughly 15% to 30%, while those with minimal obscuration saw little to no decline. The dotted trend line tracks closely with the data points, and the timing of the dips matches the eclipse’s progression too closely for random variation. Local factors like population density, time of day, and cloud cover explain some scatter, but the overall signal is consistent.
National Variations Tell the Story
Looking at individual country trends, the effect is starkest where the eclipse was deepest. Iceland, Spain, and Portugal experienced the most dramatic declines, with traffic changes ranging from a modest 9.3% decrease to as much as 46.7% below baseline. In contrast, Poland and Denmark returned to normal quickly, with Denmark showing the smallest overall change. Norway and Sweden even recorded slight increases above baseline, a reminder that the eclipse’s impact was highly localized.
The analysis uses precise astronomical calculations: for each region, the apparent angular sizes of the sun and moon and their separation in the sky were used to compute the fraction of the sun’s disk covered every five minutes. This yielded both peak obscuration (0–100%) and the exact moment of maximum eclipse. National figures aggregate regional data, with average obscuration used to set the country-level max-eclipse time. To separate the eclipse effect from a typical Wednesday evening, the baseline was the median of the three previous Wednesdays, matched slot-by-slot on time of day, preventing any single anomalous week from skewing results.
The findings underscore a simple truth: digital traffic is a reflection of human attention. When a shared physical event like a total eclipse pulls people outdoors and away from their screens, the internet slows—not due to technical failure, but because of a collective pause. As patterns normalized shortly after the eclipse passed, the data left behind offers a clear snapshot of how a cosmic event can temporarily reshape continental-scale online behavior.
