Rain is arriving in fewer, shorter bursts, a study of 19,183 weather stations finds

Half of a typical location’s yearly rain now falls in about 0.4 fewer event-days with each passing decade, according to a global analysis of daily gauge records from 19,183 weather stations. The work, led by Su and colleagues and published in Earth’s Future in 2026, covers 1980 through 2022 and finds that wet spells are getting shorter at nearly half of the stations that met its data-quality bar.

Help yourself and the planet: The Cool Down makes it easy to help yourself while helping the planet, with smart, simple changes to save more and waste less in your everyday life.

Get The Cool Down →

We may earn a commission if you sign up.

The finding concerns how rain is bunched in time, not how much of it falls. Annual totals can stay the same while the calendar days that deliver them shrink, and that is the pattern the authors describe across much of the planet.

The research team, which includes Su, Miao, Hu, Ji, Zhao and Gou and is affiliated with Beijing Normal University, built a new measure called WE50. It counts the combined duration of the wettest consecutive rain events needed to deliver about half of a station’s annual wet-day precipitation. The Watchers’ summary of the paper, written by Reet Kaur and published on Nov. 6, reports that the global average is 9.4 events spread over 27.8 event-days, and that the figure is falling by 0.4 days per decade. The summary quotes the result as a change in how precipitation is organized through time, and it carries the paper’s caveats along with its numbers.

WE50 and the 19,183-station record

The team began with 39,199 stations and kept 19,183 after requiring more than 30 years of daily data, fewer than 10 percent missing daily values and adequate yearly completeness. Event thresholds were set locally, using each station’s 1981-2010 mean wet-day rainfall, so a drizzle-prone coastline and a desert station are judged against their own norms. The paper is Earth’s Future article e2026EF008471, named “Global changes in precipitation concentration through consecutive event organization.”

Africa was left out of the continent-level analysis because long, consistent gauge records were scarce. Coverage elsewhere is uneven, which means densely monitored regions carry more weight in the global statistics.

Shortening wet spells in Europe, South America and Oceania

A significant decline in WE50 appeared at 45.7 percent of stations, a significant increase at 9.8 percent and no significant trend at 44.5 percent. The sign is not universal, then, and a bit under half of locations show nothing clear either way. News of Bahrain’s coverage, published on Nov. 5, adds that Europe has the highest average WE50 at about 35.1 days, with Asia at 18.8 and Oceania at 18.6.

Europe and South America drive much of the shortening. Mean event duration fell by roughly 0.08 days per decade in Europe and 0.10 in South America, and the paper attributes their increased concentration mainly to shorter events. In Oceania the change leaned more on the amount of precipitation, with the number of selected events falling by 0.2 per decade. Asia moved the other way on per-event rainfall, rising by 0.9 millimeters per event per decade.

Globally, the number of events needed to deliver half the year’s rain showed no significant trend, at minus 0.01 per decade. What changed was duration: the mean length of the selected events dropped by 0.06 days per event per decade, about 86 minutes, and 38.3 percent of stations posted a significant decrease in event length.

Gauge limits on flood and drought claims

The result is easy to over-read, and the broader climate science points the same cautious direction. The IPCC’s sixth assessment, in its chapter on weather and climate extremes, judges human influence, mainly greenhouse gas emissions, likely to be the main driver of the global-scale intensification of heavy precipitation over land. That finding concerns intensity of extremes, a different quantity from the timing metric in the Su analysis. Shorter wet spells do not automatically mean worse floods, and the authors say so: the study did not measure whether the detected WE50 trends caused changes in flood or drought occurrence. The decomposition that splits the trend into duration and amount components is diagnostic and not causal.

The cleanest independent context comes from a conference abstract by Tiantian Xing, Carlo De Michele and Gunter Bloschl, presented at the EGU General Assembly in May 2026. Their work on compound spatial precipitation covers 1980 to 2024 and finds that the area hit by simultaneous rain extremes is expanding faster than the intensity of those extremes, a gap that is largest in the tropics. The abstract offers no numeric rates and does not use the WE50 measure, so the two studies are complementary rather than confirming one another.

One robustness check in the Su paper matters for interpretation. Replacing the local threshold with a fixed 1 millimeter per day cutoff raised the global WE50 to 33.1 days and lowered the event count to 8.4, but the main regional contrasts survived. In a separate comparison, 93.0 percent of WE50 days overlapped with the single wettest days, meaning the new measure largely tracks the same stretches of calendar that older daily-extreme metrics flag. The open question is whether shorter wet spells translate into more runoff trouble, which the gauge data alone cannot answer.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

Help yourself and the planet: Get The Cool Down →

We may earn a commission if you sign up.