Researchers at the University of Michigan have proposed a new metric that measures how much of a region’s annual rainfall arrives in short, intense bursts rather than in steadier, more manageable storms. The Extreme Precipitation Dependency Index, or EPDI, quantifies the share of yearly precipitation delivered by events above the 99th percentile, and the study’s projections suggest that share could rise by about 15% to 20% in some regions under a 4 degrees Celsius-above-pre-industrial warming scenario. The finding reframes a familiar climate story: water is not simply disappearing, but arriving in ways that make it harder to capture, store, and use.
What the EPDI Actually Measures
Most existing precipitation metrics track totals or averages, which can mask how rain is distributed across a calendar year. The EPDI takes a different approach. Developed by University of Michigan researchers and summarized by the school’s sustainability center, the peer-reviewed study isolates the fraction of annual precipitation that falls during extreme events, defined as those exceeding the 99th percentile of historical intensity. A rising EPDI means a region is getting a larger portion of its water budget from rare, heavy downpours rather than from moderate rainfall spread across weeks or months.
That distinction matters for reservoirs, aquifers, and farmland alike. When rain arrives in concentrated bursts, much of it runs off hardened or saturated soil before it can soak into groundwater reserves. Reservoirs designed for steady inflows can overflow during deluges and sit half-empty during the long dry stretches in between. The index gives planners a single number to track how fast this imbalance is growing in any given watershed, and a recent overview on Phys.org highlights the study’s finding that, in some places, a large share of annual precipitation is already delivered by rare extreme events.
A Paradox of Shrinking Supply Amid Heavier Storms
The EPDI builds on a tension that scientists have documented for years. A global analysis of freshwater availability found that water supplies are shrinking in many basins even as climate change generates more intense precipitation events. That paradox, more rain falling but less water available for human use, stems from higher temperatures driving faster evaporation and longer dry intervals between storms. The net effect is that total annual rainfall can hold steady or even rise while usable supply drops.
The record-breaking drought across the Western United States from 2020 to 2021 illustrated this dynamic in real time. Research hosted by the NASA Goddard Institute traced the drought’s severity to compounding pressures from persistent dryness and elevated evaporative demand. Those same pressures can increase a region’s dependence on episodic extreme rainfall for replenishment, because moderate storms no longer deliver enough moisture to offset losses between events. The EPDI essentially quantifies how far a region has moved into that regime, where a few storms determine whether a year is wet enough for crops, ecosystems, and cities.
At the global scale, this shift dovetails with broader research on how warming alters the hydrologic cycle. Work published in Nature’s climate journal underscores that higher temperatures intensify both heavy precipitation and drought risk, tightening the link between extreme events and long-term water security. The EPDI translates that high-level insight into a practical diagnostic tool for specific basins and sectors.
Projected Risks for Croplands and Northern Forests
The stakes extend well beyond municipal water systems. According to the study described in Water Resources Research, 54% to 96% of global rain-fed croplands may be adversely affected by rising extreme-precipitation dependency, alongside significant risks to boreal regions. Rain-fed agriculture, which receives no irrigation and depends entirely on natural rainfall timing, is especially exposed. Farmers who once relied on predictable wet seasons could face longer droughts punctuated by floods that damage crops rather than nourish them.
In practical terms, a higher EPDI for croplands means more years where planting decisions hinge on whether one or two key storms arrive on time. If they come too early, fields may remain dry during critical growth stages; if they come too late, seeds can fail, and mature plants can be flattened or washed out. Soil erosion also accelerates under intense rainfall, stripping away nutrients and reducing the land’s capacity to absorb future storms, which further entrenches the dependence on extremes.
Boreal forests across Canada, Scandinavia, and Russia face a related threat. These ecosystems depend on snowmelt and steady summer rain to sustain soil moisture through short growing seasons. A shift toward fewer but heavier precipitation events could leave soils dry for longer stretches, raising wildfire risk and stressing tree species already pushed to their thermal limits. As the EPDI rises in high-latitude regions, forest managers may see more years that oscillate between flash flooding and fire-prone dryness, challenging traditional approaches to conservation and timber planning.
Federal Data Confirm Extremes Are Intensifying
Independent U.S. government datasets reinforce the pattern the EPDI captures. The U.S. Geological Survey has released a suite of climate-related data products that include annual extreme-precipitation metrics for the contiguous United States, derived from CMIP6-LOCA2 daily precipitation data and processed with the xclim toolkit (see USGS dataset details here). These variables track maximum one-day, five-day, and ten-day precipitation totals, along with counts of days exceeding the 90th, 95th, and 99th percentile thresholds. Taken together, the metrics offer a granular, county-level view of how extreme rainfall events are shifting in frequency and magnitude across the country.
USGS guidance on flood-frequency analysis has long underpinned the way engineers and planners estimate the likelihood of damaging flows in rivers and streams. As the tails of the precipitation distribution swell, those statistical tools are being pushed beyond the climate they were calibrated for. That makes complementary metrics like the EPDI valuable for highlighting where standard design assumptions may no longer hold.
Separately, a NOAA team built an interactive library of historical extreme precipitation events to support a specific practical goal: modernizing Probable Maximum Precipitation estimates that determine the design specifications for high-risk infrastructure (as described by NOAA’s storm library project). As extreme events intensify, PMP estimates based on older climate data can risk underestimating the storms that critical structures must withstand. Integrating EPDI-style diagnostics with such event libraries could help identify locations where infrastructure is most exposed to a growing reliance on the most violent storms.
Why Current Coverage Misses the Real Problem
Much of the public discussion around extreme weather focuses on individual disasters: a hurricane season, a catastrophic flood, a drought emergency. That framing treats each event as an anomaly. The EPDI challenges that assumption by showing the structural shift underneath the headlines. The problem is not just that storms are getting worse; it is that the entire water cycle is reorganizing so that a growing share of usable precipitation depends on those worst storms arriving at the right time and place.
This distinction has direct consequences for how cities, farms, and utilities plan ahead. Infrastructure built for a climate where rain fell in moderate, frequent doses cannot simply be scaled up to handle rare deluges. Flood channels, stormwater capture systems, and reservoir operations all need redesigning around a rainfall regime that delivers more water in less time, with longer gaps in between. The cost of that redesign, and the risk of delaying it, grows with every percentage point the EPDI climbs.
For policymakers, the index also reframes adaptation as a question of managing volatility rather than just scarcity. Even regions that are not projected to become dramatically drier overall may see their water security erode if a rising fraction of supply is tied to a handful of extreme days. By tracking that dependency now, governments and utilities can prioritize investments in flexible storage, soil health, floodplain restoration, and demand management that make communities less vulnerable to the whims of the most intense storms.
Ultimately, the EPDI does not replace existing measures of drought, flood risk, or annual rainfall. Instead, it exposes a hidden dimension of climate stress: how much our lives and economies are coming to depend on the behavior of the atmosphere’s wildest days. Recognizing and responding to that dependence may prove as important as counting every additional millimeter of rain.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
