The discourse surrounding precipitation changes in a warming climate (both public discussion and even scientific one at times) is complicated by widespread conflation of changes in averages vs extremes (and also actual vs *potential* evaporation/evaporative demand). [Thread] Increased temperatures rapidly raises "ceiling" on both precipitation & evaporation intensity--but same is *not* true for typical/average values of either! Regional mean precip can increase or decrease with warming, and actual evaporation is constrained by local H20 availability. This is the basis for "Expanding Atmospheric Sponge" effect that we coined to help visualize practical implications stemming from more heavier downpours (i.e., flash floods) & also more extreme evaporative demand (e.g., faster-developing droughts, more intense wildfires).
The analogy works well even w/caveats: A larger sponge only has *potential* to absorb more water--but will not necessarily do so unless water is present! Likewise, even a huge saturated sponge will not actually relinquish water unless actively wrung out. nature.com
www.nature.com
The "expanding atmospheric sponge" and its role in driving more extreme rainfall and faster/deeper droughts is enough to explain a large majority of why climate change is so clearly linked to these types of extreme events. But there is also more to the story, as discussed below. Continental interior and/or arid land areas, which have "functionally finite" water in soils and plants potentially available to contribute to evapotranspiration versus "functionally infinite" water over oceans, typically constrain actual evaporation relative to its potential. Further, the resulting (long-predicted) relative humidity decrease over land now appears to be stronger than earlier expected for reasons that remain TBD. Possible culprits include underestimated land-atmosphere feedbacks or possibly ENSO variability changes contrary to models. However, the fact that most of the *very most extreme* precip events in any given location are usually associated with strong/ long-range anomalous moisture transport from elsewhere (usually warmer ocean regions) means they will still increase even across continental interiors! As an aside: meteorological mechanisms that accomplish that kind of long-range moisture transport fluxing evaporated water from oceanic /large terrestrial water body sources to distant regions include atmospheric rivers, cyclonic warm conveyor belts, low-level jets, and etc. For these reasons, we cannot simply look to trends in mean precipitation, or mean evaporation, to tell us anything about changes in *extreme* precipitation or evaporative demand (i.e., atmospheric "thirst"). To do so would be deeply misleading. Yet it remains common practice. Conflating mean changes vs extreme changes is also dangerous when talking about floods, as well. Minor to moderate severity river floods do, indeed, show small or ambiguous trends despite warming; there is certainly no obvious climate signal here. BUT (and it's a big one): But...countervailing effects of increasing *extreme* precipitation (occurring only occasionally) & (more consistently) increasing evaporative demand potentially yield situation where minor/moderate floods decrease but *most extreme floods still increase!* nature.com
www.nature.com
As if that were not enough: most of the floods in the news right now are *flash floods* (sudden inundations over minutes to hours, not gradual inundation from large rivers flooding over days to weeks). This is important because... There is a stronger argument that climate change will, and likely already has, increased risk of *pluvial* flooding (i.e., direct inundation by runoff from local rainfall, including urban flooding) & flash flooding on fast-responding watersheds (as opposed to river flooding). That's because the countervailing effects above are less important during extremely intense short-duration downpours (i.e., the kinds that lead to flash flooding and rapid inundation in urbanized settings with fixed stormwater drainage capacity). In cases of pluvial flooding, and for many kinds of flash flooding in urban settings/on fast-responding watersheds, the sheer intensity of rainfall becomes the most important factor. And torrential downpours lasting minutes to hours are clearly increasing in a warming climate. Recently, there have been a lot of "bad takes" from folks who don't really understand climate, meteorology, or hydrology regarding our remarkable American "Flash Flood Summer" of 2025. I hope this thread can help contextualize, inform, and debunk (as appropriate). But let me be clear: Increasingly extreme downpours are one of the clearest meteorological manifestations of a warming atmosphere. This is already detectable in most place with adequate records (including eastern US); it will almost certainly become so elsewhere sooner vs later. It is absolutely appropriate to talk about climate change in the context of our American Flash Flood Summer 2025. It's certainly not the only factor at play; no serious scientist would argue that, either. But excluding it from the discussion would be just as disingenuous.