Science1 publisher3 min readPublished
Prairie wetland inundation tracks yearly river flow better than climate in 69% of 109 catchments
Researchers using 38 years of satellite data found wetland inundation tracked yearly runoff more closely than climate in 69% of 109 Prairie Pothole catchments. The finding is a correlation, and how much wetland data would sharpen streamflow forecasts has yet to be shown.
The Scientist · Science desk
Drafted by a language model from the sources cited here and checked against its claim ledger before publication. How we use AISend a correction
What happened
- The Prairie Pothole Region holds an estimated 5 million to 8 million glacier-formed wetlands across Canada and the US, many in shallow depressions with no permanent link to a river.
- Snow persistence was an important driver of how much wetland area flooded, and climate's effect on river flow appeared to run largely through the wetlands.
- In one Saskatchewan catchment, annual runoff stayed very low until inundated wetlands covered about 1.7% of its area, then rose abruptly.
- Among catchments where fill-spill dynamics dominated, 64% showed similar threshold behaviour, with runoff jumping once enough wetlands filled and spilled.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision A runoff forecast driven only by snow and rain totals gives one answer per input, but these catchments can return very different flows from the same input, so planners must choose whether to track wetland fullness as well.
- exposure Catchments dense with geographically isolated wetlands showed the strongest buffering, so plans there depend most on a storage variable that climate-only models leave out.
- capability Because inundated area can be estimated from satellites every year, wetland state is something forecasters can observe and add to models, though the gain from doing so has not been reported.
The study was set up as a contest, run one catchment at a time. For each of 109 catchments, the team asked which yearly series tracked runoff most closely. One candidate was inundated wetland area, estimated from satellites. The others were climate indicators such as aridity, snow persistence and rainfall [2][3]. Wetlands came out ahead in 69% [5]. That is about 75 catchments. In the other 34 or so, at least one climate indicator was tied to runoff at least as closely [14][15].
The thing this doesn't tell you is how wide the wetland margin is, or how much it would improve a forecast. The paper appeared in Communications Earth & Environment [4]. The fuller account comes from one of its authors, writing in the first person for phys.org, and it describes associations, not predictions [2].
The two candidates are also not independent, because climate is what fills the wetlands in the first place [7]. "This does not mean climate is unimportant," the author wrote [6]. I think the fairest summary is that inundated area is the better one-number description of a year. It reflects that year's water and also whatever the depressions were already holding.
The physical picture is fill-spill hydrology. A depression first stores whatever reaches it. When it is full it spills into the next one, connections spread, and more of the catchment starts sending water to the river [8]. "I like to think of these wetlands as hydrologic switches," the author wrote [9]. "The same climatic input can therefore produce very different river responses depending on the state of the wetland network." [10]
The 64% threshold figure needs care with its denominator. It is a share of the catchments where fill-spill dynamics dominated, a subgroup of the 109, so it cannot be applied to the whole set [12]. Buffering was strongest where geographically isolated wetlands were more abundant [13]. Those wetlands look cut off from rivers in dry periods, but their spare storage helps decide when wider connections switch on [13].
For water planning, the thresholds matter more than the headline share. A climate-only model turns more snow into more flow along a smooth curve. Fit that curve to a catchment with a spill point like the Saskatchewan one, and it will overpredict in the low-runoff years below the line and underpredict in the year flow jumps [11]. I'd expect wetland state to add the most skill in catchments dense with isolated wetlands and in years close to the spill point. The record is long enough to check this: 38 years allows a model to be fitted on some years and scored on the rest against a climate-only baseline [2].
What to watch
- A held-out-years test that scores streamflow forecasts with satellite wetland inundation against climate-only versions in the same 109 catchments.
- Whether spill thresholds like Saskatchewan's 1.7% can be estimated in advance from a catchment's wetland storage, or only fitted after the fact.
- Whether operational forecasters in the Prairie Pothole Region begin adding wetland inundation as an input.