Researchers led by Professor Yasuo Nihei of Tokyo University of Science, with Mr. Ryutaro Otsuka and Mr. Kengo Ohno, modeled a real 2023 flood in the Ichinomiya River Basin, where drainage channels extend into mountainous areas. They simulated rainfall runoff, river flow, and flooding under different drainage channel configurations and assessed effects on water levels, discharge, and the basin’s overall water balance.

“In typical large-scale flood analyses, rivers are primarily modeled explicitly, while the influence of smaller channels and drainage pathways upstream is often represented indirectly through runoff parameters. However, since channels play a crucial role in draining rainwater into rivers, we believed that directly reflecting them in the model would allow us to reproduce water movement within the basin more closely to real-world conditions while reducing reliance on parameter adjustments,” says Prof. Nihei.

The Ichinomiya River Basin covers 203 km2 in Chiba Prefecture, Japan, and has experienced repeated typhoon-related flooding. After severe flooding caused by Typhoon Bualoi in 2019, flood-management measures were developed under the RBDRSA project. In 2023, Typhoon Yun-yeung caused another flood while these measures were still being implemented, providing a real event for the researchers to model.

The team divided the watershed into cells representing the land surface, small drainage channels, and larger rivers, and defined drainage channels using a minimum upstream drainage area (Amin), a threshold determining which parts of the watershed are represented as channels. They ran the model with seven configurations, using Amin values from 1.00 to 0.025 km2, with lower values progressively adding smaller channels.

The drainage network strongly affected water movement. Including more small channels provided additional pathways for rainwater to reach rivers, reducing water retained on slopes and inundation depth in flooded areas. It also increased water volume in channels and rivers and caused river water levels and discharge to rise earlier during the initial flood stage, highlighting the importance of drainage configuration in determining both the magnitude of simulated flooding and the timing of water-level rises.

Of the seven configurations, an Amin of 0.1 km2 was sufficient to reproduce the main drainage channels, while values between 0.1 and 0.2 km2 provided good accuracy for reproducing river water levels. This threshold range was identified for the Ichinomiya River Basin and the September 2023 flood; appropriate values may differ depending on watershed topography, drainage channel characteristics, and modeling conditions. Including even smaller channels did not always improve the model and could reduce accuracy by representing channels that do not actually exist.

According to the researchers, the threshold upstream drainage area may be a more appropriate measure than conventional drainage density for determining how far upstream channels should be represented, particularly when topographical characteristics of the watershed are important. However, the appropriate value depends on watershed characteristics, and further validation across different river basins and flood events is necessary.

“The Amin method for designing waterway networks presented in this study can be used to improve models that comprehensively evaluate rainfall runoff, river flow, and flooding across an entire watershed,” says Prof. Nihei. These findings could help the RBDRSA project improve evaluation of flood-management measures, contribute to more reliable flood risk assessments, and support quantitative evaluation of basin-wide measures. Further validation in other watersheds and under different flood conditions could support more quantitative assessments of their effectiveness under future climate change, although this study did not directly simulate future climate scenarios.

The study, titled “Impact of drainage channel network on hydrologic and hydraulic responses over the entire basin,” was made available online on September 30, 2026, and will be published in Volume 68 of the Journal of Hydrology: Regional Studies on December 1, 2026. The work was supported by the River Fund from the Japan River Foundation, and the authors declare no known competing financial interests.

The question of how much detail a drainage network requires in a flood model touches on a long-standing challenge in hydrology. Watersheds are intricate systems in which rainwater follows countless paths before reaching a main river, and no model can represent every one of those paths explicitly. Modelers therefore face a trade-off: representing more channels captures more of the real drainage structure but demands more computational resources and more detailed input data, while representing fewer channels forces the model to compensate through calibrated parameters that may not hold under different flood conditions. The approach tested in the Ichinomiya River Basin offers a way to navigate this trade-off by letting a single, physically meaningful threshold decide where explicit channel representation begins.

The concept of a minimum upstream drainage area has deep roots in terrain analysis. When digital elevation models are used to trace the flow of water across a landscape, every point in the watershed accumulates drainage from the land upstream of it. Setting a threshold on that accumulated area effectively draws a line between hillslope and channel: cells draining areas larger than the threshold are treated as channels, while the rest are treated as land surface. Lowering the threshold pushes that line further upstream, extending the modeled network into progressively smaller headwater hollows. This is why the seven configurations tested by the researchers produced networks of increasing extent as the threshold dropped from 1.00 km2 to 0.025 km2.

Traditional drainage density, defined as the total length of channels per unit area of the basin, has long served as a summary measure of how finely dissected a landscape is. However, drainage density is a single aggregate number that says little about where channels sit within the basin or how the network connects hillslopes to the main river. The researchers’ suggestion that a threshold drainage area may be more appropriate than drainage density reflects the fact that flood response depends not just on how much channel exists, but on where it exists. In steep, forested headwaters, the upstream extent of the network determines how quickly rainwater is collected and delivered to larger channels, which in turn shapes the timing of downstream water-level rises.

The finding that adding ever smaller channels can degrade model accuracy deserves particular attention. It might seem intuitive that more detail is always better, but the drainage networks derived from elevation data are only as good as the data themselves. At fine thresholds, the extracted network can include features that do not correspond to real channels on the ground, introducing artificial flow pathways that route water in ways the actual landscape does not. In effect, the model can become over-specified, representing a drainage system that exists only in the digital terrain data rather than in the watershed itself. This means the goal is not maximum detail but the right level of detail, matched to both the terrain and the quality of the underlying data.

The timing effects observed in the simulations carry practical weight for flood forecasting. River water levels that rise earlier in a simulation give forecasters an earlier signal of an emerging flood, and the study showed that the configuration of the drainage network directly controlled when simulated levels and discharge began to climb during the initial stage of the Typhoon Yun-yeung flood. A model whose network is too sparse may delay the simulated response, potentially underestimating how quickly a river will react to intense rainfall. Because the researchers modeled an actual flood event, their results provide a concrete demonstration that network design influences not only how much flooding is predicted but when it is predicted to begin.

The Ichinomiya River Basin offers a useful setting for this kind of test precisely because its drainage channels extend into mountainous terrain, where the question of upstream extent is most acute. In lowland basins, channels are typically well mapped and their representation is less ambiguous. In mountainous and forested headwaters, however, the network can stretch far upstream, and the point at which it should be truncated is far less obvious. The threshold range of 0.1 to 0.2 km2 identified for this basin gives modelers working in similar terrain a starting point, while the authors’ caution that appropriate values depend on watershed topography and channel characteristics underscores that the threshold should be validated locally rather than transferred blindly.

The study’s limitations frame the scope of its conclusions. The threshold range was derived from a single basin and a single flood event, and the researchers note that further validation across different river basins and flood conditions is necessary before the method can be applied with confidence elsewhere. The work also did not directly evaluate early warning systems or evacuation decisions, even though more reliable flood simulations could ultimately inform such applications, and it did not simulate future climate change scenarios. What the study does establish is a principled, transferable framework: by tuning a single physically interpretable parameter against observed water levels, modelers can determine how far upstream a drainage network needs to be represented to reproduce basin-wide hydrologic and hydraulic behavior without relying on arbitrary choices or excessive parameter adjustment.

For basin-wide flood management programs such as RBDRSA, the practical implication is that the fidelity of the drainage network is not a secondary modeling detail but a first-order control on simulated flood behavior. As computational models take on a larger role in evaluating flood-management measures before they are built, ensuring that those models represent the watershed’s drainage structure at an appropriate level of detail becomes a prerequisite for trustworthy quantitative assessment.

Subject of Research: Flood modeling

Article Title: How much of a basin’s drainage network should flood models capture?

Article References: How much of a basin’s drainage network should flood models capture?. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: flood modeling, drainage network, Ichinomiya River Basin, hydrology, Tokyo University of Science, rainfall runoff, river discharge, inundation, RBDRSA, watershed, much, basin

The new technique, called the Robust Variance-based Event Identification Method, or RVEIM, was developed by Mohammad Masoud Mohammadpour Khoie, Danlu Guo, and Conrad Wasko and published in the journal Hydrology and Earth System Sciences. Its central insight is that a runoff event is best detected not by hunting for peaks in the streamflow record, but by watching for sudden surges in the variance of that record. When streamflow variance briefly and clearly exceeds what is normal for a catchment, the stream is doing something unusual, and that something is almost always a response to rain. By anchoring event detection to this statistical signature of physical process, RVEIM needs only two user-defined parameters, a dramatic reduction from the half-dozen or more that conventional methods demand.

The problem RVEIM addresses is one that has quietly undermined event-based hydrology for years. Because no instrument directly measures where a runoff event begins or ends, researchers must rely on rules: a threshold to filter out trivial flows, a minimum spacing to keep consecutive events independent, a search window to match each flood peak with its causative storm. None of these rules has a ground truth to validate against, so their values are chosen by expert judgment or local calibration. Sensitivity analyses have shown that even within accepted ranges, changing these values can alter the number, length, and volume of identified events by amounts comparable to the climatic differences between regions. In one striking example cited by the team, simply reversing the direction of the pairing search, from runoff to rainfall rather than rainfall to runoff, introduced as much variability as differences in climate itself.

Worse, the conventional two-step workflow, detecting runoff events first and pairing them to rainfall afterward, can produce physically impossible results. When a mismatched pair is created, the calculated runoff coefficient, the ratio of runoff volume to rainfall volume, can exceed one, implying that more water left the catchment than fell on it as rain. Such events are clear flags of identification error, yet some methods generate them routinely. RVEIM sidesteps this by detecting and pairing simultaneously: it identifies independent rainfall events first, then searches for the corresponding streamflow response within a window whose length is not fixed by the user but computed from the catchment’s characteristic rainfall-to-runoff lag plus the duration of each individual storm. This time-variant window mirrors the causal chain of runoff generation and allows the method to capture both brief, sharp responses and long, multi-peak events without extra tuning.

The variance threshold at the heart of the method is also derived rather than imposed. The team first applies the widely used Lyne and Hollick digital filter to estimate baseflow, the slow groundwater-fed component of streamflow. They then use a median absolute deviation test to exclude baseflow values inflated by event conditions, and take the variance of the remaining, baseflow-dominated records as the threshold below which streamflow behavior counts as normal. Because this threshold emerges from each catchment’s own data, it transfers across climates without recalibration. In tests across eight representative catchments spanning desert, grassland, temperate, subtropical, tropical, and equatorial zones, the variability of event characteristics under RVEIM stayed within roughly fifteen percent of the mean, while two benchmarking methods, a local maxima approach and the more elaborate Detrending Moving-average Cross-correlation Analysis, exhibited uncertainties of forty-three to ninety-three percent.

Physical plausibility improved as well. Across the representative catchments, the percentage of events with runoff coefficients exceeding one remained below about four percent under RVEIM, and was essentially zero for the DMCA benchmark, whereas the local maxima method produced consistently higher and more variable rates. The differences mattered most in wet catchments such as those in western Tasmania, where frequent storms and high soil moisture make event separation genuinely difficult. There, the three methods diverged sharply in their pictures of catchment behavior: RVEIM attributed far fewer events to near-zero runoff coefficients than its competitors, a result consistent with the hydrological reality of a saturated, fast-responding landscape.

With its robustness established, the team unleashed RVEIM on all 467 unregulated Hydrologic Reference Stations monitored by the Australian Bureau of Meteorology, catchments with at least thirty years of quality streamflow record and minimal human interference. The resulting atlas reveals a continent organized by water. Runoff coefficients climb toward the coasts and reach their highest values in Tasmania, where abundant rainfall, low evapotranspiration, and saturated soils conspire to send most event rainfall straight into streams. In the arid interior, by contrast, most storms vanish into infiltration and evaporation, leaving runoff coefficients clustered near zero. Northern Western Australia presents an intriguing anomaly: despite modest rainfall event volumes, its catchments yield substantial runoff, a consequence of low evapotranspiration and high antecedent soil moisture that leave little room for losses.

The most consequential finding is a systematic climate gradient in the shape of the runoff coefficient distributions. In desert and grassland catchments, the empirical distribution rises steeply at near-zero values, indicating that the vast majority of events produce negligible runoff, a signature pointing to infiltration-excess processes in which rain falls faster than parched soils can absorb it only during intense bursts. In temperate, subtropical, and equatorial regions, the distributions rise gradually toward higher coefficients, the hallmark of saturation-excess runoff from already-wet catchments. Tropical catchments occupy an intermediate, mixed regime. Moving from dry to wet climates, the distributions also become less skewed and more uniform, echoing patterns previously documented in Austria and the European Alps and suggesting the gradient is a global hydrological regularity rather than an Australian curiosity.

The implications reach well beyond academic taxonomy. Flood estimation, infrastructure design, and projections of how catchments will respond to a changing climate all depend on event-scale runoff coefficients, and biased event identification can distort all of them. By capturing the full spectrum of event magnitudes rather than privileging large floods or short responses, RVEIM offers a consistent yardstick for comparing catchments across regions and for tracking how runoff generation shifts as droughts, land-use change, and warming reshape the water cycle. The method has been released as a function in the open-source hydroEvents R package, and its developers argue that because its parameters are unlikely to require adjustment between climates, it should transfer readily to continents beyond Australia. If it does, hydrologists worldwide may finally agree on where one storm ends and the next river begins.

Subject of Research: A parsimonious rainfall-runoff event identification method applied across 467 Australian catchments to characterize event-scale runoff processes

Article Title: Characterising runoff processes for Australia: insights from a parsimonious rainfall-runoff event identification method

Article References: Mohammadpour Khoie, M. M., Guo, D., & Wasko, C. (2026). Characterising runoff processes for Australia: insights from a parsimonious rainfall-runoff event identification method. Hydrology and Earth System Sciences, 30(18), 5947-5970. https://doi.org/10.5194/hess-30-5947-2026

Image Credits: AI Generated

DOI: 10.5194/hess-30-5947-2026

Keywords: hydrology, rainfall-runoff, runoff coefficient, event identification, catchment response, Australia, baseflow separation, flood estimation, climate gradient, streamflow, RVEIM, large-sample hydrology

The scale of the analysis is extraordinary. Drawing on more than two million rainfall–runoff events, the researchers assessed the behavior of more than 80,000 catchments across 97 countries. A catchment, in hydrological terms, is the area of land that drains to a particular point in a river network, and understanding how it converts rain into streamflow is one of the oldest problems in the discipline. What makes this study unusual is its focus. Rather than mapping where rivers are or how much water they carry, the team mapped how consistently landscapes perform that conversion from one storm to the next. The answer, it turns out, varies dramatically across the globe in ways that follow identifiable climatic and physical patterns.

The researchers sorted catchments into three behavioral classes. In simple catchments, a given amount of rain delivers roughly the same amount of water to the river every time, producing a predictable, repeatable response. In complex catchments, similar amounts of rainfall can generate very different runoff responses, with the outcome shaped by antecedent conditions and by which parts of the landscape are actively supplying water to the stream during any given storm. Between these extremes sits a third, intermediate group, which the researchers found can respond to rain in more than one stable way, perhaps two or three distinct modes, and switches among them depending on circumstances. The geographic verdict was stark: complex behavior covered 87 percent of the land area evaluated, some 121 million square kilometers, while simple catchments accounted for just 1.5 percent.

That asymmetry carries real consequences. Simple catchments are so rare worldwide that they register almost as curiosities, concentrated along the Coast Mountains of Canada’s British Columbia and the U.S. Pacific Northwest, with smaller pockets in the northern and western United Kingdom, parts of Ireland and northern Spain, and in Tasmania and New Zealand. Forested catchments near Bamfield on Vancouver Island’s west coast, where rain arrives almost every week in winter, exemplify this rare class, and this stretch of coast holds one of the largest concentrations of simple catchments anywhere in the study. Complex catchments, by contrast, dominate across Africa and much of Asia, across most of France, Spain, eastern Germany and Denmark, across the central United States and across most of South America. Intermediate catchments are common across the eastern United States, where landscapes appear to hold several stable response modes in reserve.

What separates the predictable few from the unpredictable many? The strongest single factor was not how much rain a region receives but how often meaningful rainfall occurs over time. Where rain arrives regularly, soils, subsurface stores and the pathways connecting the landscape to the stream remain in similar states from one storm to the next, so a comparable input of rain is more likely to produce a comparable output of runoff. Where rain is less persistent, a catchment repeatedly swings between wet and dry conditions. A storm arriving after a dry spell may largely soak into soils and subsurface storage while sending relatively little water to the river, whereas an identical storm arriving when the landscape is already saturated can send a far larger share of that rain straight downstream. The rhythm of rainfall, in other words, sets the variability of response.

Dr. Ameli framed the finding as a challenge to a long-standing mental model in hydrology. We often imagine a catchment as a pipe, he noted, in which a certain amount of rain in produces a predictable amount of runoff out. But most catchments behave more like dynamic systems with a memory, in which how much of today’s rainfall becomes runoff depends both on conditions created by earlier weather and on which parts of the landscape supply water to the river during the storm. It is not simply the amount of rain that matters, he explained; it is also the rhythm of rainfall. Persistent rain keeps the landscape in a more consistent state, while alternating wet and dry conditions allow the same catchment to behave very differently from one event to the next.

Climate did not act alone. The researchers also found that a region’s overall water availability strongly distinguishes simple from complex catchments, and they examined this using established measures including an aridity index, which compares how much water the atmosphere can draw off a landscape with how much rain actually falls on it, and the long-term balance between rainfall and the water lost to evaporation or taken up by plants. Catchments with simpler, more consistent behavior were generally found in wetter and more humid conditions, while complex catchments were more commonly associated with drier settings, where less water remains once evaporation and plant use are accounted for. Taken together, these results suggest that overall water availability establishes the broad climatic setting, and that rainfall rhythm determines whether the landscape is likely to remain in a stable hydrologic state or alternate among different states.

The physical character of a catchment added a further layer of refinement at local and regional scales. Although climate emerged as the dominant influence, terrain mattered too. Simple catchments tended to occur in steeper landscapes, where slopes can move water rapidly and maintain more direct connections between hillslopes and stream channels, contributing to a more consistent runoff response from one storm to the next. This helps explain why the steep, rain-drenched mountains of the Pacific Northwest and coastal British Columbia host some of the world’s most predictable catchments, while gentler terrain in seasonally variable climates produces the complex, state-dependent behavior that dominates the global map.

The practical payoff of the work lies in modelling and prediction. The study and its accompanying products give hydrologists a common language for comparing catchments in different parts of the world: two catchments on opposite sides of the planet may convert rain to runoff in the same way, and the classification makes that visible at a glance. It also offers a practical guide to how a catchment should be represented in a computer model. A simple catchment can be captured by a straightforward model, while a complex one demands a model that allows different parts of the landscape to take the lead at different times, tracking how active runoff zones shift as conditions change. For flood forecasters, water managers and anyone charged with predicting how a river will respond to the next storm, that distinction could be decisive.

To make the results immediately usable, the team has released an interactive global map of the classification along with a free web application that returns a classification for any user-supplied catchment boundary, and both are publicly available. For a field in which much of the world’s stream network is ungauged, meaning no instrument continuously records its flow, a tool that infers behavioral class from climate and landscape attributes alone extends the reach of hydrological knowledge far beyond the instrumented record. The research, published as a peer-reviewed observational study in Nature Water under the title describing a global classification of hydrologic functional diversity in gauged and ungauged catchments, appeared on 11 September 2026, and its authors declare no competing interests. Its central message is likely to resonate well beyond hydrology: the landscapes that deliver our water are not passive pipes but systems with memory, and reading that memory correctly may be the key to anticipating the floods of the future.

Subject of Research: Global classification of rainfall–runoff behavior and hydrologic functional diversity in catchments

Article Title: Why most rivers don’t respond to storms the same way twice

Article References: Why most rivers don’t respond to storms the same way twice. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: hydrology, catchments, rainfall-runoff, flooding, Nature Water, runoff prediction, aridity index, climate, watersheds, flood modelling, University of British Columbia, global mapping