Rain falling onto snow can create a flood risk greater than the rainfall alone suggests. Rain-on-snow flooding occurs when liquid precipitation combines with water released from an existing snowpack and reaches rivers faster than the landscape can store or convey it. Yet rain does not always produce rapid melting: a cold snowpack may initially absorb or refreeze incoming water. The outcome depends on snow temperature, stored water, wind, humidity, rainfall and the condition of the ground below. Understanding the process means treating snow as an active store of both water and energy, rather than a white surface that disappears whenever rain arrives. That distinction helps explain why similar-looking winter storms can have very different downstream consequences.
Key takeaways
- Rain-on-snow events combine new rainfall with an existing frozen water store.
- A cold snowpack can retain or refreeze water before releasing it.
- Warm humid air and wind can contribute strongly to melt energy.
- Wet or frozen ground can increase rapid runoff under suitable conditions.
- Flood severity depends on catchment-wide timing, not rainfall totals alone.
What is rain-on-snow flooding?
A rain-on-snow event begins when rain reaches an existing snow cover. Flooding is one possible consequence, not part of every event. The snow can melt, store liquid water, transmit rainfall, refreeze it or undergo several of these processes in different layers.
When substantial water leaves the snowpack and flows towards channels, rainfall and snowmelt contributions can combine. If delivery exceeds the available storage and channel capacity, river levels can rise sharply.
In colder environments, an important outcome may instead be an ice crust or internal ice layer. Such icing can affect ecosystems and travel without producing a major downstream flood. The term describes an interaction, not a single inevitable result.
Snow depth is not the water supply
Snow contains both ice and air. A deep layer of light powder can contain less water than a shallower dense pack. Snow water equivalent measures the amount of liquid water represented by the snow, making it more useful for water accounting than depth alone.
The distribution matters as much as the average. Wind, elevation, vegetation and earlier weather can leave some parts of a catchment deeply covered while others have little snow. A measurement near one road may poorly represent mountain slopes.
A rain-on-snow forecast therefore needs to know where rain will overlap with stored snow water. Rain falling below the snow-covered zone and snow falling above it produce a different response from widespread rain across an extensive existing pack.
Cold content: why melting may be delayed
Snow below its melting point must gain energy before much of it can melt. This thermal deficit is often described as cold content. Incoming rain can cool and freeze within the pack, releasing latent heat while adding ice.
As the snow warms and its capacity to retain liquid changes, more water may pass through. A pack already near the melting point can respond differently from one that has experienced a prolonged cold spell.
Different layers can behave differently at the same time. Surface melt does not establish that the full depth is warm or that water is already leaving its base. Forecasting must consider the internal state as well as surface weather.
Does the rain itself supply all the heat?
No. Rain carries heat, but atmospheric exchanges can be very important. Warm moving air transfers sensible heat, while condensation of water vapour onto a cold surface releases latent heat.
Clouds and the atmosphere also emit thermal infrared radiation. During a warm humid storm, those exchanges can support melting despite limited sunshine. Wind affects turbulent exchange between the air and the snow surface.
The balance differs among events. It is therefore misleading to imagine rain-on-snow flooding as simply warm raindrops dissolving snow, or to assume one energy source dominates everywhere.
A calculation that puts rain heat in perspective
Take a hypothetical twenty millimetres of rain arriving at 5°C onto snow at 0°C. That rain represents twenty kilograms of water per square metre. Cooling it to 0°C releases about 418,000 joules per square metre, using water’s approximate heat capacity.
If all that energy melted ice already at 0°C, it would melt about 1.25 kilograms per square metre, equivalent to 1.25 millimetres of water. This calculation isolates the sensible heat carried by rain.
It does not include condensation, air-to-snow heat transfer, radiation, ground heat or refreezing. Nor does it predict the event’s total runoff. Its purpose is to show why rainfall heat alone may explain only part of the observed snow response.
Water can be stored and routed inside snow
Liquid water occupies spaces between ice grains and can move unevenly through the pack. Layering, ice lenses and differences in grain structure influence the routes. Some water may travel rapidly along preferential pathways.
Refreezing changes both the water balance and the structure of the pack. It removes liquid temporarily and releases heat, potentially making later water transmission easier once the cold content is reduced.
As a result, the timing of water leaving the snow can differ from the timing of rainfall. A delayed pulse can arrive after the heaviest rain, complicating interpretation based only on a rain gauge.
What happens when water reaches the ground?
Water reaching the ground can infiltrate, fill depressions, move through soil or run over the surface. Existing soil moisture affects how much additional storage remains.
Frozen ground can restrict infiltration, particularly where ice blocks pores, but not all frozen soils behave identically. Snow insulation, soil structure, ice content and thawing influence the response.
Earth Site’s explanation of water storage and transfers provides the wider framework. Snow release is an input to the catchment system, not automatically an equal volume arriving at the river at the same moment.
Why elevation and the rain–snow boundary matter
Mountain storms can deliver rain at lower elevations and snow higher up. If warmer air raises that transition, a larger snow-covered area may receive rain. The overlap can greatly change the area contributing liquid water.
A single temperature threshold is a simplification. The vertical temperature profile, humidity and melting of falling snow influence precipitation type. Forecasts use multiple observations and models rather than only the temperature at one valley station.
Catchments with a broad band of land near the transition elevation can be especially sensitive to its movement. Geography determines how much snow-covered area is exposed, so the same regional storm need not affect neighbouring basins equally.
How forest cover complicates the picture
Trees alter snow accumulation by intercepting precipitation and changing wind distribution. Canopies also shade snow, emit thermal radiation and modify turbulent air movement.
Openings may accumulate a different snowpack and experience different wind exposure from adjacent forest. During a warm storm, these differences can affect both the available water and the energy supplied for melt.
The net effect is site- and event-dependent. Research on clearcut openings in the western Cascade Range examined these processes under specific conditions; it should not be converted into a universal claim that every forest change produces the same flood response.
From hillside release to a river peak
Water from different parts of a basin takes different times to reach the outlet. Flood severity can increase when tributary contributions overlap closely, but diminish when storage and travel times spread them out.
Seasonal river-flow patterns show how snow storage shifts water delivery through time. Rain-on-snow events can compress part of that release into a shorter period, though the amount depends on the conditions.
Reservoir operations, wetlands, floodplains and channel restrictions modify the downstream response. Ice jams can add another hazard, but they are a distinct mechanism and are not required for rain-on-snow flooding.
What forecasters need to observe
Relevant information includes snow water equivalent, snow temperature, rainfall amount and type, wind, humidity, soil wetness, ground frost and river levels. Spatial coverage matters because mountain conditions can vary sharply over short distances.
Models combine these observations with estimates of snow energy and water movement. Some use temperature-index approaches; others represent energy exchanges more explicitly. Both need testing against observed events.
The article on stream gauges and river-model calibration explains why downstream observations are essential. A model should reproduce timing and volume, not merely produce a plausible-looking peak.
Two contrasting hypothetical storms
In the first scenario, rain falls on a cold, dry snowpack above relatively dry ground. Some rain refreezes, and much of the later drainage infiltrates. The snow changes substantially, but the river response remains limited.
In the second, a warm humid windy storm reaches a widespread pack already near melting. The ground is wet, and several tributaries receive rain and meltwater together. The same rainfall total could then contribute to a much larger river rise.
These are teaching scenarios, not forecasts. Their value is to identify the observations needed to distinguish outcomes. Rainfall amount alone leaves too much of the system undescribed.
Water accounting across a snow-covered catchment
Suppose a hypothetical storm delivers thirty millimetres of rain while the snowpack loses fifteen millimetres of water equivalent through melting and drainage. If other snow exchanges are negligible, forty-five millimetres is available below or around the pack over the relevant area.
That is not automatically forty-five millimetres of river runoff. Some water can refill soil storage or remain in surface depressions. Conversely, river flow may also include groundwater and water released from earlier events. State the boundary before adding the terms.
Area weighting is essential where only part of a catchment has snow. If the fifteen-millimetre snow contribution applies to half the basin, its catchment-average contribution is 7.5 millimetres, not fifteen. Rainfall may also vary across the basin.
This simple exercise explains why a snow-station reading cannot be added directly to a valley rain total and labelled a forecast. A credible estimate needs spatial representation, a clear interval and the routing of the resulting water.
Conclusion
Rain-on-snow flooding is a compound process involving stored snow water, energy exchange and catchment routing. Rain can add water while warm humid conditions help release an older store, but cold snow and available ground storage can delay or reduce the response. Understanding those controls makes winter flood explanations more useful than assuming rain always melts snow immediately or that deep snow alone guarantees a flood.
Frequently asked questions
Does every rain-on-snow event cause flooding?
No. Rain may be retained or refrozen, and the ground may absorb later drainage. Flooding requires sufficient water delivery relative to the available storage and channel capacity.
Can flooding continue after the rain stops?
Yes. Water can continue draining from snow and moving through tributaries. The river peak may lag behind rainfall because of storage and travel time.
Is warm rain always the main source of melt energy?
No. Turbulent heat exchange, condensation and radiation can be important or dominant. Their relative contributions vary with weather and snow conditions.
Why do nighttime temperatures matter?
Overnight cooling can reduce melt and encourage refreezing. Persistently warm humid nights may allow continued energy input and water release instead of a strong daily pause.
Can this article predict a local flood?
No. Local risk requires current forecasts, observations and official warnings. The explanation identifies the processes involved, not the conditions at a particular river today.
