Why Hillsides Burned by Wildfire Flood More Easily - and What the Numbers Show

When a hillside burns due to wildfire, the next heavy rainfall can produce a flood far larger than the same storm would have caused before the fire. New research published in Water Resources Research has put numbers to this long-suspected pattern.

Across seven catchments burned by wildfire in the western United States, the team found that post-fire flood peaks were at least twice as large as those from comparable pre-wildfire storms in three out of four cases. The largest difference recorded was 39 times greater.

The study also identified a clear pattern in the kinds of storms that produced these post-wildfire floods. They were not extraordinary weather events. Most were storms of a type that recur roughly once a year - common rainfall, made dangerous by a landscape recently burned by wildfire.

For those working on UK moorlands, where wildfire occurrence has grown over the past two decades, the findings offer a useful piece of the broader evidence base on how ground burnt by wildfire behaves under rain. It is also important to note that there is no evidence to suggest that controlled burning causes flooding.

What the researchers did

The team developed what they call a "Paired Storms Framework". Rather than comparing burned catchments to neighbouring unburned ones - a method that struggles when storms vary across short distances - they compared each post-fire flood to floods from earlier, similar storms in the same catchment, before it burned. Same hills, same storm characteristics: only the burn status differed.

They identified 26 large post-wildfire floods across seven catchments, ranging from 14 to nearly 3,000 square kilometres in area, with burned proportions from 10% to 100%.

For each post-wildfire flood, they searched the catchment's pre-wildfire record for storms with matching rainfall intensity, total depth, spatial coverage, season, and antecedent moisture (the rain that fell in the days before). Twenty of the 26 floods had at least one paired pre-fire storm to compare against.

What they found

Three findings stand out.

First, post-wildfire flood-producing storms shared a recognisable signature. They tended to occur in wet years, deliver high one-hour rainfall intensities and large total depths, and cover the entire catchment and burned area. Storms confined to small patches of the hillside rarely produced major post-wildfire floods.

Second, the storms themselves were not unusual. Most had return intervals of one year or less - meaning storms of that size occur roughly every year on average. The flooding was unusual; the rain was not.

Third, the amplification of flood peaks varied considerably. Across all 20 paired floods, the median post-wildfire flood was roughly four times the size of its pre-wildfire counterpart, but in some catchments the multiplier reached 39. In two cases, post-wildfire flood peaks were no larger than their paired storms - suggesting the relationship is not uniform.

The researchers also noted that flood amplification tended to decrease with time since wildfire, as vegetation recovered, and to increase with the proportion of catchment area burned.

A note on storm sequencing

One of the more practical observations concerns the order in which storms arrive. In several catchments, the first major storm after the fire produced a disproportionately large flood relative to its rainfall. Subsequent storms of similar size produced smaller floods.

The authors suggest this may reflect a "first flush" effect: loose ash and sediment, readily mobilised after wildfire, are washed out early on, after which floods become less extreme.

What this means for moorland management

The study is American, but three points have clear relevance to British moorlands.

The first is that the conditions producing post-wildfire flooding are not exotic. A common winter storm falling on a recently burned catchment is the typical trigger, not an exceptional one.

Where British moorlands have been affected by wildfire - as several have in recent dry summers - heightened flood awareness is warranted for ordinary rainfall in the months that follow, not only for major storms.

The second is that catchment-wide burns matter more than localised ones. Floods scaled with burned area in this study, supporting the case for fire management strategies - including controlled burning and firebreaks - that limit the proportion of any single catchment burning at once.

The third is that recovery takes time. The study used a three-year post-wildfire window, and amplification effects persisted across it. Restoring vegetation cover after a wildfire is not only a habitat question but a flood-risk question for downstream communities.

Don't Miss the Latest Moorland News

Get news of policy shifts, new research and land management changes that affect upland Britain, delivered free every week. If it matters on the moor, you'll hear it from us first.

Sign Up Now