The Rain After the Fire

Tuesday, September 22nd, 2026 — The Thomas Fire started north of Santa Paula, California, on December 4, 2017, and burned for more than a month across the mountains above Santa Barbara and Ventura counties. By early January the flames were mostly out. The neighborhoods directly below the burn, including Montecito, had spent weeks on evacuation orders and had largely been spared. People were back in their houses.

Before dawn on January 9, 2018, a short, intense band of rain fell on those slopes. The burned ground did what burned ground does. Soil and rock that the vegetation had held in place came off the mountain in a slurry, funneled into the creek channels, and hit the alluvial fans where the houses were. Boulders the size of cars moved through living rooms. Twenty-three people were killed. More than 100 homes were destroyed, and hundreds more were damaged. A large share of the people who lost houses that morning had not lost them to the fire. They lost them to the rain that came after it.

The claims that followed split the street. Neighbors with the same mud in the same room, and policies that responded in opposite ways. Adjusters who work wildfire need the sequence, because the fire file and the debris-flow file are different claims, often on different policies, and the second one is frequently denied for reasons the policyholder will not believe until someone walks them through the words.

What the fire does to the slope

Chaparral and forest hold soil. Roots bind it, litter slows runoff, and the canopy breaks the impact of rain. A high-intensity fire removes that in a few hours. It also bakes the top layer of soil. Waxes from burned vegetation move into the soil and coat the particles, so the surface sheds water instead of absorbing it. Geologists call the result hydrophobic soil. The U.S. Geological Survey treats recently burned, steep basins as debris-flow hazards for years afterward, and the trigger is often a short burst of heavy rain rather than a multi-day flood. A short, hard shower, well under an inch, is enough on the wrong basin. The water cannot soak in, so it runs off and takes the loose soil, ash, and rock with it. In a channel, that mixture gains speed and depth and comes out the mouth of the canyon as a debris flow.

A debris flow is a moving mass of water, mud, and rock, with enough water to behave like a fluid and enough solid material to knock a house off its foundation. That mix is why the coverage argument exists. Flood policies and homeowners policies draw the line in different places, and a debris flow sits on the line.

The hazard does not end with the first storm. Burn scars remain unstable through several rainy seasons, until vegetation returns and the hydrophobic layer breaks down. A house below a scar that survived January can still be in the path the following winter. Anyone scoping a wildfire loss in steep country should be asking where the slope above the house drains, and whether the policyholder has been told that the fire was not necessarily the last event.

Three policies, one morning

Take three houses on the same Montecito street, all hit by the same flow. The coverage answer is different in each one, and the difference is in the form, not in the mud.

The homeowners policy with a standard earth-movement exclusion. The ISO homeowners form excludes earth movement, and the definition typically lists earthquake, landslide, mine subsidence, mudflow, mudslide, sinkhole, subsidence, and the erosion, sinking, rising, shifting, expanding, or contracting of earth. Most forms also carry anti-concurrent causation language, so the exclusion applies whether earth movement is the sole cause or one cause among others. A debris flow that is made of soil and rock fits that definition on most unendorsed homeowners policies. Fire is a covered peril. The fire did not knock this house down. The flow did. The fire claim, if the house also burned, is payable. The debris-flow damage to the unburned house is not, unless an endorsement or a state-specific form says otherwise.

That answer feels wrong to the policyholder, and the wrongness has a structure you should be ready to explain. They bought fire coverage. A fire happened. Then their house was destroyed. The causal chain is real. The policy does not insure every consequence of a covered event. It insures the perils it names, and it removes the perils it excludes, including when an excluded peril follows a covered one. Say that in ordinary words. "The fire is covered. Damage from mudflow and landslide is excluded even when the fire is what made the slope fail." Then stop. Do not invent a coverage theory to soften it, and do not deny the fire damage that is covered because the mudflow denial is the larger number.

The NFIP flood policy. The National Flood Insurance Program defines flood to include mudflow. The program's definition is specific. A mudflow is a river of liquid and flowing mud on the surface of ground that is normally dry, as when earth is carried by a current of water. Landslide, slope failure, and saturated earth moving without that river of mud are not flood, and NFIP does not pay them. On a true debris flow of the Montecito type, where a channelized slurry of water and soil ran down a creek and across the fan, mudflow coverage is what the flood policy is for. The adjuster still has to document that it was a flow and not a dry landslide that the rain merely accompanied. High-water marks will not look like a river flood. Look for the deposit: unsorted mud, rock, and debris, a path that follows the drainage, upstream scarring in the channel, and boulders that could not have rolled there without water carrying them.

NFIP building coverage is capped at $250,000 and contents at $100,000, and the basement limitations still apply. A flood policy can pay the mudflow and still leave a large gap. Tell the policyholder both facts.

The auto policy and everything else. Vehicles carried off or buried are usually comprehensive losses, which is the same small piece of useful news that shows up in a river flood. Detached walls, landscaping, and the pool that filled with silt follow the same earth-movement analysis as the house. Do not assume the whole property is one coverage decision. Walk it piece by piece.

What the argument turns on

Where a claim is disputed, the dispute is almost always about which word fits the physical evidence. Mudflow, mudslide, landslide, and flood are not synonyms in a policy, even though the news will use all four by the following afternoon.

• Mudflow, for NFIP, wants a current of water carrying earth across normally dry ground. A channel, a slurry, and a deposit of mixed mud and rock support it.

• Landslide and slope failure want a mass of earth that failed and moved, with water as a contributor rather than the transporting current. A section of hillside that dropped onto a house, even if it was wet, is the pattern.

• Flood, in the homeowners sense, is the rising and overflow of water the earth-movement exclusion does not already capture. On most homeowners forms both the flood and the mudflow are excluded, so the distinction matters less than it does under NFIP. On a form that excludes flood but says nothing about mudflow, or the reverse, the distinction is the claim.

• Fire and smoke remain covered wherever they can be separated. A house with a burned wing and a mud-filled ground floor needs two scopes. Combining them into one "the mountain did this" estimate is how covered fire damage gets delayed behind an excluded mudflow denial, or how excluded mud removal gets paid inside a fire estimate.

State case law has not treated every one of these facts the same way, and California in particular has a long record of fights over efficient proximate cause. If the policy has anti-concurrent causation language and the state enforces it, the exclusion holds even where the fire set the slope up to fail. If the state applies an efficient-proximate-cause statute and a covered peril is found to be the predominant cause, the analysis changes. That is a coverage question for the carrier's coverage counsel, not a judgment to make in the driveway. Your job is to describe the mechanism accurately enough that counsel can use the description. "Mud in the house" is not that description. "Channelized debris flow from the burned basin above the property, deposit of mud, rock, and boulders throughout the first floor, no fire damage to the structure" is.

Documenting it before it disappears

Debris is cleared quickly after these events, partly by the homeowner, partly by public works trying to reopen a channel before the next rain. The evidence is the deposit. Once it is hauled off, the file is an argument about memory.

Photograph the path along with the house. Where the material entered, what it destroyed on the way, the size of the largest rocks, the height of the splash or the deposit line on the walls, and the upstream channel if you can see it safely. Photograph a cross-section of the deposit if one is exposed: mud, ash, rock, and organic debris mixed together, rather than bedded like a river sand. Note whether the material is still wet, and whether there is standing water separate from the mud. Get the rainfall, which the local ALERT gauges and the National Weather Service will have, and the fire perimeter, which the incident team publishes. A map with the house, the channel, and the burn scar on it belongs in the file.

Do not walk an unstable deposit or a house that has been moved off its foundation. These scenes kill rescuers in the first day and they stay unstable. If the public works barricade is up, the photograph is taken from behind it, and the inspection waits.

Ask what the fire claim already paid. If a prior adjuster scoped ash and smoke, or a burned outbuilding, those photographs show the property between the fire and the flow. They are the baseline that keeps the two claims from being paid as one or denied as one.

The houses still below the scar

Most of the homes below a large fire are still standing when you arrive to adjust the fire. The debris flow is a forecast. USGS and the local flood-control district often publish burn-scar hazard maps within days of containment, with the basins ranked by how likely they are to produce a flow in a given storm. Local emergency management orders evacuations. The policyholder will still ask you, because you are standing there.

The honest answer is limited and useful. The slope above the house burned. Burned slopes can produce debris flows in the first heavy rain, and the hazard lasts more than one winter. Their homeowners policy is unlikely to cover that damage. A flood policy may cover the mudflow portion and will not cover a landslide. If they do not have flood coverage, the time to hear that is before the rain, not after it. Local emergency management is the source for warnings and evacuation, and you should know the name of that office before you are asked.

Write the fire estimate as a fire estimate. If you see a house sitting at the mouth of a burned channel, note it, tell the policyholder what you told the last one, and do not fold a future mudflow into the fire scope as preventive work. The policy pays for the damage that happened. The warning is a conversation. Both of them count, and they are not the same entry in the file.