6.5 Weather-Related and Natural-Disaster Deaths
Key Takeaways
- Disaster mortality is classified as directly related when the environmental force itself causes death, indirectly related when unsafe or unhealthy conditions created by the event cause it, and possibly related when the association is plausible but cannot be confirmed.
- Generator carbon monoxide poisoning during post-storm power outages is the archetypal indirect disaster death, and it clusters in the days after landfall rather than during the storm.
- Heat-related deaths require scene-derived evidence — ambient and indoor temperature at discovery, air conditioning status, window and clothing state — because there is no specific autopsy finding and postmortem core temperature is unreliable after delayed discovery.
- The interval between a disaster and a death does not by itself sever the causal relationship: injuries and conditions arising from the event remain event-related when the sequence is unbroken.
- Weather events overwhelm offices in surges, so the investigator role includes rapid case ascertainment, consistent certification language, and reporting to public health surveillance so the count can guide response.
6.5 Weather-Related and Natural-Disaster Deaths
ABMDI groups "investigate weather-related deaths" with plane and train crashes and with bombings under Investigating Multiple Fatalities. That is the correct frame. A heat wave is a mass-fatality incident that arrives slowly, produces its decedents in separate apartments across a city, and never generates a single scene.
Direct, Indirect, and Possibly Related
Disaster mortality surveillance uses a three-way classification, and getting a death into the right bucket is a certification and a public-health task at once.
Directly related deaths are caused by the environmental force of the event itself: drowning in storm surge or floodwater, blunt and penetrating trauma from wind-borne debris or structural collapse, burns and smoke inhalation in a wildfire, lightning strike, burial in a landslide.
Indirectly related deaths are caused by unsafe or unhealthy conditions that the event created: carbon monoxide poisoning from a generator or a charcoal grill used indoors, myocardial infarction while clearing debris, a death from an untreated chronic condition when dialysis or oxygen supply failed, a motor vehicle crash at a dark intersection with the signals out, a fall while boarding a window, electrocution on a downed line, carbon monoxide from a vehicle running in a snow-blocked driveway.
Possibly related covers deaths where the association is plausible but cannot be established — often decomposed decedents found after an event with no scene evidence tying the death to it.
The classification decision is tested by asking whether the death would have occurred when it did if the event had not happened. As in all manner determinations, an interval between the event and the death does not by itself break the causal chain: a man who sustains a crush injury in a tornado and dies of sepsis five weeks later remains a tornado death.
Event-Specific Mechanisms
Heat
Heat waves kill more people in the United States in a typical year than any other weather hazard, and they kill quietly. The decedents are disproportionately elderly, socially isolated, poor, chronically ill, psychiatrically medicated, and living on upper floors without functioning air conditioning. Several drug classes impair thermoregulation: anticholinergics, antipsychotics, tricyclic antidepressants, diuretics, beta blockers, and stimulants including methamphetamine and cocaine.
There is no specific autopsy finding for hyperthermia. The determination is built from the scene, which means the investigator owns it:
- Ambient outdoor temperature and indoor temperature at discovery, measured and recorded, not estimated
- Air conditioning present, functioning, and switched on; fans running; windows open or sealed; heating inadvertently on
- Decedent's clothing and bedding — heavy clothing or layered bedding in a hot room
- Time last known alive and the duration of the heat event
- Antemortem core temperature if EMS or a hospital recorded one, which is far more valuable than any postmortem measurement
Postmortem core temperature is of limited value after delayed discovery, because a body in a hot room continues to gain heat. Decomposition is accelerated and may be advanced within a day, which itself is scene evidence of a hot environment.
Hurricanes and tropical systems
Direct deaths are dominated by drowning in storm surge and freshwater flooding, with trauma from structural failure and wind-borne debris second. The indirect toll then builds over the following days and often exceeds the direct toll: generator carbon monoxide poisoning, cardiac events during cleanup, falls from roofs and ladders, chainsaw injuries, electrocution from downed lines and from re-energized equipment, interrupted dialysis and oxygen therapy, and deaths during evacuation of medically fragile patients from care facilities.
Generator carbon monoxide deaths concentrate in the days after landfall and cluster in households, so a single case should prompt a check for others in the same structure and a public-health warning.
Tornadoes
Deaths are overwhelmingly blunt and penetrating trauma — crush injury from structural collapse, impact after being airborne, penetrating injury from wood and metal debris. Sheltering behavior is central to the investigation: whether the decedent was in a manufactured home, a vehicle, an interior room, or a below-grade shelter, and whether warning was received. Identification and recovery are complicated by bodies transported substantial distances and by severe commingling of remains with debris fields.
Floods
Drowning predominates, and in the United States a large share of flood deaths occur in vehicles driven into water of underestimated depth and current. Additional mechanisms include hypothermia in cold water, trauma from debris, and post-flood electrocution and infection. Recovery is delayed and bodies are transported by current, so the recovery location is not the incident location.
Wildfires
Deaths result from thermal injury and inhalation of smoke and combustion products, with carbon monoxide and cyanide toxicity, and from trauma during evacuation. Many decedents are found in or near vehicles on evacuation routes or in residences after failed evacuation, and the elderly and mobility-impaired are heavily overrepresented. Remains are frequently severely burned and fragmentary, so identification runs on dental comparison, DNA from dense cortical bone, and medical-device serial numbers, with careful recovery of cremated remains from the structure footprint.
Winter storms and cold
Mechanisms are hypothermia, carbon monoxide from indoor heating with unvented devices and from vehicles with blocked exhaust, myocardial infarction during snow clearing, falls on ice, motor vehicle crashes, and structural collapse under snow load. Hypothermia deaths carry two findings the investigator should recognize: paradoxical undressing, where a hypothermic person removes clothing as peripheral vasodilation produces a false sensation of heat, and terminal burrowing, where the person is found in a confined or hidden space. Neither is evidence of assault, and both are easily misread. Wischnewski spots — dark gastric mucosal erosions — support but do not prove hypothermia.
Lightning
Lightning deaths usually result from cardiac arrest and respiratory arrest rather than burns. Findings include Lichtenberg figures, transient arborescent, fern-like skin markings that fade within hours and must be photographed immediately; metallic burns where jewelry, belt buckles, zippers, or coins conducted heat; tympanic membrane rupture; magnetized or damaged personal electronics; and singed hair. Clothing may be shredded or blown off. The scene shows a strike path, a damaged tree or structure, and displaced soil or turf.
Earthquakes
Deaths result from crush injury and structural collapse, with crush syndrome and rhabdomyolysis in those extricated alive, asphyxia from entrapment in dust and debris, and secondary fires, tsunami, and landslides.
| Event | Typical direct deaths | Typical indirect deaths |
|---|---|---|
| Heat wave | Hyperthermia | Cardiac and respiratory decompensation |
| Hurricane | Storm surge and flood drowning, debris trauma | Generator CO, cleanup cardiac events, falls, interrupted medical care |
| Tornado | Blunt and penetrating trauma | Cleanup injury, electrocution |
| Flood | Drowning, often in vehicles | Electrocution, infection, hypothermia |
| Wildfire | Thermal injury, smoke inhalation | Evacuation crashes, cardiac events |
| Winter storm | Hypothermia | CO poisoning, cardiac events during snow clearing, falls, crashes |
| Earthquake | Crush injury, collapse asphyxia | Crush syndrome, secondary fire, disrupted medical care |
Surge, Surveillance, and Consistency
The operational problem in a weather disaster is surge with no single scene. Cases arrive over days from many locations, morgue capacity and transport are stressed, staff are themselves affected, and the office may be running on generator power.
Three practices carry disproportionate weight:
- Rapid case ascertainment. Establish a mechanism to capture every death that may be event-related, including deaths in hospitals, care facilities, and shelters, and flag them with a common identifier so they can be reviewed as a set rather than one at a time.
- Consistent certification language. Agree the wording before the cases arrive — how the event is named in the cause statement or the how-injury-occurred field and how direct, indirect, and possibly related cases will be distinguished. Inconsistent certification is the main reason official disaster death tolls are later revised upward by large margins.
- Report to public health in near real time. The count is not bookkeeping. Generator carbon monoxide clusters, heat deaths among residents of a specific building type, and deaths among dialysis patients are all actionable while the event is still running, and the medical examiner or coroner office is frequently the first to see the pattern.
Where local capacity is exceeded, mortuary support is requested through state emergency management and the emergency support function structure rather than by direct call, and mutual aid among neighboring offices is arranged before rather than during the event.
Four days after a hurricane makes landfall, three members of a household are found dead in a home with no power. A portable gasoline generator is running in the attached garage with the connecting door open. How should these deaths be classified for disaster mortality purposes?
An 81-year-old woman living alone is found dead during a prolonged heat wave. Which scene documentation is most important to supporting a hyperthermia determination?
A man found dead outdoors after a winter storm is partially undressed, with his jacket and shirt removed and folded nearby, and he is wedged beneath a porch. There is no evidence of assault. How should the investigator interpret these findings?
A decedent struck by lightning is examined within an hour of the strike. Which finding must be photographed immediately because it will not persist?