Breaking Investigation

Every time you fly, there's a 1 in 7 chance your airport is in the top 15 most dangerous for bird strikes

We analyzed 316,839 wildlife strike incidents reported to the FAA over 35 years. What we found should change how you think about takeoff.

316K

Incidents analyzed

Years of FAA records

500%

Increase since 1990

The Big Picture

Bird strikes have exploded 500% since 1990. Here's why nobody talks about it.

FAA Wildlife Strike Database, 1990–2025 · 316,839 total incidents

In 1990, the FAA recorded fewer than 1,600 wildlife strikes on U.S. aircraft. By 2019 — before the pandemic briefly interrupted — that number had climbed past 17,000 per year. That's a 500% increase in three decades. And yet most passengers have no idea this is happening.

The dramatic rise isn't because birds are getting more aggressive. It's a combination of three factors: more flights, aircraft becoming quieter (birds can't hear them coming), and improved reporting systems that capture incidents that previously went unrecorded.

What you need to know

The COVID dip in 2020 briefly reduced strikes by nearly 50% as flights plummeted — the clearest proof that strike frequency directly tracks air traffic volume. As travel rebounded in 2022, so did strikes.

89%

Strikes cause zero damage

Most birds simply bounce off. It's the 11% that don't that make headlines.

81K

Strikes happen on approach

Landing is your most dangerous moment. That's when aircraft are lowest and slowest.

Types of incidents our AI found

Machine learning identified 4 distinct risk profiles hiding in the data.

Airport Rankings

Is your airport on the list? The 15 most strike-prone airports in America — ranked.

Total reported wildlife strikes · FAA Database 1990–2025

Denver International holds a dubious crown: more wildlife strikes than any other American airport. But the list contains some surprises — and if you're flying through Atlanta, you might want to know where Hartsfield-Jackson lands.

#	Airport	State	Total Strikes	Strike Volume

"Denver's geography puts it at the intersection of major migratory flyways — it's not poor management, it's location."

— Analysis based on FAA Wildlife Strike Database

Quick Hits · Things that will surprise you

The most dangerous moment of your flight is the last 3 minutes

81,663 strikes — 26% of the entire database — happen during approach. Birds are most active at low altitudes, exactly where descending aircraft spend their most vulnerable moments.

The FAA can't identify the bird in 37% of all strikes

Over 121,267 incidents list the species as "unknown bird." When a bird hits a jet engine at 200 mph, there's often not enough left to identify. Better wildlife monitoring at airports could change this.

Mourning doves are more dangerous than eagles — here's the math

Large raptors get the headlines but mourning doves account for 16,236 reported strikes. Small birds at high frequency cause more total damage than rare large-bird events. Volume beats size.

Our AI found a cluster of incidents nobody has named before

Machine learning on all 316,000 records identified four distinct strike profiles. One cluster — small aircraft at low altitude — shows a disproportionate damage rate that standard reporting misses entirely.

The Culprits

The 15 bird species most likely to ruin your flight, ranked by how often they show up in FAA reports

Species frequency · FAA Database 1990–2025 · Known species only

The humble mourning dove — a bird you probably see in your backyard every morning — is the single most frequently identified species in the database with 16,236 recorded strikes. The American kestrel, barn swallow, and killdeer round out the top five known species.

Large birds like eagles and vultures are genuinely dangerous when they hit, but they're rare. The real volume comes from small, abundant species that flock near airports in huge numbers.

When It Happens

Landing is when you're most at risk. This chart explains exactly why.

Strikes by phase of flight · All 316,839 incidents

The Sully lesson

The 2009 "Miracle on the Hudson" happened during climb — the second most dangerous phase. Captain Sullenberger's plane hit a flock of Canada geese at 2,818 feet. The data shows climb accounts for 28,848 strikes in the database — nearly 1 in 10 of all incidents.

The Science

We built an AI that can predict whether a bird strike will damage your plane. It's accurate 82.6% of the time.

Machine learning regression model · Trained on 163,832 historical incidents

Using the same type of AI that powers fraud detection at banks, we trained a model on 35 years of FAA data. Given what we know about the aircraft, the airport, and the phase of flight — can we predict how badly a wildlife strike will damage a plane?

The answer, surprisingly, is yes. Our best model explains 82.6% of the variation in damage outcomes — far better than chance. And the most important predictor wasn't the bird's size, or the airport, or even the weather.

It was one simple question: was damage indicated immediately after the strike?

82.6%

Model accuracy (R²)

Explains variance in damage outcomes

0.25

Average prediction error

On a 0–4 damage scale

Models tested

Linear, Ridge, Lasso, Random Forest

"The fact that a simple linear model explains 82% of damage outcomes suggests the system is far more predictable than the aviation industry lets on."

— Runway Risk Data Team · Georgia Tech Analysis, 2026

The Hidden Patterns

AI found 4 types of wildlife strike incidents hiding in 35 years of data. One is far more dangerous than reported.

K-Means clustering · 204,790 incidents · 4 distinct risk profiles identified

When you feed 200,000 wildlife strike records into a machine learning algorithm and ask it to find natural groupings — with no instructions about what to look for — what does it find?

Four distinct incident types emerge. Three are roughly what you'd expect. The fourth is surprising: a cluster of small aircraft at low altitude with a damage rate that's disproportionately high relative to how rarely these incidents make the news.

90K

General aviation, minimal damage

The largest group. Smaller aircraft, rural areas, grasslands. Most strikes in this cluster cause no damage and go unreported in mainstream media.

83K

High-traffic commercial airports

Major hubs during takeoff and landing. Denver, DFW, O'Hare. High frequency but relatively well-managed thanks to dedicated wildlife programs.

17K

Commercial jets, confirmed damage

The 8% of incidents involving heavy commercial aircraft where damage was confirmed immediately. These are the ones that make it onto your news feed.

15K

Small aircraft, underreported risk ⚠️

The cluster the data didn't expect. Low-altitude small aircraft with a damage confirmation rate that suggests significant underreporting in general aviation.