Grass Lake
Grass Lake · Minneapolis · Sat Oct 3, 2026 · 11:49 PM

Truck into Grass Lake: a reconstruction

The truck's path off the end of W 60th St, played against audio a neighbor's security system recorded that night. Ground heights come from USGS lidar; the flight is plain projectile physics. Move the sliders to see where other speeds would have put it.

Section along W 60th St, looking northheights ×3 · truck drawn 2× size · feet above sea level
Crash replay: curb to resting spot, true scale, looking north
Plan: north is updistances in feet
pathshoreline◯ modeled landing■ where the truck rests (photo fix, ±30 ft)truck drawn 3× size; shadow shows it airborne
23:48:53.0CDT At rest

Quiet until about 11:49:23. Turn your volume up; the crash is loud.

loudnesscurb strike (anchor) and modeled splash3rd burst: nose-down to flat
Lands—
vs. where it rests—
Time in the air—
Peak height over curb—
Modeled splash—
Loudest sound on tape23:49:29.35

What the model assumes

  • The truck hits the curb at 23:49:27.2, the peak of the first loud burst, and leaves the ground there.
  • It flies freely until it meets the ground or the water. No air drag; over 2 seconds that's a small effect.
  • The curb, then ~20 ft of flat boulevard at 860 ft, then the bank drops 29 ft to the water at 830 ft.
  • Coming off a downhill street onto flat ground and a 6-inch curb tips the nose up; the takeoff angle stands in for that.
  • The third burst (23:49:31.15) is the truck dropping from nose-down to flat in the water, a second splash.
  • Sound reaches the recorder from the curb and from the landing spot with nearly the same delay, so the 2.2 s gap between bursts is close to the real gap.

What fits

Only one combination lands the truck where it sits and makes the splash line up with the loudest sound: about 75 mph with a 12° takeoff. At 75 mph it would still be in the air for 2.2 seconds and peak about 8 ft above the curb. Slower than about 38 mph, it can't clear the bank at all.

A truck that skimmed after landing would have landed short of where it rests, which would mean it was going faster, not slower. Either way, these numbers are a reconstruction, not a measurement.