Presets
Playback
Parameters
Camera
Layers
Quality
Model check
The science
What is modelled
A Eurasian lynx (18 kg) stalks a snowshoe hare across snow. Three pieces of physics decide the hunt. First, the large furry paws spread the weight, so the lynx sinks only part of the way into the snow, and compacting snow costs energy. Second, every footfall makes noise: the quieter and the slower the lynx walks, and the deeper the snow that muffles it, the closer it can get before the hare hears it. Third, the final burst is limited by traction and by muscle power, and the lynx wins only if it reaches the hare before the hare gets to cover. The page repeats the hunt with hares of different hearing and shows the resulting catch probability.
Equations
- Snow, Bekker law with exponent 1: pressure
p = k z. A paw of area A carriesp = m g / (n A)with n paws on the ground (3 while stalking, 2 in the burst), so the sinkage isz = min(H, p/k), never more than the snow depth H. The work of compacting the snow under one paw isW = ∫ p A dz = k A z² / 2; four footfalls per stride give a resisting forceFsnow = (4/stride) W. - Stalking cost per metre:
C0 m + (4 / 0.8 m) W / ηwithC0 = 4.3 J/(kg m)(a standard order of magnitude for a mammal of this size on bare ground) and a muscle efficiencyη = 0.25. - Footstep noise heard at distance r:
L(r) = L1 − 20 log10 r − α r(spherical spreading plus 0.01 dB/m of air absorption) withL1 = 48 + 30 log10(v / 1 m/s) − 0.15 H[cm]dB. A hare with hearing threshold θ notices the lynx inside the distance whereL(r) = θ; thresholds vary between hares as a normal distribution (mean 20 dB, standard deviation 3 dB), sampled at 41 evenly spaced quantiles. - Burst:
m dv/dt = [min(m amax, P/v) − m g μ − Fsnow − ½ ρ CdA v²] (1 − (v/vcap)4)withamax = 9 m/s²(traction),P = 1000 W(peak propulsive power),μ = 0.05, and a speed cap that falls as the paws sink,vcap = 14 m/s √(1 − z / 0.6 m)(top speed of similar animals scales with the square root of the effective leg length). The top speed is the smaller of the cap and the solution ofP = F(v) v. The hare reacts after 0.25 s, accelerates at 8 m/s² to 10 m/s and runs straight to cover; the lynx wins if the gap closes first, within a limit of 15 s.
Checks run in the page
- Sinkage from the Bekker law and the compaction work against its numerical integral; the sinkage cap at shallow snow.
- Spherical spreading (6.02 dB per doubling), the detection distance as a root of the level equation, and the 30 log10 speed scaling.
- Burst top speed from the cubic power balance against time stepping, and the traction-limited start against constant acceleration.
- Chase logic: with constant speeds the capture limit equals
vl tr + dc (vl/vh − 1). - The paw area and length of the Blender lynx against the numbers used in the model.
Honest simplifications
- The constants are representative values, not measurements of lynx: the paw area (110 cm², about 40 g per cm² of loading for a walking lynx), the snow stiffness, the noise level and its snow absorption, the hearing thresholds, the propulsive power, the speed cap and the hare speed are plausible orders of magnitude chosen to show the trade-offs. Real success rates of lynx hunting hares depend on terrain, cover, hare density and many things not modelled.
- The model is one-dimensional: the animals run on a straight line, the hare flees directly away and the lynx never takes a curved approach. Sinkage is the same in the whole hunt, the snow is uniform, and no crust layers or drifts are modelled.
- The hare hears only the footstep noise; vision, wind, scent and the lynx's use of cover are left out. The lynx sees the hare at all times.
- The gaits are animated for looks (leg swing and a body bob) and are not computed from the legs' mechanics.
- This is a predator-prey educational model; no animal is harmed and the page shows one hare at a time at a distance.
What to try
- Raise the stalking speed from 0.3 to 2 m/s and watch the catch probability fall as the noise rises (the first graph). The second graph shows a window: a burst started farther than a few metres never reaches the hare before cover.
- Switch to "Deep powder snow": the paws sink and the top burst speed falls, so the best burst distance moves closer.
- Press "Run 400 hunts" to estimate the catch probability by sampling random hares.