Day 1Chick mass — kgFed — % of need (max 100)Commute —Trips/day —Climb — m/sFPS 0

Drag to orbit, wheel / pinch to zoom. The scene is compressed: the real commute is kilometres long, so soaring birds loop over the colony while the numbers use the true distances. Chick size follows the simulated mass; the fish ball shrinks as the fishing group herds it.

Pelican Island Colony: Thermal Soaring, Group Fishing and Chick Provisioning

Do the fish that parents bring home across the lake feed the chicks enough? Soaring, group size and distance decide.

Thermal and glider

Fishing group

Colony and trips

Time

Live readouts

Climb rate vs bank angle

Catch per bird vs group size

Chick mass over the season

Independent numerical checks

Controls

Drag / one finger orbits, wheel / pinch zooms; the view buttons jump to the nests, the fishing group or the thermal column.
Time scale = simulated days per real second. Sliders change the model immediately; presets set several at once.
Group size is the number of pelicans herding the fish together (also the number swimming in the scene).

The science

Glide polar. Drag coefficient C_D = C_D0 + k C_L², k = 1/(π e AR). Sink rate w_s = ρ v³ S C_D / (2 m g); in a banked turn the lift coefficient is C_L = 2 n m g / (ρ S v²) with load factor n = 1/cos φ, so sink grows roughly as n^3/2. Best glide L/D = 1/(2√(C_D0 k)); minimum sink at C_L = √(3 C_D0/k).

Thermal climb. Updraught w(r) = w₀ (1 − (r/R)²) inside radius R. Circling radius r = v²/(g tan φ): steeper bank → smaller radius → stronger lift but more sink. Climb c = w(r) − w_s is maximised over airspeed and bank (grid search then refinement; the page cross-checks with a second method). A real thermal is turbulent and asymmetric; the parabolic profile is a textbook simplification.

Cross-country speed. Climb at c, then glide at airspeed v with sink s(v): average speed V = v / (1 + s(v)/c) (still air), maximised over v ("speed to fly"). Wind and thermal spacing are not modelled.

Flapping. Pennycuick-type: induced + parasite + profile power over muscle efficiency 0.23, plus basal metabolism BMR = 3.79 m^0.723 W; soaring is taken as 2×BMR (illustrative). The commute energy for flapping vs soaring is the difference the page reports.

Group fishing. Phenomenological: catch per bird Q(n) = Q₀ (1 + β (1 − e^{−(n−1)/n₀})) / (1 + γ (n − 1)): herding raises fish density (benefit saturates) while neighbours interfere (cost grows). The optimum n* is found numerically; the fish-ball radius in the scene scales as the cube root of the herding gain. These constants are illustrative, not fitted to a species.

Colony halo. Many pairs deplete nearby fish (Ashmole's halo): effective distance D = D₀ + 0.15 √N km.

Chicks. Daily demand 0.35 M^0.75 kg fish per chick. Parents deliver (trips × catch − own need 1.06 kg/day) split across the brood. Growth is logistic dM/dt = r M (1 − M/M_max) scaled by the fed fraction; a long deficit shrinks mass slowly. Fledging threshold 4.0 kg at day 70 (illustrative).

No predation, disease or weather beyond the thermal strength is modelled. Fish are drawn bigger than life and welfare of the animals is the focus: the page shows what fails (starved chicks) as a modelled outcome only. Models (pelican, nests, chicks, fish, islands, lake) built with Blender and rendered with Three.js.